Citation
Exploring the nexus of infrastructures, environment, and health in Indian cities

Material Information

Title:
Exploring the nexus of infrastructures, environment, and health in Indian cities integrating multiple infrastructures and social factors with health risks
Creator:
Sperling, Joshua B. ( author )
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (281 pages). : ;

Subjects

Subjects / Keywords:
Health planning -- India ( lcsh )
Social conditions -- India ( lcsh )
Health planning -- Delhi (India) ( lcsh )
Social conditions -- Delhi (India) ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
The overarching goal of this thesis is to explore and assess infrastructure-environment-health interactions in Indian cities, addressing social factors such as wealth and literacy, as well as the provision of multiple infrastructures. Five main studies are conducted. First, exploration of Delhi all-cause mortality data and survey of local experts on associations between infrastructures, environment, and health outcomes. Key findings include: a) that 50% of deaths in Delhi are reported with cause not classified (demonstrating the need for bottom-up study to supplement hospital data) and b) that ~19% of classified deaths by cause in Delhi, India could be related to infrastructure or infrastructure-related environmental factors. Second, review of epidemiology studies relating health outcomes to infrastructure and pollution exposure in Indian and Asian cities is conducted to help identify initial evidence and gaps for infrastructure-related health effects and quantification of differential risk based on social factors (e.g. low socio-economic status (SES)). Third, top-down analyses using national survey of under age-five mortality rates (U5MR) by multiple infrastructure conditions are studied while addressing confounding social factors. A key finding is that the relative risk for under-five mortality rates are 860% higher in Urban India for those lacking multiple basic infrastructure provisions relative to improved conditions for low SES condition and limited literacy households. These analyses demonstrates limited literacy household sensitivity and importance of considering multiple infrastructures together over single infrastructure improvements. Fourth, bottom-up comparative community study helps characterize infrastructure, environment, extreme weather conditions and local sustainability priorities. A key finding was that households deprived of infrastructure provisions would prioritize that first over pollution or extreme weather conditions. In addition, both low SES communities studied were different in their coverage of all infrastructures except cooking fuels. In the high SES area, infrastructure conditions were ranked as a highest priority (e.g. drainage) with pollution and climate-related extreme weather events still higher priorities than low SES areas, which selected water supply, parks and open space, and drainage as highest priorities. Multiple dimensions of access to healthcare conditions in the same neighborhoods were explored next with findings indicating the two low SES areas to have similar travel costs to reach care and different abilities to pay for care. The high SES area also had higher accessibility to care yet with quality of care less acceptable relative to low SES areas that had issues with wait times, affordability, and access- suggesting future study should address such factors and effects on health outcomes. Finally, data availability, needs, and challenges are explored for computing health benefits of multiple infrastructure interventions, while also identifying preliminary intervention scenarios and who may benefit more or less by age, gender, and SES. These efforts offer a preliminary approach to helping prioritize future decision-making in Asian cities by demonstrating initial methods that can be useful for modeling risks and interactions between infrastructure provisions, environment, and health.
General Note:
Thesis (Ph.D.)--University of Colorado Denver. Civil engineering
Thesis:
Includes bibliographic references.
System Details:
System requirements: Adobe Reader.
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Joshua B. Sperling.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
900879819 ( OCLC )
ocn900879819

Downloads

This item has the following downloads:


Full Text
EXPLORING THE NEXUS OF INFRASTRUCTURES, ENVIRONMENT, AND
HEALTH IN INDIAN CITIES: INTEGRATING MULTIPLE INFRASTRUCTURES
AND SOCIAL FACTORS WITH HEALTH RISKS
By
JOSHUA B. SPERLING
B.S., Civil Engineering, University of Colorado Boulder, 2007
M.Eng., Environmental and Sustainability Engineering, Univ. of Colorado Denver, 2010
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Civil Engineering
2014


This thesis for the Doctor of Philosophy degree by
Joshua B. Sperling
has been approved for the
Department of Civil Engineering
by
Anu Ramaswami, Advisor
Jo Ann Silverstein, Chair
Wes Marshall
Debbi Main
Siddharth Agarwal


Sperling, Joshua B. (Ph.D., Civil Engineering)
Exploring the Nexus of Infrastructures, Environment, and Health in Indian Cities:
Integrating Multiple Infrastructures and Social Factors with Health Risks
Thesis directed by Professor Anu Ramaswami.
ABSTRACT
The overarching goal of this thesis is to explore and assess infrastructure-
environment-health interactions in Indian cities, addressing social factors such as wealth
and literacy, as well as the provision of multiple infrastructures.
Five main studies are conducted. First, exploration of Delhi all-cause mortality
data and survey of local experts on associations between infrastructures, environment,
and health outcomes. Key findings include: a) that 50% of deaths in Delhi are reported
with cause not classified (demonstrating the need for bottom-up study to supplement
hospital data) and b) that -19% of classified deaths by cause in Delhi, India could be
related to infrastructure or infrastructure-related environmental factors.
Second, review of epidemiology studies relating health outcomes to infrastructure
and pollution exposure in Indian and Asian cities is conducted to help identify initial
evidence and gaps for infrastructure-related health effects and quantification of
differential risk based on social factors (e.g. low socioeconomic status (SES)).
Third, top-down analyses using national survey of under age-five mortality rates
(U5MR) by multiple infrastructure conditions are studied while addressing confounding
social factors. A key finding is that the relative risk for under-five mortality rates are
860% higher in Urban India for those lacking multiple basic infrastructure provisions
relative to improved conditions for low SES condition and limited literacy households.
m


These analyses demonstrate sensitivity of limited literacy households and importance of
considering multiple infrastructures together over single infrastructure improvements.
Fourth, bottom-up comparative community study helps characterize
infrastructure, environment, extreme weather conditions and local sustainability
priorities. A key finding was that households deprived of infrastructure provisions would
prioritize that first over pollution or extreme weather conditions. In addition, both low
SES communities studied were different in their coverage of all infrastructures except
cooking fuels. In the high SES area, infrastructure conditions were ranked as a highest
priority (e.g. drainage) with pollution and climate-related extreme weather events still
higher priorities than low SES areas, which selected water supply, parks and open space,
and drainage as highest priorities. Multiple dimensions of access to healthcare conditions
in the same neighborhoods were explored next with findings indicating the two low SES
areas to have similar travel costs to reach care and different abilities to pay for care. The
high SES area also had higher accessibility to care yet with quality of care less acceptable
relative to low SES areas that had issues with wait times, affordability, and access-
suggesting future study should address such factors and effects on health outcomes.
Finally, data availability, needs, and challenges are explored for computing health
benefits of multiple infrastructure interventions, while also identifying preliminary
intervention scenarios and who may benefit more or less by age, gender, and SES.
These efforts offer a preliminary approach to helping prioritize future decision-
making in Asian cities by demonstrating initial methods that can be useful for modeling
risks and interactions between infrastructure provisions, environment, and health.
The form and content of this abstract are approved. I recommend its publication.
Approved: Anu Ramaswami
iv


To my wife, Ariel, family, friends and mentors whom have all given me a lifelong
appreciation for learning, exploring, and making a difference in this world.


ACKNOWLEDGMENTS
I want to acknowledge my advisor, Prof. Anu Ramaswami and committee
members including Prof. Debbi Main, Prof. JoAnn Silverstein, Prof. Wes Marshall, and
Dr. Siddharth Agarwal for their insightful guidance, encouragement, and support in
building an interdisciplinary and international professional network. I especially want to
thank Dr. Ramaswami, my primary advisor, for offering extensive professional training,
mentorship, and insightful supervision throughout my doctoral study. Im very
appreciative for all her time, contribution of ideas, and mentorship.
I would also like to thank the NSF IGERT Sustainable Urban Infrastructure
program at University of Colorado Denver for supporting this work, providing initial
funding and the opportunity to conduct international research with the local assistance of
Dr. Agarwal, Dr. Tapan Kalita, Dr. Ajay Nagpure and Navneet Baidwan for community-
based fieldwork in India. This research was also supported and influenced by the United
States India Fulbright-Nehru fellowship and Indo-US Science and Technology Forum:
Research in Science and Engineering Fellowship in India (thanks to my research hosts,
Dr. Agarwal at Urban Health Resource Centre and Dr. Singhal at TERI University
Department of Policy Studies, Delhi, India), the NSF Research Coordination Network on
Sustainable Cities and NSF Partnerships for International Research and Education. Each
opportunity has provided a unique platform and network for building on acquired
knowledge and methods for continued learning and study, locally and globally.
The UC-Denver IGERT program faculty, coordinators, students as well as
colleagues, family, and friends in diverse localities also enriched this journey. I hope to
build on these various professional and personal relationships for a long time to come.
vi


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION.........................................................1
Research Objective.................................................1
Rationale..........................................................1
Key Question, Unique Features of Indian and Asian Cities, and Contributions .... 5
Dissertation Organization..........................................8
Key Contributions: Linking Multiple Infrastructure Sectors, Inadequate
Infrastructures, and Urban Health.................................10
II. EXPLORING HEALTH OUTCOMES AS A MOTIVATOR FOR LOW-CARBON
CITY DEVELOPMENT: IMPLICATIONS FOR INFRASTRUCTURE
INTERVENTIONS IN ASIAN CITIES...........................................12
Abstract..........................................................12
Introduction......................................................13
Literature Review.................................................15
Methods...........................................................22
Results...........................................................26
Discussion: Implications for Infrastructure Interventions.........31
Conclusion........................................................34
Acknowledgments...................................................35
Appendix. Chief registrar for NCT of Delhi Re-analysis of births and deaths
2008 report for Delhi.............................................36
III. A REVIEW OF EPIDEMIOLOGICAL STUDIES RELATING TO
INFRASTRUCTURE AND POLLUTION IN ASIAN CITIES: EVIDENCE AND GAPS
FOR INFRASTRUCTURE-RELATED HEALTH EFFECT ESTIMATES IN DELHI. 35
Abstract..........................................................35
vii


Introduction
36
Methods..................................................................37
Literature Review........................................................38
Results: Risk Factors and Health Effect Estimates........................38
Discussion: Implications for Infrastructure-Related Health Interventions.57
Conclusion...............................................................62
Appendix. Health Determinants of a Population and To Which Factors Health
Effect Estimates are Most Sensitive......................................63
IV. UNMASKING THE ROLE OF MULTIPLE INADEQUATE BASIC
INFRASTRUCTURES IN INFLUENCING UNDER FIVE MORTALITY IN ALL-
INDIA, URBAN INDIA & DELHI......................................................64
Introduction.............................................................65
Literature Review........................................................68
Methods..................................................................71
Results..................................................................75
Discussion...............................................................81
Appendix A. Delhi Analyses and All-India Principal Component Analysis....94
Appendix B. All-India DHS Analyses: Progressions by Wealth Quintiles.96
Appendix C. Diarrheal Incidence Analyses.................................97
Appendix D. Additional DHS Re-Analyses for Six Indian Cities, DHS Urban vs.
Rural India, and Delhi vs. All-India.....................................98
V. INFRASTRUCTURE CONDITIONS AND SUSTAINABILITY PRIORITIES IN
ASIAN CITIES: A COMPARATIVE STUDY OF THREE NEIGHBORHOODS IN
DELHI, INDIA...................................................................101
Abstract............................................................101
Introduction: Why Asian Cities, Civil Infrastructures & Local Priorities.102
Objectives..........................................................105
Review of Surveys on Global-to-Local Priorities.....................106
viii


Methods..................................................................108
Results and Discussion...................................................114
Conclusions..............................................................122
Appendix A. Delhi District-level Infrastructure Condition (Census India, 2011)
.........................................................................123
Appendix B. Preliminary Characterization using Survey, Transects, and a
Preliminary Descriptive Matrix of Physical Infrastructure and Social
Infrastructure...........................................................125
Appendix C. Prioritization Visual Used for Survey........................127
Appendix D. Survey Visuals: Levels of Satisfaction and Inconvenience.....128
Appendix E. Assessing Levels of Satistfaction and Inconvenience Experienced
.........................................................................129
Appendix F. Initial Analyses Discussion by Individual Infrastructures....130
Appendix G. Household Survey Instrument Used.............................147
VI. A FOCUS ON EXPERIENCES WITH AND BARRIERS TO ACCESS TO
URGENT HEALTHCARE IN THREE NEIGHBORHOODS OF DELHI, INDIA.......................174
Abstract.................................................................174
Introduction.............................................................175
Rationale................................................................176
Literature Review and Defining Access....................................177
Methods..................................................................180
Results..................................................................183
Discussion...............................................................191
Conclusions..............................................................191
Appendix A. Supplementary Analyses.......................................192
Appendix B. Consent Form and Household Survey Instrument Used............200
IX


VII. WHAT IS KNOWN AND WHAT IS NEEDED TO ESTIMATE HEALTH
BENEFITS OF INFRASTRUCTURE INTERVENTIONS: CASE STUDY OF DELHI,
INDIA......................................................................209
Abstract............................................................209
Introduction........................................................210
Literature Review...................................................211
Results of First-Order Computation..................................217
Discussion: Synergies in Sustainable Infrastructure Development and Urban
Health..............................................................223
Conclusion..........................................................224
VIII. SUMMARY & CONCLUSIONS...............................................227
REFERENCES.................................................................232
Chapter 1.......................................................2322
Chapter 2........................................................233
Chapter 3........................................................239
Chapter 4........................................................253
Chapter 5........................................................257
Chapter 6........................................................260
Chapter 7........................................................262
x


LIST OF TABLES
Table
11.1 Review of Health Benefits and Risks of Infrastructures.......................16
11.2 Tracing Health Outcomes to Infrastructure and Environment-Related Factors....19
11.3 Short-Term (24-hour mean) Environmental Exposure Guidelines and Delhi
Compliance........................................................................21
11.4 Comparative Health Indicators for Delhi, Mumbai, and India...................27
II. 5 Summary of Local Expert Opinion on 2008 Deaths..............................32
III. 1 Available Health Effect Estimates for CVD and Total Mortality Risk.........41
111.2 Available Health Effect Estimates for Respiratory Mortality.................45
111.3 Available Health Effect Estimates for Airborne Infectious Disease Mortality.48
111.4 Available Health Effect Estimates for Road Safety...........................49
111.5 Available Health Effect Estimates for Diabetes..............................50
111.6 Available Health Effect Estimates for Waterborne Disease Pathogens..........53
111.7 Available Health Effect Estimates for Cancer................................56
111.8 Study Key Findings and Identified Knowledge Gaps............................58
IV. 1 Review of Studies In Past Ten Years Using the DHS: 2002-2012............... 84
IV.2 Exploring Correlations: All-India Pearsons correlation (r) Matrix With U5M
Incidence.........................................................................86
IV.3 Correlation Matrix: Aggregate Analyses where U5MRs are computed for Wealth
Quintiles (n = 5) and Plotted Against Changes in Other Variables Due to Wealth....87
IV.4 Summary of U5MRs and RRs for a Proxy Attribute of Access to Healthcare.......87
IV. 5 Summary of All-India U5MR, IMR, NMR, U5MR-ND Analyses for Presence and
Absence of Infrastructure Combinations............................................93
V. l Review of Small Sample of Example Surveys on Global and Local Priorities...108
xi


V.2 Example Analyses on Time to Fetch Water and Motor Vehicle Ownership
V.3 Summary of NM & BJ T-Test Results for Key Infrastructure Conditions...
116
118
VI1 Review of Studies Characterizing Access to Healthcare and Providing Quantitative
Findings in Terms of Experiences with Access to Healthcare............................179
V. 2 Access to Healthcare: Costs to Reach Care and Affordability of Urgent Care......188
VI. 3 Comparing Qualitative Responses on Knowledge and Experiences With Health
Facilities along with Length of Residence.............................................189
VII. 1 What We Know and What is Needed to Explore Health Benefits of Infrastructure
for the Case of Delhi, India..........................................................212
VII.2 Literature Review Integrating Multiple Infrastructure-Related Health Risk
Categories and Health Effect Estimates.................................................213
VII.3 Summary of more health effect estimates useful for intervention scenarios.......216
Xll


LIST OF FIGURES
Figure
1.1 Unique Features of Indian Cities and the Case of Delhi, India...................6
1.2 WHO Framework Model on Social Inequalities and Environmental Risks..............8
II. 1 Mortality by Age: Delhi (2008)................................................27
11.2 Mortality by Cause (if classified) in the NCT of Delhi (2008)..................28
11.3 Estimating Air Quality-Related Cardiovascular & Respiratory Mortality Reduction in
Delhi...............................................................................30
III. 1 Example Relative Risk Findings for Infant and Child Mortality................52
111.2 WHO Framework Model on Social Inequalities and Environmental Risks............59
111.3 A preliminary schematic representation / conceptual diagram for study of the nexus
of urban infrastructures, environment and public health.............................60
111.4 A revised schematic diagram for studying infrastructures, environment-climate, and
health with consideration of biophysical and socio-biological risk factors..........61
IV. 1 Initial Exploration of Infrastructures, Health/Banking/Education, and Infant / <5
Mortality Per 1000 births for Slum vs. Non-Slum (DHS/NFHS-3)........................83
IV.2 Infrastructure deficiencies: common in all Indian cities, improve with SES.....84
IV.3 Example U5MRs by Wealth and Literacy from Lowest to Highest (India & Urban-
India)..............................................................................88
IV.4 Flow Charts for India and Urban India: Controlling for Wealth and Literacy.....89
IV.5 India and Urban-India Multiple Infrastructure Presence / Absence vs. U5MR......90
IV.6a RR of Infrastructure Condition Combinations: India (top) v. Delhi (bottom)....91
IV. 7 Breakdown of All-India Neonatal, Infant, and Remaining Under-5 Mortality.....92
V. l % of Births as Institutional Deliveries by Delhi Districts...................110
V.2 Northeast (NE) Delhi Study Neighborhoods on East side of Yamuna.................Ill
xiii


V.3 Closest Nearby Air Quality Station to NE District (West of Yamuna River).......112
VIA Socio-Economic Factors By Study Area: Avg. Monthly Expenditures................113
V.5 Study Neighborhoods Infrastructure Conditions Snapshot......................113
V.6 Summary of Results in Study Areas Compared with NFHS-3, 2006 Indicators 117
V.7 Summary of Results from the Three Study Areas.................................117
V. 8 Percent of Total Votes for Top Three Priorities Among Infrastructures, Environment,
and Climate-Related Extreme Weather Events by Study Area..........................121
VI. 1 Geographic Distribution of % Slum population to Total Population (Census, 2001)
and Births as Institutional Deliveries by Delhi Districts (DLHS, 2008)............ 177
VI.2 Physicians Per 100,000 population by Country and Life Expectancy vs. Physicians
Per Capita by Country (UC Atlas of Global Inequity and WHO)........................180
VI.3 % of Births as Institutional Deliveries by Delhi Districts...................181
VIA Socio-Economic Factors By Study Area: Avg. Monthly Expenditures................185
VI. 5 Comparing Self-Reported Access to Healthcare in Three Neighborhoods.........186
VII. 1 Example Analysis using Cropper et al. HEE from Thesis Chapter II...........218
VI.2 Rajarathnam et al. HEE for N02 and PM 10, with PM 10 by Age and Gender.......220
VIA Thesis Ch. IV HEE Initial Application: Multiple Basic Provisions to Low SES Delhi
Households Lacking Basic Provisions (Using DHS Survey and Delhi Data)..............222
VIA First-Order Computation of Reduced Traffic Fatalities.........................223
VIII. 1 Schematic Representation of a Preliminary Analytical Framework............229
xiv


LIST OF EQUATIONS
Equation
III. 1 Estimating Excess Mortality..........................................25
IV. 1 Computing Under-Five Mortality Rates (U5MR)...........................72
IV.2 Computing Relative Risk of Under-Five Mortality Rates..................73
xv


CHAPTER I.
INTRODUCTION
Research Objective
The overarching goals of this thesis are to explore and assess infrastructure-
environment-health interactions in Indian cities. More specifically, this thesis seeks to 1)
explore the extent to which multiple civil infrastructures (e.g. water, sanitation, energy
infrastructures) and related environmental factors (e.g. air and water quality, extreme
events) can shape health outcomes and 2) begin to understand these interactions while
also addressing issues of socioeconomic conditions, literacy, and access to healthcare.
Insights are used to begin modeling the multiple risks and linkages between
infrastructure provisions, environmental conditions, and health. Efforts using the case of
Delhi, India offer a preliminary approach to helping inform future engineering, planning,
and policy decision-making on infrastructure development for health benefits in Indian
cities such as Delhi, India specifically by presenting an improved knowledge base on how
infrastructures can shape health risks and benefits, and, what data is available and needed.
Rationale
Asian Cities will Dominate Future Urbanization and Global GHG emissions:
Seven billion people now live on this planet (UN, 2011) and over half of humanity now
lives in cities (UN, 2007). In the next 20 years, nearly 60 percent of the worlds people
will be urban dwellers, with rapid growth in Asian cities. Trends suggest urban
populations are expected to soar to five billion from more than three billion today and
this will include 60% of Chinas population (current 46% urban), 41% of Indias
1


population (currently 30%), and 87% of the USA population (currently 82% urban) by
2030 (Ramaswami & Dhakal, 2011). Meanwhile, these same countries are estimated to
contribute -46% of global C02 emissions from energy use (IEA, 2010) and -47% of
global greenhouse gas (GHG) emissions (UNFCCC, 2010). GHG emissions are being
generated from various engineered infrastructure sectors serving cities such as
transportation, energy generation / distribution, buildings, water supply, waste /
wastewater management, telecommunication and industrial production (Chavez &
Ramaswami, 2013; Ramaswami, 2013).
Inadequate Infrastructures. Urbanization, and Health in Asian Cities: The current
infrastructure conditions in Asian cities are often quite poor. In Delhi, India, for example,
estimates suggest -40-50% of the population are living in slums or slum-like conditions,
which the United Nations defines as households that lack access to improved water,
sanitation, sufficient-living area, durability of housing, and security of tenure (UN, 2007).
Currently, 16% of households lack access to drinking water taps (putting residents at risk
of waterborne illnesses), 6% lack access to latrines, and 8% use wood, dung, and charcoal
for cooking (MolID, 2009). In slum areas of Delhi, 19% lack water supply on premises,
76% lack an improved private toilet facility, 52% use kerosene or solid fuels, 35% lack a
pucca (solid structure) house, and 48% of these households have overcrowded sleeping
areas with more than 5 persons per sleeping room (Census, 2011). As a result, a large
proportion of the population may be suffering today due to current local conditions.
Infrastructure-Environment Interactions and Health in Asian Cities: While civil
infrastructures enable economic development in rapidly expanding and newly emerging
Asian cities, they also pose risks to the environment (e.g. pollution, resource depletion,
2


GHG emissions) that can impact public health in various ways. For example, Delhi
average pollutant concentrations can be up to four times higher than national outdoor air
quality standards for residential areas (putting residents at risk of cardiac and respiratory
problems), and up to 18 times higher than drinking water quality standards (Sperling and
Ramaswami, 2012). In addition, approximately 1.8 million people, primarily in India and
China, die prematurely from exposure to black carbon in combustion emissions (WHO,
2010). Environmental pollution concentrations and health risks seem to be trending in the
wrong direction with reports highlighting the rise of inadequacy of limited drainage
connections to wastewater treatment outlets (Census, 2011), electricity demand, which is
set to double from 2009 levels by 2015 leading to large increases in fossil-fuel burning
infrastructures (Central Electric Authority, 2013), motorization: roughly 1000 vehicles
added to Delhi roads daily (CSE, 2012), and health risks, with roughly one-third of all
medically certified causes of death due to cardiovascular and respiratory illness (NCT of
Delhi, 2008). In addition, surveys are suggesting current local human health risks from
infrastructure and environmental factors may be more important then future (uncertain)
climate considerations.
Public Health. Infrastructures, and Climate Change: High vulnerability of Asian
cities to current and future climate-related hazards (flooding, extreme heat), amplified by
poor socioeconomic and infrastructure conditions also exists. In fact, episodes of extreme
weather events such as extreme heat, drought, flooding, and infectious disease outbreaks
are already taking a toll on urban health and these impacts are expected to be further
exacerbated by global climate change and associated increases in climate variability
(Bush et al., 2011). Heat-related deaths for Delhi have been estimated to increase overall


deaths by 2% and future CVD-related deaths are estimated to rise by 4% per degree
Celsius increase in temperature greater than 20 degrees Celsius (95% confidence interval)
(Hajat, 2005). Flooding events are also impacting the lives of many Delhi residents,
especially slumdwellers who often live near open sewage drains- with the effect of
sewage flooding their homes every year during heavy rains, outbreaks of malaria,
dengue, and other vector and water-borne diseases (Aggarwal, 2013; De et al., 2013).
Infrastructure Priorities: A key thesis question is what do Delhi residents in
different socioeconomic strata prioritize in terms of infrastructures? Do they prioritize
provision of basic infrastructure that can improve health and livelihoods, clean up of
polluting infrastructures, or improved management of extreme weather events? Are there
infrastructure interventions that can address all three of these? What are the trade-offs?
Quantifying Infrastructure Pathways for Healthy and Low-Carbon Asian Cities:
Measuring both synergistic and antagonistic infrastructure pathways toward low carbon
and healthy cities is of importance. For example, providing clean drinking water systems
can place new energy demands on cities by requiring new water treatment plant facilities
and can result in increases in GHG emissions, meanwhile reducing waterborne diseases.
In addition, infrastructure upgrades such as moving from improperly ventilated cook
stoves burning solid fuels (dung, wood, et) to clean cook stoves and from commuting in
diesel vehicles to clean mass transit can have both health and climate benefits. As such,
this thesis defines and explore synergistic pathways achieving GHG emission reduction
and health benefits and antagonistic pathways that may increase health benefits, while
worsening GHG emissions or vice versa. Continued efforts in this area can help to
identify many tradeoffs across infrastructure sectors that have yet to be evaluated.
4


Evidence Base and Understanding of Priorities: Quantitative knowledge is
missing about how multiple inadequate infrastructures and lack of quality basic services
including water supply, sanitation, and access to urgent medical care can have important
health impacts. This thesis attempts to improve understanding of such knowledge gaps
for rapidly growing cities in Asia like Delhi, India, through quantitative data analyses. In
addition, local priorities are identified for planning, designing, and constructing of new
engineered infrastructure systems, which can be of significant importance to society as
they affect health outcomes today and may still be in place for many years to come.
Key Question, Unique Features of Indian and Asian Cities, and Contributions
This thesis will assist the beginning of understanding the key research question:
what are the interactions between inadequate and polluting infrastructures, environmental
conditions, and health outcomes in Indian and Asian cities and how are these interactions
modified by socioeconomic and access to healthcare conditions? Top-down studies using
secondary data and bottom-up studies using primary data collection efforts in different
Delhi neighborhoods are used to inform and help answer these questions. Both offer
important insights, some of which are used to compute health benefits of infrastructure
interventions; with other aspects of these studies used to inform understanding of local
conditions and realities.
In addressing this research question, several unique characteristics of Indian and
Asian cities are identified and explored, four of which are shown in the schematic below -
that they often a) lack adequate basic infrastructures, b) have high levels of pollution, c)
have limited causes of death records, and d) lack access to quality healthcare services.
5


Rationale: Unique Features of Indian Cities & the Case of Delhi, India
Multiple Inadequate Basic
High Levels of Pollution & Pop
Density Exposed to Pollution
Infrastructures
Delhi, India: 16% of households lack
access to drinking water taps, 6% lack
access to latrines, 8% use solid fuels for
Delhi pollution concentrations 4x higher
than nat'i outdoor air standards; up to
18x higher than drinking water quality
standards (BOD levels). (CPCB, 2010)
cooking (MoUD, 2009)
1
These unique features are interrelated and
critical to exploring the infrastructure-
environment-health nexus in Indian cities
Causes of Death Undocumented
Lack of Access to Quality
Healthcare
>66% of households in 8 Indian cities do
Even in larger Indian cities like Delhi, not utilize govt, health facilities due to lack
more than 50% of deaths occur with Qf accessibility 8. poor quality services
causes undocumented (NCT Delhi, 2008) (Goli, 201l); 9% of Delhi deaths by cause
are birth-related (Sperling & Ramaswami,
2013)
Figure 1.1 Unique Features of Indian Cities and the Case of Delhi, India
These unique features are important, interrelated, and integral to the thesis
research question for several reasons. First, with -50% of deaths occurring outside
institutions where cause of death records are maintained for the capital city of Delhi,
India, an incomplete understanding of mortality in Asian cities and its associations with
other risk factors will continue to be the case. Improving understanding as to how
inadequate infrastructures might inhibit access to quality urgent healthcare services may
be important if to improve both the care provided for serious health episodes and the
mortality records themselves. Second, Asian cities are faced with issues on both proactive
and reactive sides of health risk prevention lack of proactive health risk mitigation to
avoid inadequate and polluting infrastructures causing diarrhea, respiratory illness, and
road accidents to name a few; and lack of reactive risk mitigation of providing critical
care to those facing such serious illness. Third, exploring current mortality associated
with high pollution exposures, lack of infrastructures or climate-related hazards can only
be assessed to a certain extent if mortality records are inadequate and quality healthcare
access such an important issue that it may confound potential quantitative associations.
6


In order to address and improve understanding of these unique features, this thesis
makes four key contributions (shown from largest to smallest): (1) A first quantitative
data analysis in Delhi and Urban India examining correlation of under-five mortality with
lack of adequate basic infrastructures while controlling for the confounding effects of
access to healthcare, literacy, and socioeconomic status; (2) Bottom-up study of
infrastructure conditions, priorities, and preliminary health diary reports in three
neighborhoods of Delhi, India; (3) Characterizing experiences with and barriers to access
to urgent healthcare in these neighborhoods with the impact on health not yet able to be
determined due to differences in infrastructure conditions by neighborhood; and (4)
Using the integration of bottom-up and top-down data analyses, literature review, and
collected mortality data for Delhi -first-order health benefit computations are modeled
related to findings on infrastructure provision, environmental conditions, and health.
These contributions align with the WHO (2010) framework shown below, which
this thesis aims to utilize and build on specifically by exploring the roles of multiple civil
infrastructures, socioeconomic and environmental conditions in shaping health effects.
Characteristics of access to / quality of health services are also explored using top-down
secondary data re-analyses and preliminary bottom-up community primary data analyses.
7


Figure 1.2 WHO Framework Model on Social Inequalities and Environmental Risks
Dissertation Organization
This dissertation consists of eight chapters. First, Chapter 1 provides an
introduction to the research objectives, rationale, main research question, contributions,
organization, and proposed outputs. Chapter 2 provides a preliminary baseline assessment
that explores the extent to which civil infrastructures (i.e., water, sanitation, energy,
transport and building infrastructures) and environmental factors (e.g. air and water
quality) associated with these infrastructures shape current urban health outcomes in
cities in Asia using the case of Delhi, India. Analyses on current mortality data and a
preliminary survey of local expert opinion indicate up to 19% of all recorded deaths in
Delhi, India may be infrastructure-related. While preliminary, the findings suggest health
outcomes may be a large factor in motivating low-carbon development in Asian cities.
Chapter 3 includes literature review of existing epidemiology, infrastructure, and
environmental exposure studies and models, with emphasis on documentation of the
impact of social disparities as they modify health effects estimates and outcomes.
8


Evidence of associations between current urban health outcomes and infrastructure /
environment-related factors are reviewed to determine the state of evidence and gaps in
forming a database of health effect estimates specifically for Delhi, India and Asian cities
(while presenting additional studies where useful that are often global or from North
America or Europe). The goal of developing such a database is to begin developing first-
order computations of health benefits from infrastructure interventions (in Ch. 7). Gaps
identified in the literature inform where bottom-up primary data collection efforts and
community health studies are needed (e.g. access to healthcare impacts on health).
Chapter 4 then describes results from a top-down analysis of multiple civil
infrastructure conditions (including housing, drinking water, toilet facility, cooking
fuels), socioeconomic conditions (wealth and literacy), proxy attributes of access to
healthcare conditions (affordability and distance / transportation to healthcare facilities)
and under-five mortality rates while separating out confounding health determinants.
These analyses were conducted for Delhi, India; five additional Indian cities (Hyderabad,
Indore, Chennai, Mumbai, Kolkata), All-India, and urban-vs.-rural conditions.
Chapters Five and Six then present the results of a bottom-up community study in
India that aims to fill certain knowledge gaps and form baseline neighborhood-level
information for preliminary comparative assessment of current health, environment,
extreme weather, and access to quality urgent healthcare conditions for people living in
three neighborhoods of Delhi having different infrastructure and socioeconomic levels.
In Chapter Seven, results and new analyses are utilized for developing a first
order assessment of how different infrastructure intervention scenarios can mitigate
health (e.g. mortality) and climate (greenhouse gas emissions) risks, offering a
9


preliminary approach to help prioritize future decision-making in Asian cities. Finally,
Chapter Eight summarizes dissertation findings and proposes future research needs in
studying the infrastructure-environment-climate-health nexus in Delhi and other cities.
Key Contributions: Linking Multiple Infrastructure Sectors, Inadequate
Infrastructures, and Urban Health
The broad impacts of this thesis are two-fold: 1) quantitative risk assessment
linking multiple infrastructures; and 2) using place- and evidence-based data in cities to
study the infrastructure-environment-health nexus. While others have looked
quantitatively at energy and fossil-burning infrastructures, environment and health at
country and global levels (Pruss-Ustun and Corvalan, 2007), qualitatively at
infrastructure-environment-health in cities (Campbell-Lendrum and Corvalan, 2007), few
studies have attempted to quantitatively assess health risks of multiple infrastructures
(e.g. water, sanitation, energy, transportation, and housing) and in infrastructure-deprived
areas in Indian cities. Relevant literature and large datasets are analyzed and synthesized
across multiple infrastructures and health impacts that have often remained separate in
the past. Social and environmental determinants are also linked in Chapter 4 analyses. As
a result, this dissertation makes three original contributions:
(1) An Improved Knowledge Base Addressing Inadequate Infrastructure: study
provides quantitative analyses examining associations and correlations of under-
five mortality and inadequate infrastructure / housing conditions in Indian cities.
(2) A Place-based Study of the Infrastructure-Environment-Health Nexus: These
complicated but important relationships are investigated in top-down analyses and
bottom-up community studies of infrastructures, environment, socioeconomic
10


conditions, local priorities, and health outcomes within three neighborhoods of the
rapidly growing megacity of Delhi, India.
(3) New Lines of Inquiry Linking Multiple Urban Infrastructures with Health:
methods are developed to explore the key hypothetical question: if improving
health is a primary objective in Asian cities, what levels of health benefits may or
may not be achieved with infrastructure-related investments that focus on
improved health (some of which may also have GHG co-benefits)? This question
(or approach) addresses specific features of Asian cities (e.g. poor infrastructure
conditions, high levels of environmental pollution, lack of access to urgent
healthcare, and undocumented causes of death), and is partially motivated by the
transposed question of other recent research looking at the public health co-
benefits of household energy GHG reduction strategies (Wilkinson et al., 2009),
transport GHG reduction strategies (Woodcock et al., 2009), low-carbon
electricity generation (Markandya et al., 2009), food and agriculture GHG
reduction strategies (Friel et al., 2009), and reducing short-lived greenhouse gas
pollutants (Smith et al., 2009). Using thesis findings and other literatures, a
preliminary summary of what we know and what is needed to compute or
estimate health benefits from potential infrastructure intervention scenarios is
described for Delhi, India. Challenges to estimating health benefits for
populations and subpopulations by social factors, and linking to first-order
computation of GHG co-benefits for air pollution alternatives are also described.
11


CHAPTER II.
EXPLORING HEALTH OUTCOMES AS A MOTIVATOR FOR
LOW-CARBON CITY DEVELOPMENT: IMPLICATIONS FOR
INFRASTRUCTURE INTERVENTIONS IN ASIAN CITIES
Abstract
Sustainable urban infrastructure interventions can help achieve both public health
and low-carbon goals in cities. This paper explores the extent to which civil infrastructure
(i.e., water, sanitation, energy, transport and building infrastructures) and environmental
factors (e.g. air and water quality) associated with these infrastructures shape current
urban health outcomes in cities in Asia using Delhi, India as a case study. Current
mortality data for Delhi are used as context to estimate the extent to which urban health
outcomes are shaped by infrastructure and infrastructure-related environmental factors,
some of which could directly or indirectly reduce mortality through low-carbon
interventions. Mortality data along with a preliminary survey of expert opinion indicate
up to 19 percent of all recorded deaths in Delhi may be infrastructure-related. More
detailed epidemiology studies and infrastructure models are needed to confirm these
initial findings. The findings suggest public health outcomes may be a large factor in
motivating low-carbon development in Asian cities.
12


Introduction
Over half of humanity now lives in cities. By 2030, nearly 60 per cent of the
worlds people will be urban dwellers, with 60% of Chinas population living in cities
(currently 46% urban), 41% of Indias population (currently 30%), and 87% of the USA
population (currently 82% urban) (Ramaswami & Dhakal, 2011). Meanwhile, these same
countries are estimated to contribute a total of 46% of global C02 emissions from energy
use (IEA, 2010) and -47% of global greenhouse gas emissions (UNFCCC, 2010). These
greenhouse gas (GHG) emissions are being generated from various engineered
infrastructure sectors serving cities such as transportation, energy generation /
distribution, buildings, water supply, waste / wastewater management,
telecommunication, and industrial production (Hillman & Ramaswami, 2011).
Engineered infrastructures provide many benefits to society, including economic
development. At the same time, they pose risks to the environment (e.g., resource
depletion, groundwater depletion, GHG emissions) and can impact public health in
different ways. Infrastructure improvements such as moving from biomass fuel sources to
clean cookstoves or installing wastewater treatment facilities can improve public health.
On the other hand, fossil fuel combustion in current infrastructure can pose a risk to
health i.e., traffic congestion, air pollution, traffic accidents, and associated mortality.
Public opinion surveys in China and the US suggests that local environment and
public health concerns can be a more important motivator for low-carbon infrastructure
interventions than the abstract goal of mitigating climate change (Lo, 2010; PRC, 2010).
A few authors have discussed the relative role of carbon mitigation in the context
of local sustainable development priorities. One indicator for this has been the human
13


development index (HDI), introduced in the early 1990s as a new way of measuring
development by combining indicators for health (e.g. life expectancy at birth and under
age-five mortality), education (expected years of schooling), and living standards (gross
national income per capita) using a scale of 0 (worst) to 1 (best). Recent research on HDI
in relation to energy and carbon emissions demonstrate trends of high levels of human
development being achievable with certain energy and carbon emission minimum
thresholds, as well as shifting maximum thresholds. Several countries report stable and
high levels of HDI corresponding with both low (e.g., Japan and Costa Rica) and high
(e.g., USA) energy usage and carbon emissions. Other rapidly developing countries like
India and China are showing sharp increases in HDI for relatively small increases in
energy use and carbon emissions (Steinberger, 2009). These tradeoffs between HDI
versus energy use and carbon emissions have been evaluated in the context of local
sustainable development by Amekudzi (2011). Amekudzi, Ramaswami, Chan, Lam, and
Meng (2011) identify two types of low-carbon development objectives with Type 1 as
GHG mitigation as the primary objective (with other co-benefits as secondary objectives
such as water and energy savings); and type 2 as economic development or other
sustainable development priorities (e.g., health, reducing childhood mortality, etc) as the
primary objective with GHG mitigation as a secondary priority (or not a priority at all).
This paper asks the question if improving public health is your primary objective,
what level of GHG mitigation may be achieved (or not) as a co-benefit? This approach is
the transposed question of other recent research looking at the public health co-benefits
of household energy GHG reduction strategies (Wilkinson et al., 2009), transport GHG
reduction strategies (Woodcock et al., 2009), low-carbon electricity generation
14


(Markandya et al., 2009), food and agriculture GHG reduction strategies (Friel et al.,
2009), and reducing short-lived greenhouse gas pollutants (Smith et al., 2009).
The objectives of this study are to estimate the proportion of health outcomes that
may be related to infrastructure in order to identify whether urban health outcomes may
be a motivator for low-carbon infrastructure development in cities. Using Delhi, India as
a case study, the research question identified here is: to what extent does civil
infrastructure (i.e., water, sanitation, energy, transport and building infrastructures) and
environmental factors (e.g. air and water quality) associated with these infrastructures
shape current urban health outcomes in cities in Asia?
The paper is divided into three parts: literature review, preliminary city health
data analysis, and qualitative implications for infrastructure interventions. The
preliminary data analysis for Delhi includes gathering different types of public health
data and local expert opinion on the relationship of mortality to engineered infrastructures
and infrastructure-related environmental factors. The paper concludes with a preliminary
exploration and discussion of how infrastructure interventions can shape both health and
low-carbon goals. Such analysis provides a rationale and pathway for the twin goals of
developing healthy and low-carbon cities.
Literature Review
The literature review addresses two different perspectives: 1) The first
summarizes how individual infrastructure affects health, examining both direct and
indirect health benefits and risks; 2) The second approach traces observed negative global
health outcomes to infrastructure and infrastructure-related environmental factors.
15


Table 1 provides a summary of the first approach including thirty publications
describing health benefits and risks (both direct and indirect) of infrastructures. The
literature identified is fairly strong quantitatively in terms of environment-pollution
related health impacts and quite weak with respect to direct and indirect health benefits of
infrastructure. The exceptions are the case of water supply and sanitation and in some
cases, energy. Measurable improvements in health based on infrastructure interventions
provide an important evidence base for decision-making, but in many cases the factors
driving health outcomes can be quite complex, and are confounded by factors like socio-
economic status and accessibility (and transport costs) to hospitals and health services.
Table II. 1 Review of Health Benefits and Risks of Infrastructures.
Infrastructure Literature Review of Health Benefits
Sector___________(Direct and Indirect)________________
Transport
Access to work, school, & essential health
services (Killoran et al., 2011; Babinard &
Roberts, 2006); wealth, job creation,
economic development (Mohan, 2004)
Energy
Enables improved standards of living
(Pasternak, 2000), extended hours &
expanded services for hospitals (Hess,
2011; Schwartz et al., 2011); enables
cooking, boiling water, space heating,
cooling; electricity; transport enabling
access to livelihoods; social networks;
industrial production; & communication
(Wilkinson, 2007; Saatkamp et al., 2000;
__________McMichael, 1994)_________
Reduces waterborne illnesses and prevent
spread of animal-borne disease pathogens
(Butala, 2010); reduces infant mortality
(UNW-DPAC, 2011) & mosquito-related
illnesses (Gunther, 2011)
Hazardous Protects water, air, & soil by promoting
Waste proper storage & disposal of toxic waste
Management__________________(Guerriero, 2009)__________
Water Supply
& Sanitation
Literature Review of Health Risks
(Direct and Indirect)______________________
Accidents (WHO, 2004); transport-related
urban outdoor air pollution affecting
asthma exacerbation, acute & chronic
bronchitis, respiratory / cardiovascular
illness, & lung cancer (Samet, 2000); lack
of mobility as causal factor in maternal &
neonatal mortality (Molesworth, 2006)
Outdoor and indoor air pollution (Listorti,
2004; IEA, 2010), injury risks, and
industrial hazards (Venkataraman et al.,
2010); cardiovascular disease; respiratory
disease; bronchitis; asthma; and eye
infections (Kammen, 2011)
Diarrhea (Montgomery, 2007),
schitosomiasis, intestinal helminths;
trachoma; trypanosomiasis (Eisenburg et
al., 2001); malnutrition (Gleick, 2002);
cholera; typhoid (WQHC, 1995); lung,
bladder & skin cancer (Smith, 2000)
Toxic chemicals and waste affect airway
diseases and brain, lung, & gastrointestinal
__________cancer (Rushton, 2003)_____________
16


The second approach involved exploring health outcomes, and how they relate to
possible infrastructure and environmental causal relationships. The health outcomes that
are reported in the global health literature that may relate to infrastructure include:
Airborne Pollution and Health Risks: In 2000, urban outdoor air pollution
caused 799,000 deaths globally, an estimate calculated by considering effects of
particulate matter (PM10 at 10 pm or less and PM2.5 at 2.5 pm or less) on health
for all cities with populations over 100,000 (Ezzati, 2006). More recently (2010),
World Health Organization (WHO) estimates that every year, urban outdoor air
pollution causes 1,3 million deaths worldwide, while indoor air pollution from
improperly ventilated cook stoves burning solid fuels is responsible for 1,6
million deaths worldwide (WHO, 2011).
Waterborne Pollution and Health Risks: In 1995, water-borne diseases caused
more than five million deaths worldwide. Of these, about four million deaths were
of children below age five (Gadgil, 2003). More recently (2008), estimates
suggest annual mortality from waterborne diseases is closer to 3,6 million deaths
per year (WHO, 2008). While 2.5 billion cases of diarrhea occur each year among
children under-5, and estimates suggest overall incidence has remained relatively
stable over the past two decades, diarrhea mortality has declined over past two
decades from an estimated 5M deaths among under-5 children to 1.5M deaths in
2004 (Boschi Pinto, 2009).
Transport Accidents: In 2002, there was an estimated 1.2 million deaths from
road traffic accidents: an average of 3242 deaths per day (WHO, 2004). More
recently (2010), road deaths accounted for 1.3 million deaths worldwide while
also causing between 20 million and 50 million non-fatal injuries every year
(WHO, 2011). In India, where rapid motorization is occurring at a rate of 10% per
year, transport accidents has become a growing public health concern, with the
country having experienced an average increase of about 4% per year in total
number of traffic fatalities in the period 1997-2003, and the rate having increased
17


to 8% per year since then. In 2007, 114,000 persons were killed in traffic
accidents (Mohan, 2004; WHO, 2004).
Cancer: In 2000, cancer was estimated to account for about 7 million deaths
(12% of all deaths), with more than 70% of all cancer deaths occurring in low-
and middle-income countries. More recently (2007), cancer accounted for 7.9M
deaths worldwide. Projections suggest an increase to an estimated 12 million
deaths in 2030 (WHO, 2011).
Diabetes: In 2000, an estimated 959,000 deaths worldwide were caused by
diabetes (WHO, 2004). More recently (2011), estimates suggest 346 million
people worldwide have diabetes currently, and projects that diabetes deaths now
estimated at 3,8 million deaths worldwide per year will double between 2005
and 2030 (WHO, 2011). In India, 40 million suffer from diabetes, and this figure
is estimated to go up to 80 million by 2025. In Delhi, estimates suggest three
million suffer from this disease (DFI, 2011).
Traffic accidents, air pollution and waterborne disease are more readily linked with
infrastructures or infrastructure-related environmental factors. Indeed, protection of
public health is the basis on which WHO and Indias Central Pollution Control Board
have established the following:
Air quality standards limiting ambient concentrations in air of particulate
matter (PM), ozone (03), nitrogen dioxide (N02), and sulfur dioxide (S02).
These concentrations are often exceeded in many Asian cities, and levels
beyond these standards help estimate the excess mortality from pollution
episodes, which typically occur from vehicle emissions.
Drinking water and river quality standards addressing pathogen content and
limiting concentration of certain carcinogens in drinking water (e.g., arsenic,
benzene, nickel, cadmium, etc). Levels above these prescribed standards also
often occur in Asian cities, as shown in Table 3.
In the above cases, adverse health impacts can be expected when infrastructure
condition (i.e., traffic, untreated water, etc.) are such that these environmental water or air
18


quality standards are exceeded. In contrast, other diseases such as diabetes and
cardiovascular disease (CVD) risk are more difficult to link to infrastructure, however,
careful qualitative-quantitative studies are uncovering such linkages, as summarized in
Table 2. As an example: "Development of type 2 diabetes mellitus is influenced by built
environment, which is, the environments that are modified by humans, including homes,
schools, workplaces, highways, urban sprawls, accessibility to amenities, leisure, and
pollution. Built environment contributes to diabetes through lack of access to physical
activity and increased prevalence of obesity in less walkable neighborhoods (Saelens et
al., 2003). With globalization, there is a possibility that lifestyles from western countries
like the US may be replicated in developing countries such as India, where the
underlying genetic predisposition makes them particularly susceptible to diabetes
(Pasala, 2010). Thus, diabetes -which is a major health risk category is linked to the
built environment infrastructure in this example. Similarly, each of the health risk
categories in Table 2 below is described broadly in the left column, while the
infrastructure relationship is summarized in the middle column.
Table II.2 Tracing Health Outcomes to Infrastructure and Environment-Related Factors.
Health Risk Categories Infrastructure and Environment-Related Factors References
Heart Diseases & Heart Attacks Fossil-energy from transportation, energy generation, industrial production and buildings infrastructure causes air pollution (among 14 identified triggers for heart attacks, along with alcohol use, anger, physical exertion, and others). Studies on obesity suggest built environments and motorized transport as major contributors to sedentary lifestyles and increased heart disease risks. Infrastructure linkages', energy, transportation, buildings, and industrial production. Nawrot, 2011; Mobley, 2006; Gordon- Larsen, 2006; CDC, 2011
Airborne Pollution and Health Risks Fossil-energy related indoor and outdoor airborne pollutants (e.g., sulfur dioxide, particulates, carbon monoxide), minimal distance to nearby highways, overcrowded homes, sleeping conditions, and lack of proper sanitation are contributors to serious health issues including tuberculosis, bronchitis, asthma, pneumonia, diphtheria, influenza, whooping cough, and leprosy (with inhaled respiratory droplets being an airborne transmission pathway). Infrastructure linkages: energy, transportation, buildings, sanitation, and industrial production. Shaw, 2004; Ezzati, 2006; Fang et al., 2008; WHO, 2011; Agarwal, 2005; Gupta, 2007
19


Table II.2 (contd.)
Unsafe drinking water, insufficient water for hygiene,
Waterborne sanitation conditions, and water-related insect vectors are
Pollution major contributors to diarrhea, dysentery, jaundice, malaria,
and Health typhoid, acute poliomyelitis, cholera, food poisoning, and liver
Risks disease. Infrastructure linkages', water supply and sanitation
(e.g., wastewater and sewage treatment).
Gadgil, 2003; Gleick,
2002; Boschi Pinto,
2009; WHO, 2004 /
2011
Hazardous waste exposure and environmental contaminants (e.g., methane, benzene, cadmium) are contributors to cancer (influencing the likelihood of developing brain, lung, and Rushton, 2010;
Cancer gastrointestinal cancer) and central nervous system damage. Arsenic contaminated water also contributes to lung, bladder, and skin cancer. Infrastructure linkages', water treatment and hazardous waste management. Griffith, 1989; Guerriero, 2009
Together, Tables 1 and 2 suggest infrastructure and infrastructure-related
environmental factors can have tremendous local and global impact. However, detailed
epidemiology study is still needed to quantitatively understand infrastructure and
infrastructure-related environmental factors at the city-scale, particularly for diseases
such as diabetes with complex infrastructure dependencies. Further, epidemiological
studies comparing households with and without urban infrastructures can address
confounding factors (e.g., socio-economic status and accessibility to hospitals and health
services) with regard to resulting health outcomes. Before embarking on such detailed
study, we describe preliminary methods for exploring relationships between urban health
and infrastructure using a mixed quantitative and qualitative study addressing the case of
Delhi. We also present Table 3 as initial context for comparing existing environmental
guidelines (where infrastructure and environment-related health associations are well-
understood) to current Delhi pollutant levels.
20


Table II.3 Short-Term (24-hour mean) Environmental Exposure Guidelines and Delhi
Compliance.
Environmental Compliance Standards Pollutant Concentration Guidelines
Outdoor Air WHO 2005 Guidelines: PMio of 50 pg/m3 National Ambient Air Oualitv Standards fNAAOSi of India:
Quality -> Sensitive Areas:
PMio of 75 (Tg/m3; -> Industrial Areas: PM10of 150 pg/m3; -> Residential Areas: PMio of 100 (rg/m3
Delhi Average
Pollutant
Concentrations were...
Health-Related Rationale
for Standard
Up to four times higher
than the NAAQS
residential standard.
State of Environment for
Delhi 2010 Report:
2008 PMi0: 400 |xg/m3
Quote: PM
concentrations have
remained (1997-2010)
very high compared to
the national ambient air
quality standard.
Short-term exposure to a
PM 10 concentration of 150
pg/m3 would be expected to
translate into roughly a 5%
increase in daily mortality, an
impact that would be of
significant concern, and one
for which immediate
mitigation actions would be
recommended (WHO, 2005).
Major concerns for human
health from exposure to PM-
10 include: effects on
breathing & respiratory
systems, damage to lung
tissue, cancer, and premature
death (US EPA, 2010)
WHO 2004
Guidelines:
Benzene: 0.01
mg/L
Arsenic: 0.01
mg/L
Drinking Water
Quality
India Central
Pollution Control
Board (CPCB)
Standards:
Biological Oxygen
Demand (BOD): 3
mg/1 or less (Note:
this is CPCBs C
Class Criteria set
for the use of
drinking water
sources after
conventional
treatment and
disinfection.)
Up to eighteen times
higher than CPCB
standards.
Ministry of
Environment. 2008:
Range of BOD levels: 3
to 55 mg/1
Quote: Municipal
Corporation of Delhi
(MCD) found 15% of
Delhis water (90 out of
765 samples) to be unfit
for drinking. In South
Delhi, contamination is
highest with 50% of
samples declared
polluted. (Chandel,
2009).
Standards are set with the
ultimate goal of no adverse
health effects as a result of
human uses of water. For
example, under the US Safe
Drinking Water Act, the US
Environmental Protection
Agency set standards for 90
contaminants in drinking
water and set legally
enforceable standard limits
e.g. maximum contaminant
levels (MCL). Water meeting
these standards is considered
safe to drink. (Ramaswami et
al., 2005).
21


Methods
A case study of Delhi was conducted to explore the extent to which current urban
health outcomes are shaped by infrastructure and infrastructure-related environmental
factors:
First, baseline annual mortality data from multiple data sets in Delhi are collected
and analyzed to assess data quality and to understand how the various sources of
data are different.
Next, associations are identified between infrastructure and health through a
combination of literature review and local expert opinion on the number of
observed deaths in Delhi, India as it relates to infrastructure and environmental
factors. Local experts were selected as either internationally recognized public
health and public health engineering experts based in and around Delhi; senior
professionals having in-depth knowledge and understanding of public health and
infrastructure development in Delhi; and senior professionals having in-depth
experience researching and analyzing local health and infrastructure outcomes and
associations in Delhi.
Third, estimates of excess mortality were computed for the case of air quality
where infrastructure/environment associations are well-understood to demonstrate
the scale of excess deaths related to just a single environmental/infrastructure
factor, similar to other recent studies estimating excess mortality (NRC, 2008;
Jacobson, 2007).
22


In future studies, similar computations will be needed for all infrastructure and
environmental factors to provide cities with a template for becoming both healthy and
low-carbon.
1. Delhi Mortality Data: For initial exploration of health outcomes, mortality is
selected over other indicators (e.g., morbidity) due to its policy relevance, significant /
intrinsic importance as a measure of societal well-being (Sen, 1998), and the availability
of mortality data at city-scale.
As suggested by Amartya Sen: The significance of mortality information lies in a
combination of considerations, including the intrinsic importance we attach and have
reason to attach to living; the fact that many other capabilities that we value are
contingent on our being alive; and the further fact that data on age-specific mortality can,
to some extent, serve as a proxy for associated failures and achievements to which we
may attach importance (Sen, 1998). Other metrics used by health professionals include
morbidity, hospitalization, disability-adjusted life years (DALY), missed days at work,
and height-to-weight ratios to reflect for malnutrition. Ideally, multiple health indicators
should be used including morbidity with DALYs (Murray, 1994) both of which are
sensitive to acute environmental conditions (e.g., responses to high pollution episodes).
However, public heath experts, recognizing the complexity in relating health outcomes to
infrastructure, suggested that this initial study focus on mortality data first.
Datasets: Initially, three datasets were reviewed to extract mortality data, including
the:
(1) 2010 National Health Profile by Ministry of Healths Central Bureau of Health
Intelligence (CBHI) using State / Urban Territory Data for Delhi,
23


(2) 2008 Births and Deaths (B&D) Report for the North Capital Territory (NCT) of
Delhi by the Office of the Chief Registrar, and
(3) 2008 Certified Causes of Death Report for Delhi by the Office of the Chief
Registrar.
The 2008 Births and Deaths Report for the North Capital Territory (NCT) of Delhi
was selected for further analyses as the most comprehensive, consistent, and clearly-
defined data providing 40 specific medically certified causes of death and providing more
comprehensive accounting of cause-specific deaths then other databases. For example,
data inconsistencies existed in 2009 reports at time of this research and the CBHI
National Health Profile presents data for only 29 specific medically certified causes of
death and undercounts the total number of deaths for Delhi.
The 40 specific causes of deaths as reported by the 2008 B&D Report in Delhi are
shown in the Appendix and are aggregated into twelve major health risk categories that
are shown in left column of the Appendix table and in the Figure 2 pie charts. Of these
twelve categories, six were found to be infrastructure and environment related based on
the literature review shown in Table 2. These six categories are shown hatched in Figure
2 and were used to gather local expert opinion on the relationship to infrastructure.
2. Infrastructure-Health Associations: Local expert opinions were gathered to get
their best estimate at two phenomena:
What percent of deaths could be attributed to infrastructure and infrastructure-
related environmental factors?
24


How many of those deaths are estimated to be attributed to infrastructure and
infrastructure-related environmental factors, could be avoided by access to health
services?
Overall, six experts including two government health officials, a clinical physician, a
public health professor, a public health engineer, and an internationally-recognized
development consultant participated in a ten-question survey and provided a total of
about 60 quantitative replies for Delhi which guided preliminary estimates shared in this
paper. As such, this survey approach provides a first attempt at estimating the proportion
of mortality that is likely infrastructure and environment-related in Delhi using local
expert knowledge. We emphasize that these results are preliminary and exploratory;
untangling infrastructure and health relationships are complicated and additional
epidemiological field data is much needed. The expert estimates summarized here, and
accompanied by sample quantitative analysis, provides a starting point for detailed
epidemiological study.
3. Sample Estimate of Excess Mortality: The third computation shows how
epidemiological data, where available, can in fact help make quantitative associations
between health outcomes and varying levels of infrastructure and environmental
conditions. For particulate matter concentrations in environment, as one example, the
following equation (EPA, 2006) is used:
Mortality Reduction=Pollution Change x Effect Estimate x Incidence Rate x
Population
Equation III. 1 Estimating Excess Mortality
25


The data for this computation is based upon a single environmental factor, suspended
particulate matter (SPM) smaller than about 10 microns in diameter (.PM10), which is
related to emissions from fossil combustion, e.g., in transportation and industrial
production which also accounts for over 60% of C02 emissions in India (TEA, 2011)
and almost 80% in Delhi (TERI, 2010; Chavez et al., 2011). Epidemiological studies of
mortality observed in Delhi following high PM episodes were gathered by (Cropper et al.
1997) and modeled using the Poisson model. The epidemiological study showed 4.3%
and 3.1% excess CVD and respiratory mortality, respectively, associated with an increase
of 100 pg/m3 in SPM (Cropper et al., 1997), which is the effects estimate. Incidence rate
refers to CVD and respiratory mortality observed in the base case (i.e., the current
condition). The equation above is applied in this paper to compute avoided mortality if
Delhi were to transform their transport and industrial emissions to meet Indias NAAQS
standard of 100 pg/m3 in residential areas with reductions from the current annual
average PMi0 levels of 400 pg/m3 (i.e. a reduction of 300 pg/m3 to be in compliance with
NAAQS set by the CPCB).
Results
1. Exploring Baseline Mortality, Mortality Rates, Life Expectancy, and Causes of Death
In the National Capital Territory of Delhi, the mid-year population of Delhi in 2008 was
estimated at 17,115,000 (Government of NCT Delhi, 2008a), and the number of total
deaths registered under the Civil Registration System included 107,600 deaths, of which
57,122 were institutional deaths (53%) and 50,742 were domiciliary (47%). 68,033 deaths
were of males and 39,567 were of females. As to mortality distribution by age group, 14%
of deaths were of under-5 children, 72% of deaths between the ages of 5 and 69, and 15%
26


of those greater than 70 (see Figure 1). Annual mortality rates and average life expectancy
in India, Delhi, and Mumbai are also presented below:
Table II.4 Comparative Health Indicators for Delhi, Mumbai, and India.
Health Indicators Delhi Mumbai India
Est. Annual Mortality Rate (Per 100,000 population) 629 689 989
Average Life Expectancy 72 71 69
A comparison of these two health indicators in Table 4 suggest mortality rates and
life expectancy may be related. However, life expectancy -as an indicator- is less
sensitive to changing environmental conditions on a periodic basis, e.g., acute pollution
or heat episodes that may occur a few days a year. For such purposes, mortality is a better
metric.
1-4
4%.
5- 14
3%
15-24
8%
55 64
17%
Figure HI Mortality by Age: Delhi (2008)
27


2. Delhi Mortality Outcomes Associated With Infrastructure-Related Factors
Among the total registered deaths in Delhi, over 55% were not classified by cause of
death. The remaining 45% (or 48,148 deaths) were grouped into twelve health risk
categories of which six that had a literature-based relationship to infrastructure are shown
hatched in Figure 2. Among the classified deaths, up to 62% of classified deaths could be
infrastructure-related (as shown in pie chart B). However, this is likely an overestimate of
the influence of infrastructure; to address this, we conducted local expert interviews.
A
Airborn* Oimt*
heart Dt*e**e & Ptthogen*
Mean Attach* %
15%
B
Genetic
Blood-Related
Disease of Diseases
Digestive IS
System ) Elderly
2 85* I 14*
Other Ny
Infectious
Diseases
4*
Pregnancy &
Birth
9*
1*
Waterborne
Disease
Pathogens
OS
The pie chart A on left shows percentages of total deaths in 2008for the NCT of Delhi (including
unclassified deaths); Pie chart B (right) shows percentages of total classified deaths. The hatched portions
represent health outcomes with potential infrastructure-related linkages.
Figure fl.2 Mortality by Cause (if classified) in the NCT of Delhi (2008)
As shown in the figure, the 40 causes of death (presented with reported data in
Appendix) are grouped into twelve health risk categories six of which are infrastructure
and environment-related: 1) heart diseases and heart attacks; 2) airborne disease
pathogens (e.g., bronchitis and asthma); 3) waterborne disease pathogens (e.g. cholera,
typhoid, diarrhea); 4) diabetes; 5) cancer; and 6) accidents. While epidemiology studies
are being designed, local expert interviews were conducted using questions shown in
Appendix B to estimate percentage of deaths for these six categories as (to their best
guess) related to engineered infrastructure and the environment:
28


Heart Diseases & Heart Attacks (14.7% of total deaths): the median of expert
responses suggests 25% could be infrastructure related of which 25% could be
avoided or delayed with timely access to health services.
Airborne Disease Pathogens (5.7% of total deaths): the median of expert
responses suggests 25% could be infrastructure related of which 25% could be
avoided or delayed with timely access to health services. This includes deaths
from tuberculosis, bronchitis and asthma, pneumonia, diphtheria, influenza,
whooping cough, and leprosy.
Diabetes (4.3% of total deaths): the median of expert responses suggests 35%
could be infrastructure related of which 15% could be avoided or delayed with
timely access to health services.
Cancer (2.85% of total deaths): the median of expert responses suggests 25%
could be infrastructure related of which 15% could be avoided or delayed with
timely access to health services.
Transport Accidents (1.4% of total deaths): The expert responses suggest 100%
could be infrastructure related of which 15% could be avoided or delayed with
timely access to health services. This includes only transport accident deaths,
while the above pie charts showing 3% (Pie Chart A) and 6% (Pie Chart B)
include additional types of accidents: burns, falls and drowning, accidental
poisoning, bites, and others.
Waterborne Disease Pathogens (0.7% of total deaths): The expert responses
suggest 50% to 100% could be infrastructure related of which 25% could be
avoided or delayed with timely access to health services. This includes deaths
from dystentery and diarrhea, jaundice, malaria, typhoid, acute poliomyelities,
cholera, and food poisoning.
To supplement the expert opinions, and to gain a quantitative understanding of the
order of magnitude of deaths that could be attributed to environmental pollution when
epidemiological data are available, a sample quantitative analysis is provided for the case
of outdoor air quality.
29


3. Estimating CVD & Respiratory Mortality Reduction Due to Standards Compliance
A sample calculation for Delhi mortality reductions expected from compliance
with Indias Central Pollution Control Board (CPCB) National Ambient Air Quality
Standards (NAAQS) is shown in the calculation below based on the key data provided in
Table 3. The equation utilized is also used in the model, BenMAPP, which estimates
health impacts when city populations such as Mumbai, India experience changes in air
quality (EPA, 2006). By running these sorts of equations based on environmental
exposure models and epidemiological studies, improved understanding of infrastructure-
related health effects can be achieved.
Estimating Air QualitV'R<>l**ttd Cardiovascular H* Respiratory
Mortal ity Roduction in Delhi
/Mortality f?er/u t?ti on Po Notion C/tattcfO Effoct
fs firm* trr Indt/*?nce9 fr; Popt///t ion
* F*oUtition a nniiitl change in PM 10 in
microgroms por motor cu bod ( pg/m3).
* Effoct E stimato percent change in mortality from
a cortain offoct por ug/m3 of RIVIIO.
* tndcionco Ffato baseline of deaths/person/year
from that effect.
* /atioti number of persons in a specific age
tl roup.
Sample Calculation for Cardiovascular fCVD) Deaths:
fVlortal itv Reduction 300 pg/m3 (0.043/100 pg/
m3) (12,949 deaths/17t115rOOO pers) - 17,115,000 1670
rrxcoss d?aths/year. Note: This represents a 13% reduction
in OVD deaths.
Sample Calculation for Respiratory Deaths:
Mortal ity Reduction 300 pg/m <0.031/100 pg/
m3) x (2696 deaths/17,1 15,000 pers) 17,115,000
persons 2b1 excess deaths/vear. Note: This represents
a 9% reduction in respiratory deaths.
Assumptions: This calculation assumes Delhi can roduco
RIVI toy 300 ug/m3 to moot ttoo IMAAQS residontial standard.
Cropper e t al. (1997) study on premature mortality and
standard particulate matter in Do I hi i, India reported a CVD
relative risk of 1.043 and a respiratory relative risk of 1.031
associated witto a change in SPM level of 100 ug/rn3 (used to
contpute).
Estimated Mortality Reduction: 1,921 CVD respiratory
deal hs/yeti r < 1 2% reduction of a I CVD & respir atory deal hs).
Figure II.3 Estimating Air Quality-Related Cardiovascular & Respiratory Mortality
Reduction in Delhi
30


This sample calculation shows that quantitative epidemiological models indicate 12%
of CVD and respiratory deaths (NCT Govt, of Delhi, 2008b) are pollution related in
Delhi. This is due to changes in PM only. Change in levels of other air pollutants (e.g.,
S02, NOx, unbumed volatile organic compounds) can also have significant health
impacts (Cropper, 1997; Russell, 2008).
Discussion: Implications for Infrastructure Interventions
Preliminary exploration using 2008 Delhi mortality data and local expert opinion on
the extent to which infrastructures and infrastructure-related environmental factors shape
the number of observed deaths in Delhi is summarized in Table 5 below.
In aggregate as shown in Table 5, almost 19% of recorded classified deaths may be
associated with infrastructure and environment related causes. It is important to recognize
that 55% of the total deaths are not classified, and many of the deaths in developing
world cities may not happen in institutions where causes of death records are maintained.
If all these unclassified deaths followed similar patterns, infrastructure and infrastructure-
related environmental factors shaping 19% of health outcomes could be a large factor in
motivating low-carbon development in Asian cities. While this still needs further
exploration with rigorous epidemiology data, the preliminary results in this study suggest
infrastructures can have a significant impact on health outcomes for developing country
cities (estimated at 19%).
In contrast, future heat-related deaths have been estimated for Delhi to increase
overall deaths by 2% and future CVD-related deaths by 4% per degree Celsius increase in
temperature greater than 20 degrees Celsius (95% confidence interval) (Hajat, 2005).
31


These results suggest addressing public health outcomes at the present time may be a
greater motivator for low-carbon development then future climate-related health risks.
Table II. 5 Summary of Local Expert Opinion on 2008 Deaths.
Infrastructure & Environment-Related Health Risk Category (% of classified deaths) Median % of Deaths Estimated % of All
Attributed to Classified Deaths in
Infrastructure & Environmental Factors by Local Expert Opinion Delhi That May Be Related to Infrastructure
Heart Diseases & Heart 25% 8.2%
Attacks (32.9%)
Airborne Disease Pathogens (12.7%) 25% 3.2%
Transport Accidents (3.2%) 100% 3.2%
Diabetes 35% 2.7%
(7.8%)
Waterborne Disease Pathogens (1.5%) 50% to 100% 0.35% to 0.7%
Cancer (6.4%) 25% 0.7%
As shown in Table 5, the highest percentage of mortality related to infrastructure
is heart diseases and heart attacks. These preliminary results demonstrate that
infrastructure interventions and air quality compliance strategies that decrease fossil
energy use and GHG emissions can also have significant affects on reducing risks of
cardiovascular and respiratory deaths, thereby leveraging both low-carbon and healthy
cities.
As shown in Table 2, heart diseases, airborne pathogen risks, and diabetes health
outcomes are associated with air pollution and obesity, each of which relate to those
infrastructures implicated such as transport energy. Through transportation sector
32


improvements alone, four of the aggregated health risk categories can be addressed via
cleaning up of air pollution (e.g., cleaner fuels such as low sulphur diesel, reducing
vehicle miles travelled (VMT), making vehicles more efficient), promoting mode shifts
that make communities more walkable, physically active, and perhaps more transit
dependent thereby reducing road conflicts. Such infrastructure interventions reducing
VMT, cleaning up fuel, making vehicles more efficient, and mode shifts to transit modes
all can help achieve both health and low-carbon city goals.
Industrial symbiosis and other clean energy interventions can also clean
production, thereby reducing airborne pollution (e.g. S02, NOx, and PM10). With
airborne pathogen risks, tradeoffs exist between house size efficiency (for energy
consumption and low-carbon goals) and mortality (as it relates to overcrowded sleeping
conditions affecting tuberculosis and other health outcomes). Passive strategies for
building and home design that increase ventilation and provide natural cooling reduce
energy loads and indoor air pollution.
Measuring both synergistic and antagonistic pathways toward low-carbon and
healthy cities is of importance. Synergistic pathways achieve both GHG emission
reduction and health benefits. Antagonistic pathways may increase health benefits, while
worsening GHG emissions or vice versa. For example, providing clean drinking water
systems and reducing exposures to cancer-causing pollutants through improved
hazardous waste management systems may place new energy demands on cities by
requiring new water treatment plant facilities and new transport requirements for
hazardous waste that can result in increases in GHG emissions, that reduce waterborne
diseases and cancer-related deaths. Synergistic pathways meanwhile do exist as
33


demonstrated by Miller et al. who has done a life cycle assessment of energy use in
wastewater treatment in Hyderabad, India, showing that energy investments in
wastewater treatment plant infrastructure may actually reduce overall GHG emissions, by
reducing emissions of methane and nitrous oxide from untreated sewage (Miller et al.,
2011). Wastewater treatment therefore may be net GHG mitigating.
Conclusion
To develop low-carbon cities of the future, carbon mitigation potential must be combined with
quantitative analysis of other benefits and co-benefits, including energy security and public
health, suited to the overall goals of society. A. Ramaswami & S. Dhakal, 2011
With some 1050 cities in the US (US Mayors, 2011), eight cities in China, and 40
cities in India setting low-carbon development goals (ICLEI-SA, 2009), city-scale
strategies focused on infrastructure and low-carbon interventions tailored to fit the unique
local cultural, social, economic, public health and human development aspirations in each
city, can (presumably) be more successful then failing international efforts focusing on
carbon mitigation as primary objectives. Results from this preliminary exploration
indicate up to 19 percent of all classified deaths in Delhi may be infrastructure-related,
and as shown here, reducing mortality through infrastructure interventions provides an
important driver for low-carbon city development. We emphasize that additional
exploration of infrastructure-health associations and causal relationships through field
work and epidemiological study is needed. The results of this study provide an initial
qualitative quantitative exploration that opens up a new line of inquiry.
The initial explorations in this paper can be complemented by future
epidemiological assessment of the different degrees of health hazards associated with the
infrastructure in Delhi from 2008 to 2011. Currently, initial estimates suggest 55% of
34


households live within 500 meters of roads with high levels of air pollution (putting
residents at risk of cardiac and respiratory problems), 16% of households lack access to
drinking water taps (putting residents at risk of waterborne illnesses), 6% lack access to
latrines, and 8% are using solid fuels (wood, dung, and charcoal) for cooking (Govt, of
Delhi, 2009; Govt, of India Ministry of Urban Development SLB Databook, 2009).
Additional research and a carefully constructed epidemiological study design could be
useful in establishing objective associations on mortality statistics or other health
indicators and the provision and upgrading of specific infrastructures, including water,
sewage, electricity, transport, and buildings (e.g. upgrades to pucca and kutcha housing)
within the citys geographic areas. Health effect estimates, risk factors, and potential
confounding factors for relevant infrastructure-related health outcomes need to be
identified in such a study. For example, is there a measurable improvement in households
with piped water and how can confounding factors like age, socio-economic status,
access to hospitals and health services be accounted for?
Acknowledgments
This study was supported by an Integrative Graduate Education and Research
Traineeship Grant (IGERT; NSF Grant #DGE-0654378) from the United States National
Science Foundation (NSF) to the Center for Sustainable Infrastructure Systems at
University of Colorado Denver
35


Appendix. Chief registrar for NCT of Delhi Re-analysis of births and deaths 2008 report for Delhi.
Chief Registrar for VCT of Delhi Births and Deaths 2008 Report for Delhi
No. Aggregated Health Risk Categories No. of Deaths % of All Deaths Medically Certified Cause of Death (% of All Deaths By Specific Cause) Recorded Deaths By Cause [Infrastructure Causal Relationship]
1 Heart Diseases & Heart Attacks 15876 14.75% Heart Diseases & Heart Attacks (14.77%) 15876 [transport, energy, buildings, industry]
2 Airborne Disease Pathogens 6114 5.68% Tuberculosis (2.45%) 2632 [air pollution and overcrowded conditions]
Pneumonia (1.43%) 1539 \E, T&AQ, O]
Bronchitis & Asthma (1.22%) 1316 [E, T&AQ]
Diptheria (0.23%) 345 [O & poor sanitation]
Influenza (0.15%) 161 [T minimal distance to nearest national highway]
Whooping Cough (0.06%) 60 [O]
Leprosy (0.06%) 61 [O]
3 Diabetes 4626 4.30% Diabetes (4.30%) 4626 [T&LU]
4 Cancer 3070 2.85% Cancer (2.85%) 3070 [HW, T&AQ, E]
5 Accidents 2765 2.57% Transport Accidents (1.43%) 1540 [T&AQ]
Accidental Bums (0.62%) 670
Falls & Drowning (0.17%) 187
Accidental Poisoning (0.11%) 122
Bites (0.05%) 49
Other Accidents (0.18%) 197
6 Waterborne Disease Pathogens 714 0.66% Dysentery & Diarrhea (0.19%) 206 [WS]
Jaundice (0.13%) 140 [WS]
Malaria (0.09%) 96 [WS]
Typhoid (0.08%) 90 [WS]
Acute Poliomyelities (0.08%) 87 [WS]
Cholera (0.07%) 79 [WS]
36


Food Poisoning (0.01%) 16 [poor sanitation]
7 Homicides & Suicides 333 0.31% Homicide (0.14%) 147
Suicide (0.17%) 186
8 Pregnancy & Birth 4231 3.93% Birth-Related Affecting Child (3.07%) 3303 [H]
Cerebral (Paralysis) (0.68%) 729
Pregnancy Complications (0.15%) 165 [H]
Abortion (0.03%) 34
9 Other Infectious Diseases 1672 1.55% Tetanus (0.80%) 864
Meningitis (0.61%) 652
Rabies (0.08%) 84
Syphillis (0.07%) 72
Measles (0.04%) 38
10 Disease of Digestive System 1374 1.28% Chronic Liver Disease & Cirrhosis (1.10%) 1187 [poor sanitation]
Appendicitis (0.40%) 427
Ulcer of Stomach & Duodenum (0.06%) 63
11 Genetic Blood-Related Diseases 625 0.58% Anemia (0.58%) 625
12 Elderly 6545 6.08% Senility (6.09%) 6545
A Deaths With Cause Not Classified 59314 55.12% % of Deaths Infrastructure-Related 27.93%
B Deaths With Cause Classified 48286 44.86% LEGEND for Infrastructure-Related Causes of Death (Informed bv Lit. Review): O = Overcrowded housing and sleeping conditions ; WS = Water Supply & Sanitation (e.g., inadequate drinking water or sewerage systems); E = Energy (e.g., indoor smoke from solid fuels); HW = Hazardous waste (e.g., inadequate storage and disposal of toxic waste); T&AP = Transport & associated air pollution (e.g., outdoor pollution, road accidents); H = Health Infrastructure (e.g., inadequate accessibility to hospital and/or health services)
C Total Recorded Deaths 107600 100.00%
Notes: Associations of infrastructure-related health factors to causes of death are shown in [ ] (also see Legend) and are informed by the literature below (full citations available in the references):
Shelter / Crowded Sleeping Conditions: tuberculosis (Agarwal, 2005); measles (Agarwal, 2005)
Water Supply and Sanitation: cholera (Butala, 2010); typhoid (WHO, 2011); dysentery and diarrhea (Esrey, 1991; Gleick, 2002); liver disease (WHO, 2011)
Energy: tuberculosis (Agarwal, 2005), pneumonia (WHO, 2011); cancer (Slaper et al., 1996; WHO Fuel for Life, 2006);
Hazardous Waste: cancer (Guerriero, 2009); brain, lung, and gastrointestinal cancer (Rushton, 2009; Griffith, 1989
Transportation: accidents (Mohan, 2004); heart disease (Pande, 2002; Nawrot, 2011); pneumonia / bronchitis/ asthma (Krzyzanowski, 2005); influenza (Fang et al., 2008)
Health Infrastructure: complications related to pregnancy (Molesworth, 2006); Condition such as birth injuries.. .premature originating in perinatal period (Aguilera, 2006)
37


CHAPTER III.
A REVIEW OF EPIDEMIOLOGICAL STUDIES RELATING TO
INFRASTRUCTURE AND POLLUTION IN ASIAN CITIES:
EVIDENCE AND GAPS FOR INFRASTRUCTURE-RELATED
HEALTH EFFECT ESTIMATES IN DELHI
Abstract
Many Asian cities currently lack analytic tools and an evidence-base to link
infrastructure improvements with health outcomes. This paper builds a preliminary
foundation to address such challenges by conducting a review and synthesis of literature,
models, and field data. Over eighty quantitative epidemiological studies, using literature
focused on Asian cities where possible, are identified and reviewed across seven
categories of infrastructure-related health outcomes: cardiovascular disease, respiratory,
airborne infectious diseases, transport accidents, waterborne pollution and health risks,
cancer, and diabetes. An initial database of health effect estimates is created to model
mortality risk mitigation options. Knowledge gaps are also identified where further
bottom up study may be needed for example, how access to urgent healthcare shapes
health outcomes. By reviewing epidemiology studies across Asian cities and specifically
for Delhi, India, this top-down study helps to inform areas where bottom-up field study
are still needed and also to help inform a first order scenario tool that can be used to
compute health risk reduction benefits in Asian cities and for Delhi, India in particular.
35


Introduction
Asia now represents half of the worlds urban population. Cities in India and
China alone make up 28% of world urban population and by 2030, Indian and Chinese
cities will represent 40% of global urban population. Such urbanization creates huge
demands for infrastructures, basic services, and improved environments. In response to
such challenges, the World Health Organization (WHO) in 1994 created the Healthy
Cities initiative to improve health and quality of life in global cities. Since then, this
initiative has taken the form of a long-term program aiming to place health high on
agendas of decision makers and to promote comprehensive local strategies for health
protection and sustainable development. Such goals become increasingly critical with
half the worlds population now living in cities and 1/3 of the worlds urban population
living in what are defined by the United Nations as slum conditions (UN-Habitat, 2006).
In this study, we analyze current health risks facing Asian cities posed by
infrastructures (or lack thereof) and infrastructure-related environmental factors. In Delhi,
India for example, the current urban population of 18 million is expected to soon reach
over 26 million by 2025, creating demands for not only new civil infrastructures (i.e.
water, sanitation, energy, transport and building infrastructures), but also infrastructure
upgrades that help improve environmental conditions. In Delhi, air quality is currently up
to four times worse than national ambient air quality standards and water quality is up to
eighteen times worse than the Central Pollution Control Board water quality standards;
55% of households live within 500m of roads with high levels of air pollution (putting
residents at risk of cardiac and respiratory problems), 16% of households lack access to
drinking water taps (putting residents at risk of waterborne illnesses), 6% lack access to
36


latrines, and 8% are using solid fuels for cooking (NFHS, 2006).
While an understanding of health effects associated with provision of
infrastructures and infrastructure-related environmental conditions would indeed be
beneficial, such a database for Asian cities does not yet exist. Meanwhile, various threads
of evidence on infrastructure and environment-related health impacts in Asian cities have
evolved in recent decades. This is particularly the case for outdoor air pollution in Asian
cities with now well over 140 studies (most conducted in China) published in peer-
reviewed literature presenting original estimates of health effects of outdoor air pollution
(HEI PAPA, 2004). At the same time, models that connect fossil combustion from
transportation, power generation, industrial processes with air quality and public health
are highly limited in their application.
Methods
This study conducts literature review of existing epidemiology, infrastructure, and
environmental exposure studies and models, with emphasis on documentation of the
impact of social disparities as they modify health effects estimates and outcomes.
Evidence of associations between current urban health outcomes and infrastructure /
environment-related factors are reviewed to determine the state of evidence and gaps in
forming a database of health effect estimates specifically for Delhi, India and Asian cities
(while presenting additional studies where useful that are often global or from North
America or Europe). The goal of developing such a database is to begin developing first-
order computations of health benefits from infrastructure interventions. Gaps identified in
the literature inform where bottom-up study is needed.
37


In a recent study by Sperling & Ramaswami (2012), major categories of urban
health outcomes in Delhi are traced to infrastructure (e.g. water, energy, transport) and
infrastructure-related environmental factors (e.g. air and water quality). These major
categories include cardiovascular disease, respiratory, airborne infectious disease
pathogens, waterborne pollution and health risks, transport accidents, cancer, and
diabetes. To build on this work, this review takes the approach of assessing quantitative
associations between infrastructure and health using the case of Delhi, India and global
cities where relevant studies are available. The review summarizes ~80 studies presenting
infrastructure and infrastructure-related mortality risk factors and health effect estimates.
Literature Review
While not all potential risk factors are addressed in this literature review, as this
would be nearly impossible, we highlight key risk factors and epidemiological evidence
on health effect estimates primarily from cities in India, Asia, and the USA for
comparison. In all cases, we note whether the study, in developing its health effect
estimates, consideres differential risk by age, gender, ethnicity, income levels, education
levels, and access to health services. Conclusions reached and study limitations where
explicitly mentioned by authors themselves are presented. Where primarily urban
health studies from Delhi and Asia are lacking, we identify health effect estimates for
infrastructure-related risk factors from studies in US, Europe, and OECD countries.
Results: Risk Factors and Health Effect Estimates
Infrastructure-related mortality risk factors and epidemiological HEEs are described
below for the identified seven major categories of infrastructure-related health outcomes
38


and are presented in the order of cardiovascular disease, respiratory, airborne infectious
disease, waterborne pollution and health risks, transport accidents, cancer, and diabetes.
Cardiovascular Diseases (including Heart Diseases and Heart Attacks): The major
mortality risk factor reported from the
literature for this health outcome are
emissions of particulate matter, which can
be ten times more harmful than ozone
exposure levels. In fact, more than 500,000
people die every year from diseases related
to air pollution, often attributed to
particulate matter (PM10) in outdoor air
(WB, 2000).
One study, relevant to Table 1,
notes that PM10 sources vary, and multiple
interventions are therefore needed to reduce
Definition: Particiriate matter (PM),
in the form of PM less than 10 /urn and
2.5 urn in aerodynamic diameter (also
referred to as PM 10 and PM2.5), is
inhalable material emitted directly from
motor vehicles, power plants, and other
sources or formed in the atmosphere
through reactions with gaseous
emissions (eg, nitrogen and sulfur
oxides [NOx and SOx] react to form
nitrates and sulfates, respectively).
Although the health effects of PM have
been of concern for many decades,
short-term and long-term epidemiologic
studies published in the United States
and Europe in the 1990s found
associations of PM with increased
morbidity and mortality at ambient
levels below the national air quality
limit values at the time, the basis of
action in both the European Union and
the United States to establish more
PM10 concentrations. Garg (2011) finds that 31% of PM10 emissions in Delhi, India are
attributable to fossil fuel combustion and 11% to biomass combustion, and 17% from
other sources of ambient dust such as construction activities. However, the dominant
sources can vary across cities, particularly the newly industrializing ones. Such source
information is essential to estimate the impact of infrastructure interventions (e.g. mode
shifts to transit) on air quality given that most Asian cities do not typically have transport
models or industrial emission models.
39


Studies shown in the below tables have (in some cases) identified (quantitatively)
differential risk by controlling for and also differentiating findings across subpopulations,
many of whom often have different exposures to hazards based on their different
infrastructure conditions (e.g. energy, water, sanitation, transportation, buildings),
environmental conditions (e.g. air and water quality, seasonal variation, weather and
climatic changes) and also biological (e.g. age, gender, ethnicity) and socio-economic
conditions (e.g. income, education, access to healthcare). These subpopulations may also
be more susceptible or vulnerable to the resultant health effects (WHO, 2010).
For the available health effect estimates for CVD, many of the studies are short-
term exposure studies with only a few long-term studies (e.g. the Pope et al., 1995
American Cancer Society cohort study). In the table, each study identified is
characterized by the HEE findings, whether the study was short or long-term, the
exposure concentrations for the various study localities, and the models used. HEEs in
many cases are developed as it relates to sensitivity by age, SES factors such as income,
and whether having asthma or not (e.g., Pope, 2011). Meanwhile, none of the identified
studies consider HEEs in terms of sensitivity related to inadequate or delayed access to
healthcare often a common issue in Asian cities (Goli et al., 2011).
An important limitation of this review is that many additional studies may exist in
the literature as air pollution remains a major investigation field and action domain for
improving public health in cities globally. The research articles obtained were through a a
search of air pollution and mortality initially focused on only a couple Indian cities,
followed by Asian cities, then the US and Europe where larger bodies of evidence exist.
40


Table III. 1 Available Health Effect Estimates for CVD and Total Mortality Risk.
Health Hazards Epidemiological Health Effect Estimates / Dose-Response Relationships Quantification of Differential Risk by: Age (A), Gender (G), Ethnicity (E), Education (Ed), Income (I), Season (S), hlth-care access (ATH)
PM10 (Indian cities) (1) Cropper et al. 1997 (Delhi, India): 0.43% increase in CVD mortality & 0.23% increase in total mortality per 10ug/mA3 increase in PM10; Findings based on 1991 to 1994 study of short-term exposure; avg. total suspended particulate (TSP) was 375 ug/mA3 over study period (~5 times higher than WHO annual standard); levels also exceeded WHO 24 hour standard 97% of the time. Model used'. Poisson. Study Limitation/s'. results based on limited cause-specific mortality data representing only 25% of deaths during study period; TSP no longer routinely monitored or considered good marker for health effects (Balakrishnan et al. 2010) (2) HEI short-term study by Rajarathnam et al., 2010 (Delhi, India): 0.15% increase in All-Natural-Cause Mortality per 10 ug/mA3 in PM 10 (all natural-cause excludes the following causes of death: accidents and suicides/homicides); Study of short-term exposures from 2002-2004; annual mean PM10 concentration: 170 ug/mA3 (~3 times higher than Central Pollution Control Board Natl. Ambient Air Quality Stndrd. of 60 ug/mA3 for residential areas); Model used: quasi-Poisson regression model; Study Limitation/s: AQ data from 9 of 10 stations available for <100 days per year; no monitoring on weekends/holidays; medically-certified cause of deaths for only 55% of total deaths (3) HEI short-term study by Balakrishnan et al., 2010 (Chennai, India): 0.44% (95% Cl: 0.17-0.71) increase in All-Natural- Cause mortality per 10ug/mA3 increase in PM10 concentrations; Daily avg for PM10 at 92 ug/mA3; often exceeded NAAQS 100 ug/mA3 across ind., resid., & commercial zones; Model used: quasi-Poisson generalized additive model; Study Limitation/s: stations monitor only 100 120 days / yr; no monitoring on weekends/holiday s; HARMONIZATION: Range of 0.15% 0.44% increase in all- natural-cause mortality per 10ug/mA3 increase in PM 10 The (#) below corresponds to study identified in left column. The following studies only address: (1) A (2) A; G (3) A; G; S The studies do not address E. ED. I. & ATH.
41


Table nil (contd)
PM10
(Asian, US,
& Other
Global
Cities)
(4) HEI PAPA 2004 meta-analyses of short-term studies (Seoul, Honk Kong, Bangkok, Inchon): 0.41% increase in
mortality per 10ug/m3 increase; 29 European Cities: 0.6% increase per 50ug/m3 increase; concentrations
ranged from 10 to 290ug/m3; Poisson model; (Katsouyanni et al., 2001); 90 US Cities (Samet et al. 2000):
0.41% increase
(5) Chen, 2011 (Beijing, Shanghai, and Shenyang): short-term study finding a 10-pg/m3 increase in 1-day lagged
PM10-2.5 was associated with a 0.25% (95% Cl: 0.08 to 0.42) increase in total mortality, 0.25% (95% Cl: 0.10
to 0.40) increase in CVD mortality, and 0.48% (95% Cl: 0.20 to 0.76) increase in respiratory mortality. Avg.
daily concentrations of PM10-2.5: 101 pg/m3 (Beijing, 2007-2008), 50 pg/m3 (Shanghai,2004-2008), and
49 pg/m3 for Shenyang (2006-2008); GLM Poisson model used
(6) Bell et al. 2013 (Global review: 31 European studies (12 in Italy), 24 in Asia, 8 in Canada, 7 in Latin America,
1 in Russia / Australia with some as single and multi-city studies): short-term exposure studies of
subpopulations: consistent evidence that elderly experience higher risk of PM-associated hospitalization / death
and suggestive evidence that those with lower education, income, or employment status have higher death risk.
Meta-analysis findings: 0.64% (95% Cl: 0.50-0.78) increase in death risk for older populations compared with
0.34% (95% Cl: 0.25-0.42) foryounger populations per 10ug/m3 increase in PM10. Women: 0.55% (95% Cl:
0.41, 0.70) vs. Men: 0.50% (95% Cl: 0.34-0.54) (not statistically significant difference).
(7) Yang et al. 2013. (Beijing after the 2008 Olympics): time-series analysis using generalized additive model:
1.8% increase (95% Cl: 1.21-2.40) for 10 ug/m3 increase in PM10; mean daily average PM10 concentrations
were 121.04ug/m3 from 2009-2010; CVD mortality increase per 10ug/m3 for males: 0.96% vs. females:
2.64%; elderly age 65+ : 1.97% vs. 45 and below: 0.53%
(8) Ostro, 2004 (Global Cities Short-Term Studies Review): All-natural-cause mortality % increase per 10ug/m3
change in PM10: 0.8%; 1.7% in CVD mortality (Bangkok: Ostro et al. 1999); 1.83% (Mexico City: Castillejos
et al., 2000); 1.1% (Santiago: Ostro et al., 1996); 0.8% (Incheon: Hong et al., 1999); 1.6% (Brisbane: Simpson
et al. 1997); 0.95% (Sydney: Morgan et al., 1998).
(9) Schwartz, 1994 short-term studies; All-Natural-Cause Mortality: 0.4% (e.g. Santa Clara)-0.9% (Tennessee)
increase in total mortality per 10 ug/mA3 increase in PM 10 (study includes Philadelphia, Detroit, Minneapolis,
Utah Valley, St Louis, etc); average concentrations of TSP ranged from 56 to 111 pg/m3; Model used: Poisson
(10) Dominici & Samet et al., 2002 (88 US Cities): .5% increases in total mortality per 10ug/m3 increase in PM 10
(i.e. Denver, CO: ~0.5%); Short-term exposure effect studies using the National Morbidity, Mortality, and Air
Pollution Study (NMMAPS) data base from 1987-1994; hierarchial linear model used
(1 l)Dockery at al. 1993: Harvard Six-Cities (Steubenville, St. Louis, Harriman, Watertown, Topeka, Portage) long-
term cohort study with 16 year follow-up of 8111 adults: All-cause-of-death risk ratio of 1.26 (95% Cl: 1.08-
1.47) for most vs. least polluted city; Model used: Cox Proportional Hazards Survival model
(12)Pope et al. 1995 (151 U.S. Metro areas): American Cancer Society long-term cohort study (the original ACS
study)from 1982 to 1989 of 552,138 adults: All-cause-of-death risk ratio of 1.17 (95% Cl: 1.09-1.17);
HARMONIZATION: Range: 0.25% 0.9% increase in All-Natural-Cause mortality per 10ug/mA3 increase in
PM10
(4) Only A is
addressed in
this study
(5) Only A is
addressed in
this study
(6) A, G, Ed, I,
and
Employment
(7) A; G
(8) A / E are both
addressed;
ATHis
acknowledged
as important,
but not
addressed
explicitly.
(9) DR not
addressed in this
study
(10) DR not
addressed in this
study
(11) A; G; E;
smoking;
occupational
exposure; BMI
(12) E and
smoking
addressed
42


Table nil (contd)
PM2.5 (13) Ostro et al., 2006: % increase in total mortality per 10ug/m3 change in PM2.5: 0.6% (nine California counties short-term exposures study from 1999-2002; mean daily PM2.5 concentrations: 14 to 29 ug/mA3; Model used: Poison multiple regression model using natural or penalized splines; Study Limitation/s: exposure measurement error) and 1.4% (Mexico City short-term study by Boria-Aburto et al.. 1998); 0.6% in CVD mortality and 2.2% in respiratory mortality, respectively (nine CA counties) (14) Krewski et al. 2009 (US nationwide, NYC & LA): % mortality increase per 10ug/m3 change in PM2.5 (1979- 1983: mean concentrations: 21.2 ug/m3;): All-Natural Cause Mortality: 0.04%; Ischemic Heart Disease (IHD): 0.18%; Cardiopulmonary: 1.09%; (1999-2000; mean concentrations: 14.02): All Causes: 0.06%; IHD: .24%; CP: 0.14%; random effects Cox model used (15) Laden et al. 2006 (Extended Follow-up of Harvard Six US Cities Study): % increase per 10 ug/mA3 change: All- Natural-Cause Mortality: 0.16%; CVD: 0.28-1.0% (16) Miller et al. 2007 (Long-term Cohort study separate from ACS cohort: 65,893 postmenopausal women without preexisting CVD in 36 U.S. metropolitan areas: Difference of 10 pg/m3 PM2.5 associated with 24% increase in risk of a cardiovascular event (relative risk 1.24; 95% Cl 1.09-1.41) and 76% increase in risk of CVD mortality (RR 1.76; 95% Cl 1.25-2.47) HARMONIZATION: Range of 0.04% 1.4% increase in All-Natural-Cause mortality per 10ug/mA3 increase in PM2.5 (13) A; G; E; Ed; and diabetic subpopulation. I, AH, Snot addressed. (14) E (e.g. black, white, hispanic); I (e.g. unemployment, median HH income; income disparity; Ed (e.g. < high school education). AH is not addressed. (15) TBD (16) A; G; E; diabetes
N02 (2) Rajarathnam et al., 2010: 0.84% increase in total mortality per 10ug/mA3 change (7) Yang et al. 2013: 2.63% increase in nonaccidental mortality per 10ug/m3 change (2) A; G (7) A; G
NOx (17) Greenbaum, 2011: 0.65% increase in mortality per 10ug/mA3 increase (17) DR not addressed
S02 (2) Rajarathnam, 2010: Delhi: no significant effect for S02 concentration changes (2) DR not addressed
Ozone (18) Smith, 2009: CVD: 0.22% change in mortality per 1 ug/mA3 change in pollutant (18) DR not addressed
Ozone (19) WHO, 201 l(Europe): heart disease: .3-.4% mortality increase perlO pg/m3 change (19) DR not addressed
Sulphate (18) Smith, 2009: % change per 1.0 ug/mA3 change in particle sulphate concentrations: All-Cause: 0.12-0.88; CVD: 0.09-0.36; Cardiopulmonary: 1.01; Respiratory: 0.37-0.70 (18) DR not addressed
Solid Fuels (17) Smith, 2000 (India): CVD: 0.287% increase in mortality for children <5 exposed to solid fuels (17) DR not addressed
43


Table nil (contd)
Black (19) Gan, 2011 (Vancouver): An elevation in black carbon (0.97xl0-5/m) was associated with a 4% increase in CHD (19) DR not
Carbon mortality. addressed
Diet (20)WHO GBD, 2009: Insufficient intake of fruit & vegetables (prevalence measure of 5 servings / day) is estimated to cause 11% of ischaemic heart disease deaths worldwide. (20) G, I
Note: Many estimates are based on specific thresholds of environmental pollution exposures, with estimates presented as relevant
to cardiovascular disease (CVD), cardiopulmonary, coronary heart disease (CHD), and all-natural-cause mortality.
44


Airborne Pollution Risks for Respiratory Mortality: The major mortality risk
factors and health effect estimates reported from literature for respiratory deaths (e.g.
tuberculosis, bronchitis, and asthma including ICD8 460-519, and excluding 463, 464,
and 474) are similar to those for CVD. Outdoor air pollutants such as PM, ozone, and
sulphate are all important risk factors. Other risk factors include indoor air pollutants
from use of solid fuels for cooking and associated black carbon emissions. The HEEs in
Table 2 are presented as relevant to respiratory mortality, which includes deaths due to
acute respiratory infection, asthma, pneumonia, and bronchitis. In most cases, age is
considered and in some cases gender, education, and seasonal variation is addressed.
Table III.2 Available Health Effect Estimates for Respiratory Mortality.
Health Hazards Epidemiological Health Effect Estimates / Dose-Response Relationships Differential Risk (DR) by: Age (A), Gender (G), Ethnicity (E), Education (Ed), Income (I), Access to Healthcare (AH), Seasonal (S)
PM10 (Indian cities) (1) Cropper et al. 1997 (Delhi): 0.31% increase in respiratory mortality per 10ug/mA3 increase; in PM10; Findings based on 1991 to 1994 study of short-term exposure; avg. total suspended particulate (TSP) was 375 ug/mA3 over study period. Model used: Poisson model. (1) A
PM10 (Asian, US,& Global Cities) (2) Ostro et al. 1999 (Bangkok): 3-6% increase in respiratory mortality and a 1 -2% increase in all-natural-cause mortality per 10ug/mA3 increase in PM 10 (3) Tellez-Roio et al. 2000 (Mexico Citv): 2.9-4.1% increase in resp. mortality per 10ug/mA3 PM10 increase (2) A (3) A; AH E. ED. I are not addressed
in these studies
45


Table III.2 (contd.)
PM2.5 (4) Laden et al 2006 (Extended Long-Term Follow-up of Harvard 6 US Cities'): Respiratory mortality increases bv 0.08% per 10ug/mA3 change in PM2.5 exposure levels; all-natural cause mortality by 0.16%; 1979-1997 annual avg. PM2.5 concentrations ranged from 10 to 40 ug/mA3 ; model used: Cox proportional hazards regression model (5) Franklin et al 2007 (27 US Communities short-term case-crossover study from 1997 to 20021: 1.21% increase in all-cause mortality; 1.78% increase in respiratory mortality; and 1.03% increase in stroke related mortality with a 10 ug/m3 increase in previous days PM2.5; 9.3 in Palm Beach, FL to 28.5 in Riverside, CA with mean concentration across all communities:!5.7ug/m3; model used: conditional logistic regression model (6) Ostro 2006 (9 California Counties short-term exposure study from 1999 to 2002): 10 ug/m3 change in 2-day average PM2.5 cone corresponded to 0.6% increase in all-natural-cause mortality, 0.3% in CVD; 1.3% in respiratory. Mean daily PM2.5 levels ranged from 14 pg/m3 in Sacramento and Contra Costa Counties to 29 pg/m3 in Riverside County, exceeding the U.S. EPA annual average PM2.5 standard of 15 pg/m3 in six of the nine counties; Models Used: random effects model / penalized and natural spline model (7) Englert, 2004 (Re-analysis of three long-term exposure cohort studies: Harvard 6 US Cities Study; ACS study, and AHSMOG study): key findings: difference in lung cancer estimates by gender is considerable (RR = 2.14 (male) vs. 1.20 (female) per 10ug/m3 increase in PM2.5; highest effects also linked with lowest education,especially for lung cancer (4) A; G; Ed (5) A; G (6) A; G; E; Ed; S; diabetic (7) A; G; Ed; I; and smokers
Ozone (8) Lancet Short-Lived GHG Smith, 2009 (US); 0.57% change in resp. mortality per 1 ug/mA3 change in pollutant (9) Ji, 2010; Emergency hospitalizations for total respiratory disease increased by 4.47% per 10 ppb 24h ozone among elderly w/out adjustment for publctn bias (2.97% w adjustment) (10) Bell et al, 2004 (US Natl Avg: 95 large US urban communities, including ~40% of the total US population): A 10-ppb increase in the previous weeks ozone was associated with a 0.52% increase in daily mortality and a 0.64% increase in cardiovascular and respiratory mortality. (8) DR Not Addressed (9) Study Addresses A&S. G. E. Ed. I, and AH not addressed (10) Study Addresses A&S. G. E. Ed, I, and AH not addressed
Solid Fuel Use (11) Smith, 2000 (India): ARI: 0.381% increase in mortality for children <5 exposed to household use of solid fuels; TB: 0.366% increase for children <5; Asthma: 0.315% increase (11) A
Particle Sulphate (12) Smith, 2009 (Multi-city study in 6 Californian counties): 0.7 % change in resp. mortality per lug/mA3 change in pollutant (13) Dockery et al.,1993 (US): 1.26% change in all-natural cause mortality for sulfate particles exposure between the between the most polluted city and the least polluted city (range: 4.8 to 12.8 pg per cubic meter) (12) DR not addressed (13) Only Ed addressed.
Malnutriti on (14) James, 1972 (San Jose, Costa Rica cited by CPCB, 2008 Delhi Study on Ambient Air Quality, Resp. Symptoms & Lung Function of Children): Malnourished children experience 2.7 times more bronchitis & 19 times more pneumonia than normal-weight properly nourished children (14) Only age (and weight) addressed
46


Airborne Infectious Disease Risks: HEEs have been identified in Table 3
specifically for the infrastructure-related risk factor of overcrowded sleeping conditions.
The estimates are presented as relevant to airborne infectious diseases, which includes the
following diseases: tuberculosis, influenza, diptheria, whooping cough, and meningitis.
As shown in the Table below, the only health effect estimates identified so far for
a global south city in terms of overcrowding and its impacts on TB mortality was for Sao
Paolo, Brazil (Antunes et al., 2001). Additional HEEs were developed for Canadian first
nation communities (Clark et al., 2002), in the Bronx, New York and for boroughs of
London in the UK (Lienhardt, 2001). While associations between overcrowding and
physical health have been well documented qualitatively including over 40 studies (UK
Deputy Prime Minister Office, 2004), limited evidence exists on quantitative health effect
estimates, particularly for Asian cities. The literature reviewed to date suggests an
association between overcrowding and TB in both children and adults, with
overcrowding in childhood affecting aspects of adult health.
47


Table III.3 Available Health Effect Estimates for Airborne Infectious Disease Mortality.
Health Hazards Health Effect Estimates Differential Risk (DR) by: Age (A), Gender (G), Ethnicity (E), Education (Ed), Income (I), Access to Healthcare (AH), & Seasonal (S)
Crowded Household Sleeping areas (1) Antunes & Waldman 2001 ((Sao Paolo, Brazil): estimate a 14% increase in tuberculosis (TB) mortality over a five year period due to an average increase of one additional dweller per bedroom (over one person) in the household. (2) Lienhardt, 2001 (Bronx, New York): under age-5 children living in severely crowded areas were about five times more likely to develop tuberculosis (adjusted for HIV status) than children living in areas with limited or no crowding. From 1982 to 1991, tuberculosis notification rates in London boroughs (United Kingdom) increased by 12 % for each % increase in the number of persons living in overcrowded accommodations (3) Clark, Riben et al., 2002 (Canadian first nation communities): Morbidity rate was 18.9 per 100,000 in communities with average of 0.4-0.6 persons per room (ppr), while communities with 1.0- 1.2 ppr had a rate of 113.0 per 100000. Increase of 0.1 ppr in a community was associated with a 40% increase in risk of > 2 TB cases occurring, while an increase of $ 10 000 in community household income associated with 0.25 the risk, and being an isolated community increased risk by 2.5 tunes. (4) Baker, 2008 (New Zealand): for every 1% increase in the average crowding level of a Census Area Unit (CAU) there would be a 5% increase in the expected TB count. (1) A; I; Other: foreigners / migrants significantly higher mortality risk than persons bom in Sao Paolo (88.57% increase) (2) A; G; E; Ed; I; & AH (3) I, E (4) A, I, E, Other: migrants (the total TB cases notified in the 2000- 2004 period were 1898. About 60% of these cases occurred in migrants bom in high incidence countries).
Transport Crashes: Road fatalities are considered to be an emerging epidemic, and
are listed in the top ten causes of global mortality (GBD, 2011). Recently, theres been a
trend towards the term crashes replacing accidents in the transport literature and is
therefore the term used in the remainder of this paper. WHO (2004) divides the main risk
factors for road fatalities into the following major categories:
Factors influencing exposure to risk (e.g. economic factors, land use planning
practices which influence the length of a trip or travel mode choice, high-speed
motorized traffic mixed with vulnerable road users, insufficient attention to road
layout and design, etc)
48


Factors influencing crash involvement (e.g., inappropriate or excessive speed,
presence of alcohol, medicinal, or recreational drugs; fatigue; being a young male;
travelling in darkness; defects in road design, layout and maintenance which can
also lead to unsafe road user behavior; vehicle factors such as braking, handling
and maintenance; inadequate visibility due to environmental factors, etc),
Factors influencing crash severity (e.g. inappropriate or excessive speed; seat-
belts and child-restrains not worn; roadside objects not crash protective; etc)
Factors influencing severity of post-crash injuries (e.g., lack of appropriate pre-
hospital care, appropriate care once in hospital emergency rooms, difficulty
evacuating people, hazardous materials leakage, fire presence from collision, etc)
For two-wheeler vehicle users, one major risk factor identified is crash helmets not
being worn by users; for transport fatalities, gender and age play a large role (Arnett,
2002; De Hartog, 2010; Fischbeck et al., 2012). In first row of table X, the risk factor of
mode choice is analyzed with respect to avoiding exposure to risk by shifting from road
use to rail use. The health effect estimates due to mode choice are developed using the
National Crime Records Bureau (NCRB) transport fatalities 2008 data for Delhi, India.
Table III.4 Available Health Effect Estimates for Road Safety.
Health Hazards Health Effect Estimates Differential Risk
Mode Choice /Road Safety (1) NCRB, 2008 (Delhi): From Auto to Rail: 9.7 deaths / 1% mode shift; From 2-Wheeelers to Rail: 20.6 deaths / 1% shift; 5 deaths / 1% reduction in ped fatality risk via shift to rail (1) DR not addressed
Lack of Helmets (2) Ichikawa, 2003 (Municipality of KhonKaen Province, Thailand): After enforcement of the helmet act, helmet-wearers increased five- fold; head injuries decreased by 41.4% and deaths by 20.8%. (2) by helmet wearing
Lack of Seatbelts (3) Mohan, 2004 (India): Use of seat belts, child seats and airbag equipped cars can reduce car occupant fatalities by over 30% (3) A & G; by seatbelts
Lack of Seatbelts (4) WHO, 2009 GHR p.33 (Global): Seatbelts, when used correctly, estimated to reduce risk of death in crash by 61% (5) by seatbelts
49


Table III.5 (contd.)
Lighting (5) Elvik, 2006 (meta-analysis): Using daytime running lights (DRLs) on cars shows that such use reduces the number of multi-party daytime accidents by about 10-15% for cars using DRLs; Similar results for DRLs by motorcyclists in Serembam, Malaysia (Radin, 2003) (4) DR not addressed
Vehicle Crashes (6) Novoa et al., 2009 (Spain): Graduated licensing system (31% road traffic injury reduction); vehicle electronic stability control system (2 to 4% reduction); area-wide traffic calming (0 to 20% reduction); and speed cameras (7 to 30% reduction) (6) DR not addressed
Diabetes: The major infrastructure-related health effect estimates reported from
the literature for this infrastructure-related health risk category includes levels of physical
activity, obesity, and diet. Most studies identified to date, are not Delhi or South Asia
specific and further epidemiological study is therefore needed.
Table III.5 Available Health Effect Estimates for Diabetes.
Health Hazards Epidemiological Health Effect Estimates Differential Risk by:
PM2.5 (1) Pearson, 2010 (US): short-term exposure effect study from 2004-2005 finding a 1% increase in diabetes prevalence seen with a 10 pg/m3 increase in PM2.5 exposure; comparing US counties with highest (annual mean concentration quartile of 13.8-17.1 ug/m3; EPA limit is 15ug/m3) compared to lowest levels (annual mean concentration quartile of 2.5-7.7ug/m3) of PM2.5 exposure: resulted in a >20% increase in diabetes prevalence (1) A, E, I, Ed, Obesity rates (BMI>30kg/mA2; physical activity), population and fast food establishment density, health insurance
Inadequate Physical Activity (2) De Hartog, 2010 (Hu et al, 2004 Finland): 25% reduction in diabetes mortality from walking/cycling to work (unadjusted for other domains of phy act); (Matthews, 2007 Shanghai) 21% 34% reduction in diabetes mortality for Chinese women cohort (adjusted for other phy act) (2) A, G; cyclists vs. car commuters with tradeoffs in DR by exposure to air pollution, traffic accidents, higher physical activity
Obesity (3) (WHO, 2000 (Obesity: Preventing & Managing the GlobalEpidemic): Diabetes Mellitus: ~64% of male and 74% of female cases of diabetes could have been prevented if no one had had a BMI over 25 (3) A, G, I, genetic susceptibility
Diet (4) Carter et al., 2010 (UK): Summary estimates showed that greater intake of green leafy vegetables was associated with a 14% reduction in risk of type 2 diabetes. (4) Only A, G, E, and Ed
50


Waterborne Pollution, Pathogens and Health Risks: The major infrastructure-
related mortality risk factors reported from the literature for water and sanitation-related
health outcomes (e.g. diarrhea, malaria, and malnutrition-related mortality) include water
quality (and removal of
waterborne disease
pathogens), sanitation,
water supply,
household water
storage, drainage,
management of solid
waste, and specific
thresholds of
environmental pollution
exposures. HEEs have
been a challenge to
identify for many of
A Key Challenge to Developing this Evidence Base:
Addressing Confounding Risk Factors in Water
Sanitation Studies: Inadeqaute control of confounding
variables is a major problem in all be a few of the
[water and sanitation-related health impact] studies.
The factors most likely to confound results in studies of
water or santitation are housing (pathogen survival),
crowding (hand-to-hand contamination), age (acquired
immunity), sex (activity and contact with environment),
breastfeeding (exposure to pathogens and pathogen
viability), health education (use of services), seasonality
(availability of other water sources), rural-urban
(exposure to pathogens), migration (exposure to new or
more pathogens), income (better food and medical care),
diet (quality and quantity), other infections (e.g. malaria
causing poor nutrition status), and distance to medical
care (whether proper attention is sought) (Esrey and
Habicht 1985:73-74). Completely controlling for the
large number of confounding variables that might
influence the various health indicators is an impossible
task in historical research... [in addition], many studies
assume that the presence of a particular water supply is
synonymous with use of that facility
- Van Poppel et al., 1997
these risk factors (not all) due to the existence of many other confounding risk factors
(see blue text box on right).
Additional risk factors studied include house design, wastewater treatment, and landfill
management (a more specific aspect of solid waste management).
A study by Van Poppel et al. 1997 has identified several other risk factors specific
to infant and child mortality in the Dutch municipality of Tilburg. While water supply
was not found to have an impact on the mortality of children, their study presents
51


interesting findings (per their relative risk table below) on risk factors such as sex of
child, mothers age at childbirth, childs order of birth in family (e.g. firstborn child), and
season of birth. As shown, male children, children with higher birth order, children who
were quickly followed by the birth of another child, children bom in the summer period,
and children bom to very young and old mothers had higher relative mortality risks:
Proportional hazards coefficients for infant and childhood death, based on Cox-regression
model (N-4863)
Variable Relative risk
Water supply Piped water 1.01
Other 1.00
Sex Male 1.16*
Female 1.00
Birth order of child One 0.73**
Two or three 1.00
Four or higher 1.37***
Interval to next child in months Less than 16 1.65***
16-24 1.00
25 or more 0.68*
Socio-economic group Upper class 0.94
Middle class 0.98
Lower class 1.00
Income High 0.74*
Middle 0.90
Low 0.93
Very low 1.00
Season of birth Summer 1.24***
Winter 0.95
Other 1.00
Number of rooms in dwelling Rented dwelling with two or three rooms 1.06
Four or more rooms and own dwelling 1.00
District Unhealthy 1.25***
Other 1.00
Age of mother 1.00
Age of mother squared 1.00*
Significant at 0.05 level; ** significant at 0.01 level; ** significant at 0.001 level.
Figure III. 1 Example Relative Risk Findings for Infant and Child Mortality
52


Table HI.6 Available Health Effect Estimates for Waterborne Disease Pathogens
Health Hazards Epidemiological Health Effect Estimates Differential Risk (DR) by: Age (A), Gender (G), Ethnicity (E), Education (Ed), Income (I), Seasonal (S), etc
Water Quality (1) Caimcross, 2010 (meta-analysis): Diarrhea morbidity risk reductions of 17% associated with improved water quality (with water supply from public source) see Tbl 41.7; 63% diarrhea risk reduction associated with house piped water connection (Tbl 41.7) (4a) Esrey et al., 1991 (adapted by Caimcross, 2006): Water Quality Only Intervention: 15% median reduction in diarrheal morbidity (6) Fewtrell et al. 2005 (synthesis of five urban sutides): Water treatment at point of use (inch chemical treatment, boiling, pasteurization, and solar disinfection) produced a relative risk of 0.74 (0.65-0.85) HARMONIZATION: 15% to 26% reduction in relative risk of diarrheal morbidity (1) A (children) (2) A, G (3) A, G, I, S, socioeconomic group, birth order of child, dwelling type (6) A; E; urban / rural (4a) A; G; Ed (literacy); breastfed or not; (4b) A (5) A; Ed; G; Proxy of Income; Marital Status; Household Size; Urban & Rural; birth order / spacing; (6) A; Ed; urban / rural (7) A; Ed; urban / rural; seasonal (8) A (e.g. age of mother); Ed; I; household size; marital status; urban / rural; child vaccinations
Sanitation (1) Caimcross, 2010(meta-analysis): improved exreta disposal: 36% reduction in diarrhea morbidity
Water Supply (2) Caimcross, 2006: a 50% reduction expected from excellent water supply alone (3) Van Poppel et al., 1997: child mortality RR of 1.01 for piped water versus other (relatively no impact)
Water, Sanitation and Hygiene (WASH) (4a) Esrey et al., 1991 (adapted by Caimcross, 2006): Water quantity only intervention: 20% median reduction in diarrheal morbidity; Water quantity & quality intervention: 17% median reduction in diarrheal morbidity; Sanitation Only: 36% reduction; Water & Sanitation: 30% overall & 55% reduction in child mortality; Hygiene Promotion: 33% reduction (4b) Esrey et al., 1985: Review of 67 studies on water supply & sanitation interventions / their impact on diarrheal diseases. For <5 years of age, improved water supply & sanitation demonstrated a median reduction of 27% and 30% in morbidity and mortality rates, respectively, when studies of better quality were used. (5) Gunther, 2011 (Global): average mortality reduction achievable by investment in full household coverage with water and sanitation infrastructure is 25 deaths per 1000 children bom across countries (6) Fewtrell and Kaufmann, 2005 (Meta-analysis of 46 studies): water, sanitation, and hygiene interventions had a similar degree of impact on diarrheal illness, with relative risk estimates from the overall meta-analyses ranging between 0.63 and 0.75; (7) Waddington et al. 2009 (Meta-analysis of 71 studies): water supply: RR of 0.73 to 0.75; water quality: RR of 0.65-1.09; Sanitation: RR of 0.68 to 0.78; Hygiene: 0.53 to 0.68; Multiple interventions: 0.67 (8) Gunther and Fink (2010): water and sanitation infrastmcture lowers the odds of children suffering from diarrhea by 7-17% and reduces mortality risk for children under age five by about 5-20 %(findings based on review of 172 DHS reports from 70 countries; all are short-term effects studies); Models used: Ordinary least squares / logit regression model for child diarrhea; Weibull survival model for under-five mortality HARMONIZATION: Having water and sanitation infrastmcture results in a 5% to 30% reduction in the relative risk of under-five diarrheal mortality
53


Table HI.6 (contd.)
Unsafe Wat-San (9) WHO, 2009 GHR (Global): Unsafe water and sanitation estimated to cause 88% of diarrheal deaths (7) I; A (8) See above (9) A; Ed; In; vitamin and mineral deficiencies (10) Population density; Climatic conditions; (11) A, I, Ed (literacy) (12) A (13/14) A, G, I, Ed (15) House type; waste generated; private or public facilities (16) Having insecticide treated bed nets
Water Supply (8) Gunther and Fink (2010): Private access to water supply decreases relative likelihood of diarrhea by ~14%, whereas the odds of diarrhea are reduced by only about 7% if the household uses a public pipe and/or a public well/borehole. (10) White, Bradley, and White, 1972: Typhoid fever: 80% reduction expected from excellent water supply; Trachoma: 60% reduction expected from excellent water supply; Scabies: 80% reduction expected from excellent water supply (11) Esrey and Habitch, 1986: among the illiterate group an infant was 1.36 times more likely to die if no piped water was available compared with having piped water; while among the literate group the relative odds of an infrant dying was more than doubled if no piped water was available
Sanitation (12) Messou et al. 1997 (Ivory Coast; rural setting): Diarrhea Mortality Risk Reduction: 85% (13) Esrey et al. (1991) found water supply and sanitation reduced prevalence of ascariasis by a median of 28 percent (range 0 to 83 percent) and of hookworm infection by 4 percent (0 to 100 percent). Those reductions are likely caused by sanitation rather than by the water-supply improvements. Indeed, three of the nine positive studies of ascariasis and three of the five positive studies of hookworm involved sanitation alone. It is also likely the effect of excreta disposal on Trichuris infection is similar to that on ascariasis (Henry 1981). (14) Esrey (1996): Diarrhea reduction of 13-44% for flush toilets and 8.5% for latrines (findings based on eight Demographic and Health Surveys from Bolivia, Burundi, Ghana, Guatemala, Morocco, Sri Lanka, Togo, Uganda); Limitation: selected only 8 out of the 63 DHS surveys that were available in 1995 (8) 13% reduction in diarrhea for having flush toilet vs. open defecation; for households that share toilets with several other households: no significant impact of access to a public latrine/flush toilet on child diarrhea, and the effect on child mortality is only 3.3%; but private sanitation facilities reduce the odds of diarrhea by 10% and the likelihood of dying before the age of 5 by 13% (findings from 172 DHS studies)
Managem ent of Solid Waste (15) Fobil et al. 2011 (Accra, Ghana): On basis of population and waste generation sub-component, strong evidence of a difference in risk of urban malaria mortality between least deteriorated zone (mean fraction = 0.029, 95% Cl: 0.011-0.045) and moderately deteriorated zone (mean fraction = 0.051, 95% Cl: 0.04-0.059) with a relative mortality [RR] = 1.77 between the two zones; also has estimates for modifying housing design / poor water supply/sanitation
Water- insect vectors (16) Insecticide treated bed nets (Glennon et al. Sub-Saharan Africa): Mortality RR: 0.85, 95% Cl .76-.89
54


Additional systematic literature reviews by WHO at a global scale indicate deaths
from diarrhea of children aged less than 5 years represent approximately 19% of total
child deaths (Boschi-Pinto, 2008). Such estimates are useful for quantifying future health
scenarios at city scale if the only data available is under-5 mortality counts with cause not
specified (e.g. DHS / NFHS-3 data in India and Delhi).
Additional relavent literature for developing a database of infrastructure-related
health effects estimates includes developing estimates for population as a whole, not just
under-5 diarrheal mortality. A synthesis by Walker and Black, 2010 highlights the need
for improved diarrhea specific mortality and morbidity data for different age groups
including older children and adults (as deaths for this age group can also be quite high).
Lamberti et al. 2012 also identifies the need for improved understanding of diarrhea
duration and severity across age groups due to significant global diarrhea morbidity, and
improved treatments leading to decreases in diarrhea mortality. A study by Boschi-Pinto
(2008) indicates deaths from diarrhea of children aged less than 5 represent
approximately 19% of total child deaths (Boschi-Pinto, 2008). For the case of Delhi,
India, analyses of death records for 2006 indicate the percentage of diarrheal mortality
relative to total deaths reported by cause for under-5 populations was 0.6% (likely
indicating underreporting) and for adult populations was 0.1% (NCT of Delhi Annual
Births and Deaths Report, 2006).
Although not addressed in the table above of health effect estimates, the Central
Pollution Control Board of India, the US Environmnetal Protection Agency, WHO, and
others also provide drinking water quality guidelines for hundreds of contaminants with
water meeting these standards safe to drink, and water unfit being harmful to health.
55


Cancer: Infrastructure-related health effect estimates for cancer mortality are
shown in the Table below. Our review found health effect estimates mostly available for
lung cancer as it relates to both indoor and outdoor air pollution. In U S. counties with
hazardous waste sites and ground water pollution, as identified by the Environmental
Protection Agency, estimates were also made for gastrointestinal cancers. Other cancer
risks include radon, asbestos, and other hazardous chemical exposures. Findings from
primarily European studies suggest diet, specifically fruit and vegetable intake, and
physical activity can help reduce the risk of cancer.
Table III.7 Available Health Effect Estimates for Cancer.
Health Hazards Epidemiological Health Effect Estimates Differential Risk
PM2.5 (1) Krewski et al. 2009 (US nationwide, NYC and LA): A 10ug/mA3 change in PM2.5 exposure levels: (Study Period 1979-1983):Lung Cancer mortality increases by 0.09%; (Study Period: 1999-2000): Lung cancer mortality increases by 0.14% (2) Laden et al 2006 (Six US Cities): Lung cancer mortality increases by 0.27 % per 10ug/mA3 change in PM2.5 exposure (3) Pope et al 2002 (US): Each 10-pg/m3 elevation in PM2.5 was associated with approximately a 4%, 6%, & 8% increased risk of all- cause, cardiopulmonary, &lung cancer mortality, respectively HARMONIZATION: Mm: 0.14% to Max: 8% increased risk of lung cancer mortality per 10ug/m3 change in PM2.5 exposure (1) E; I; Ed are addressed. AH is not addressed. (2) TBD (3) TBD
S02 (1) Krewski et al. Minimal effect: A 10ug/m3 change in S02 concentrations over 15-yr time window as modified by education: (Less than High School): Lung Cancer: 0.02%; (High School or More): Lung Cancer:- 0.05% (1) See above
Solid Fuels for Cooking (4) Gupta 2001 (Chandigarh, India): Cumulative exposure of > 45 age women to indoor air pollution from use of coal or wood for cooking or heating showed 0.43% increase in developing lung cancer (very few subjects were employed in high-risk occupations). (4) A; smoking vs. non-smoking; occupations; rural vs. urban
Radon Exposure (5) Pavia et al., 2003 (meta-analysis of 17 case-control studies): Residential exposure to radon at 150 Bq/m3 increases risk of lung cancer by 0.24% (5) G; I; time spent at home
Other chemicals (e.g. pesticides; petrochemi cals) (6) Parent et al., 2009 (Montreal, Canada): Evidence of a two-fold excess risk of prostate cancer among farmers with substantial exposure to pesticides [odds ratio (OR)=2.3, 95% confidence interval (Cl) 1.1-5.1], as compared to unexposed farmers. Also some increased risks among farmers ever exposed to diesel engine emissions (OR=5.7, 95% Cl 1.2-26.5). (7) Liu et al., 2008 (Taiwan): significantly higher risk of developing brain cancer for those living in municipalities characterized by highest vs low levels of petrochemical air pollution (OR = 1.65, 95% Cl = 1.00-2.73) (6) A, G, E, Ed, I, occupation histories and exposures (7) A,G, municipalities characterized by high and low levels of exposure
56


Table III.7 (contd.)
Diet (8) Key, 2011 (Europe): For cancers of the oral cavity, pharynx and larynx, Freedman et al (2008a) reported that, compared with people who consumed ~1.5 portions of fruit and vegetables each day, people with intakes of -5.8 portions per day had a relative risk of 0.71 (95% Cl 0.55-0.92), and in a study in Europe of squamous cell cancers of the oral cavity, pharynx, larynx (and oesophagus), Boeing et al (2006) reported that people with a high intake of fruit and vegetables (-7.7 portions per day) had a relative risk of 0.60 (95% Cl 0.37-0.99) compared with those with a relatively low intake (-2.5 portions per day) (9) WHO GBD, 2009: Insufficient intake of fruit and vegetables (prevalence measure of 5 servings / day) is estimated to cause 14% of gastrointestinal cancer deaths, -11% of ischaemic heart disease deaths and 9% of stroke deaths worldwide. (8) G, fruit and vegatable intake; smokers vs non- smokers; obesity; alcohol intake; (9) G, I (PDF P.46)
Physical Activity (10) Schnohr et al., 2006 (Copenhagen): Adjusted relative risks for cancer for moderate activity 0.77: and high activity: 0.73; and for all-cause mortality, moderate: 0.78 and high: 0.75 for both sexes combined. (9) G; physical activity levels
While the majority of these studies are focused on Europe and North America,
one rural versus urban study identified in Chandigarh, India presents important lung
cancer evidence for reducing indoor air pollution exposures of women age 45 and older
that are due to the use of solid fuels for cooking. It was noted in this study that residence
in urban areas did not entail an increased risk for developing lung cancer (Gupta, 2001).
Discussion: Implications for Infrastructure-Related Health Interventions
To the best of our knowledge, knowledge gaps remain for applying HEEs in
Delhi, India specifically associated with access to urgent healthcare, overcrowding, and
urban water supply and sanitation. Cancer and environmental-health effect studies are
also primarily Europe-focused. The table below summarizes relevant HEEs for Delhi and
urban health knowledge gaps for Delhi and more generally, for Indian cities.
57


Table III.8 Study Key Findings and Identified Knowledge Gaps.
Interventions Addressing... A In Exposure Concentration Health Effect Estimate (HEE) HEE Reference (Study Location)
Outdoor Air Quality & CVD Particulate Matter (PM): 300 pg/m3 0.43% increase in CVD mortality per 10pg/m3 change in PM10 Cropper, 1997 (Delhi, India)
Outdoor Air Quality & Respiratory PM: 300 pg/m3 0.31% increase in respiratory mortality associated with a change in PM 10 level of 10pg/m3 Cropper, 1997 (Delhi, India)
Airborne Infectious Disease N/A * Sao Paolo, Brazil: a 14% increase in tuberculosis (TB) mortality with an average increase of one additional dweller in the household. & Antunes & Waldman, 2001 (Sao Paolo, Brazil)
Traffic Accidents 17% mode shift of auto & two- wheelers to rail From Auto to Rail: 9.7 deaths / 1% mode shift; From 2-wheeler to Rail: 20.6 deaths / 1% shift NCRB, 2008 (Delhi, India)
W ater and Sanitation N/A Global: having water & sanitation results in a 5% to 30% reduction in relative risk of <5 diarrheal mortality (no urban HEE identified) *
Cancer Indoor air pollution from use of coal or wood for cooking or heating Exposure of > 45 age women to IAP showed a 0.43% increase in developing lung cancer. Gupta, 2001 (Chandigarh, India)
Access to Urgent Healthcare N/A * *
&
Refers to a knowledge gap where quantitative health effect estimates are unknown for Delhi, India
This review of HEEs addresses some, but not all of the key elements in the WHO
framework model shown below: e.g. environmental and socioeconomic conditions in
shaping differential exposure, individual susceptibility, and health effects. However, this
review can help inform first-order computations of health risk reduction benefits
integrating equity concerns by exploring health effects due to inequalities in conditions
and services.
58


Figure III.2 WHO Framework Model on Social Inequalities and Environmental Risks
Taking into consideration this WHO model and the literature described in this
study, the conceptual diagram below is developed for future study of urban infrastructure,
environment and health. More specifically, the diagram demonstrates how exploring both
biophysical system characteristics (e.g. infrastructure and environmental conditions) and
human-social system characteristics (e.g. socio-economic and biological) can be useful as
both of which are known to have important implications for health outcomes. Exploring
the linkages in this diagram between infrastructure, environment-climate and health-
more specifically, between exposures to hazards (left), health effect estimates (middle),
and health outcomes (right) can be useful towards assessing key risk factors to inform
quantitative decision-making on alternative health risk reduction scenarios for cities.
59


Biophysical Interventions
Social System Interventions
Figure III.3 A preliminary schematic representation / conceptual diagram for study of the
nexus of urban infrastructures, environment and public health
In the above conceptual diagram, this review has zoomed in on the health effect
estimates (HEEs), and explored linkages to the other boxes in three ways: (1) literature
review of quantitative mortality studies of infrastructure and environmental conditions
that identify differential exposures to hazards such as air pollution, and differential
susceptibility to health effects and final mortality outcomes, (2) preliminary
harmonization of HEEs from identified studies addressing risk factors and health effects
in terms of mortality outcomes; and (3) discussion of challenges to developing HEEs
(e.g. addressing confounding factors) and identification of knowledge gaps.
The efforts in this study help develop preliminary HEEs primarily for biophysical
(e.g. infrastructure and environment-related) interventions, with a somewhat smaller
60


focus on social system interventions relevant to Delhi, India. However, by identifying
differential exposure and risks to vulnerable subpopulations as related to socioeconomic
and biological risk factors where possible, important features in the above conceptual
diagram for an improved health effects evidence-base in Asian cities. Future development
of this diagram could explore behavior changes, coping strategies, and response capacity
as social system elements. The below revised version is shown to help distinguish the
biophysical (in green boxes) and human-social (in red boxes) system characteristics.
Biophysical system characteristics (e.g. infrastructure & environmental conditions) and human-social system
characteristics (e.g. socio-economic & biological) are known to have important implications for health.
Exploring these linkages specifically between exposures to hazards, health effect estimates, and health outcomes are
considered critical for improving decision-making on infrastructure-related health interventions.
Biophysical System Interventions
Infrastructure:
e.g. water, sanitation, energy
transport, buildings
Environmental conditions:
e.g. water & air quality,
GHG emissions, natural hazards,
local climate impacts

Social System Interventions
Health Risk Computation:
Mortality: # of deaths in a given
time or place)
Morbidity: # of cases of illness in
a specified community/group
Costs: e.g., healthcare costs,
lost wages due to disability/illness,
future earnings lost by premature
death
Infrastructure & Environment-
Related Risk Categories:
Cardiovascular: Heart Attacks &
Diseases
Airborne Pollution & Respiratory
Diseases
Airborne Infectious Disease &
Infrastructure
Transport Fatalities
Diabetes
Waterborne Diseases
Cancer
Figure III.4 A revised schematic diagram for studying infrastructures, environment-
climate, and health with consideration of biophysical and socio-biological risk factors
61


Conclusion
By reviewing quantitative epidemiology studies on civil infrastructure, pollution,
and seven related health-risk categories, this study develops an initial database of
infrastructure-related health effect estimates and identifies areas where further inquiry
and field study of infrastructure and environment-related health impacts are still needed.
Identified knowledge gaps include the associations between urban mortality outcomes
and inadequate access to quality urgent healthcare services, overcrowded housing, and
water supply/sanitation. Methods for filling such knowledge gaps through
epidemiological and community-based study could be beneficial so that additional
infrastructure-related health improvement scenarios can be assessed for supporting
decision-making and prioritization of actions toward healthier Asian cities.
62


Appendix. Health Determinants of a Population and To Which Factors Health
Effect Estimates are Most Sensitive
Scientists generally recognize five health determinants of a population:
Genes and biology (e.g. sex and age)
Health behavior (e.g. alcohol use, injection drug use, unprotected sex, smoking)
Social environment or social characteristics (e.g. income, gender, discrimination)
Physical environment or total ecology (e.g. where a person lives, basic
infrastructure services, and crowding conditions)
Health services or medical care (e.g. access to quality health care and having or
not having insurance / ability to pay for care)
The figure below presents estimates of how each of these five major determinants
influence population health (Tarlov, 1999). As shown, genes / biology and health
behaviors make up 25% of population health, with the remainder of health most sensitive
to social factors including social environment, physical environment and health services.
(determinants of population health)
Figure 1. (By Tarlov, 1999)
As an example from this chapters literature review, the case of outdoor air
pollution exposure-related premature mortality in Delhi demonstrates how women and
elderly are most sensitive based on HEEs in the study by Rajarathnam et al, 2010.
63


CHAPTER IV.
UNMASKING THE ROLE OF MULTIPLE INADEQUATE BASIC
INFRASTRUCTURES IN INFLUENCING UNDER FIVE
MORTALITY IN ALL-INDIA, URBAN INDIA & DELHI
Abstract
Prior studies typically have shown infrastructure provision to have small to
modest improvements to under-five mortality rates (U5MR) (e.g., a 17% reduction in
U5MR for a relative risk of 1.2), under the assumption of independence of each
infrastructure provision from the other and without accounting for confounding effects of
wealth and literacy. This study focuses on All-India, Urban India and Delhi to explore
importance as well as interactions among social factors (wealth and literacy),
infrastructural factors, i.e., the provision of multiple basic infrastructures (piped water,
sanitation, clean cooking fuels, etc.), and proxy attributes that represent access to urgent
care health (ATH) facilities. Key findings include: 1) all response variables highly
correlated with each other requiring large datasets to control for socioeconomic status
(SES) and literacy to unmask the role of infrastructure; 2) the partial literacy subgroup is
particularly sensitive to lack of infrastructure with relative risks exceeding 4.5 for
unimproved versus improved conditions; and 3) multiple infrastructure provisions
together when controlling for confounders has larger relative risks compared to single
infrastructures. Results indicate 4.9 [95% Cl: 1.2, 19.8] and 8.6 [95% Cl: 1.2, 64.6] times
higher U5MR for All-India and Urban India, respectively, for limited literacy low SES
populations with absence versus presence of toilets / taps on premises and clean fuels.
64


Introduction
This study focuses on All-India, Urban India and Delhi to explore importance as
well as interactions among several factors influencing under-age five mortality rates
(U5MR) in India and Indian cities, including: social factors (socioeconomic status (SES)
and literacy), infrastructural factors, i.e., the provision of multiple basic infrastructures
(piped water, sanitation, clean cooking fuels, etc.), and proxy attributes that represent
access to urgent care health (ATH) facilities. Mortality is selected as the response
variable over other health indicators (e.g. morbidity) due to its policy relevance, critical
importance as a measure of societal well-being (Sen, 1998), and the availability of under
age five mortality data at city-scale. Under-Five years mortality rate (U5MR), i.e.,
deaths of children aged 5 years or less reported over the past 5 years per 1000 live births
over the same period is also an important human development indicator recognized
globally in the UN Millennium Development Goals (UN, 2013). U5MR in India is
important because India is home to 20% of the worlds and three-quarters of South Asias
under-five population, respectively (UNICEF, 2008) and has historically reported U5MR
in the bottom 50 countries, globally (World Bank, 2013). While significant progress has
been made over the past decade to reduce U5MR in India by about 25% (from 88 to 61
under-5 deaths per 1000 live births from 2000 to 2010, respectively) the current 2010
U5MR of 61/1000 remains high with one child in every 16 dying prior to age five and the
U5MRs still ten times higher than high income OECD countries (61.0 vs. 5.6) (World
Bank, 2013). Thus, understanding complex interactions that shape under-5 mortality
(U5M) is particularly important for India, and by extension the world.
Understanding urban U5MR in India is particularly important given that urban
65


U5MR is declining slower than rural U5MR (Claeson, 2000), and, urbanization is leading
to rapid growth of slums in many Indian and Asian cities (Agarwal and Taneja, 2005)
where inequalities in socioeconomic conditions, infrastructure provision and public
health services are high. More than ten cities (e.g. Mumbai, Hyderabad, Agra, Delhi)
among the more than fifty cities in India with population over one million continue to
report a high % population (>30%) living in slum areas that do not have many basic
infrastructure provisions for example, 43% of slum households in India lack access to
drinking water taps, 34% lack access to latrines and 33% are using solid fuels for cooking
(Chandramouli, 2013). Slum residents in India are among the lowest by socioeconomic
status (SES) (Gupta et al., 2009) and report significant proportions of the population
(-39%) who are illiterate or partially literate (able to read only parts of a sentence)
(National Family Health Survey, 2006). Further, one third of the worlds urban
population currently lives in infrastructure deprived slum conditions, 66% of which are in
cities of Asia and Africa where slum populations are projected to increase annually at
2.2% and 4.5%, respectively (UN-Habitat, 2010; UN-Habitat, 2006). Consequently, an
improved understanding of how under-age five mortality (U5M) is shaped by the
provision of multiple infrastructures (or lack thereof) in urban India, and the interaction
of infrastructure provision with social factors delineated in large Demographic Health
Surveys (DHS), can contribute significant insights important for improving health and
human development in other cities of the developing world.
Prior studies exploring child health in India and the South Asia region have
considered the wealth-health link (Mohanty, 2009), urban versus rural health
comparisons (Bharati et al., 2008), and the role of water and sanitation (Mollah and
66


Aramaki, 2010) in shaping U5M. In addition, prior studies of urban areas specifically
using the DHS data have considered single infrastructures such as percentage of
households having improved source of drinking water, improved sanitation, electricity
and cooking gas (Goli et al., 2011) or slum versus non-slum housing (Govt, of India,
2010); and a few have also looked at multiple infrastructure variables (Das and Gautam,
2013). However, these studies have not identified or explored the confounding effects of
socioeconomic status, literacy, and access to healthcare. Most studies have typically
assumed infrastructure provision to be an independent variable and consequently report
relative risks (RR, i.e., the ratio of U5MR in unimproved versus improved conditions) of
1.7 for urban poor versus non-poor (Agarwal and Srivasava, 2009) and up to 1.2 for
having water and sanitation versus not (Gunther and Fink, 2010). Such studies suggest
small to modest improvements to U5MR (e.g., a 17% reduction in U5MR for a RR of
1.2) can occur through providing basic infrastructures, under the assumption of
independence of each infrastructure provision from the other.
A more recent study suggests that the providing multiple infrastructures, together,
may be more beneficial having synergistic multiplier effects (e.g., (Fink et al., 2011), yet
few studies directly examine the improvements in U5MR from multiple basic
infrastructure provisions while controlling for key confounders such as socioeconomic
status (SES) and literacy. To address these gaps, the three key objectives of this paper are
as follows:
(1) Conduct factor analyses to extract key variables affecting the incidence of under-
five years mortality (U5M) as reported in India and Urban India DHS data.
67


(2) To explore associations of under-5 years mortality rates, represented by U5MR
(under age five mortality rates expressed as deaths per 1000 live births) as the
response variable with three categories of predictor variables, while controlling
for confounding interactions among:
Socioeconomic factors (wealth and literacy levels);
Proxy attributes of access to healthcare (ATH) conditions; and
Infrastructure-related factors (housing, water, toilets, and fuels for cooking);
(3) Developing relative risk (RR) computations representing the mortality rates in
unimproved versus improved condition, for various combinations of basic
infrastructure provisions (water, sanitation, and cooking fuels), while controlling
for wealth and literacy (as ATH found not to be significant). We explore
improvements arising from simply the presence and absence of various
combinations of basic infrastructures and then the quality of basic infrastructures.
Literature Review
In this study, two forms of literature review are conducted. The first approach
looks only at prior studies of under-five mortality using the Demographic Health Survey.
The second approach reviews a small set of infrastructure-environment-health studies
external to DHS that may be relevant to this study. Importantly, these studies do not
explore U5MRs associated with groups of multiple infrastructure conditions (including
housing, drinking water, toilet facility, cooking fuels), socioeconomic conditions (wealth
and literacy), proxy attributes of access to healthcare conditions (affordability, distance /
transport to healthcare facilities, facilities being open when needed) while also separating
out confounding health determinants at both city and national levels. The two types of
68


studies reviewed are described below with some providing valuable insights for the
methods in this study and ideas for future research and analyses using DHS datasets.
DHS-FocusedLiterature Review: Since October 2003, 1265 peer-reviewed
journal articles have been published including focus on 68 different countries, with
research based entirely or primarily on DHS data. For India, this includes 397 papers. Per
Table 1 below, several studies have looked at socioeconomic conditions and healthcare
utilization in India (e.g. Mohanty, 2009, Bhargava, 2011) while few have considered
inadequate civil infrastructure provision, access to healthcare, and under-five mortality.
The most pertinent studies to this paper include Agarwal (2009) which explores
tuberculosis outcomes with overcrowded sleeping conditions, Das et al (2012) which
considers household living conditions such as housing, toilet facility, drinking water,
cooking fuels and its effect on child morbidity (including diarrhea, fever, and cough) in
Madhya Pradesh, and the Ministry of Health (2009) study that considers infrastructure
conditions associated with socioeconomic status, separately the overall U5MRs across
eight cities, and nutritional status associated with slum versus non-slum areas. Goli et al.
2011 explores disease conditions, immunization coverage, and separately the deficits in
basic amenities within selected Indian cities, but does not assess under-five mortality
associated with infrastructure and access to healthcare conditions (e.g. health facility
open when needed, affordability, distance / transportation) within the cities, or across All-
India. The Gunther et al. study also aims to comprehensively address water and
sanitation across countries using DHS and they observe the highest positive effect of such
infrastructure on childrens health in urban areas. The variables Gunther et al. control for
include mothers age, childs order of birth, and ownership of durable goods as a proxy of
69


households long-term income. However, they do not control for levels of literacy.
One additional noteworthy infrastructure and health quantitative analysis using
DHS India data was a recent review of tuberculosis (TB) outcomes relative to living
space density. Agarwal (2013) finds almost 50% of population in the poorest wealth
quartile are living in houses where 5 persons or more share one sleeping room, with TB
incidence up to twice as high in homes with 5 persons or more per seeping room versus
only 4 persons per sleeping room (Incidence of 423 versus 268 TB cases per 100,000
population). Similarly, TB incidence for those homes without separate cooking space can
be 2.2 times higher than those that have separate cooking space (494 versus 223 TB cases
per 100,000 population). An extension of this prior DHS India re-analyses is shown in
our results focusing on diarrheal incidence associated with members per household and
water and sanitation conditions, to compare with findings in studies by Gunther and Fink.
To the best of our knowledge, no article in this first approach for literature review
has looked at under-five mortality associated with multiples infrastructures (including
housing, drinking water, toilet facility, cooking fuels), socioeconomic conditions
(including wealth and literacy), proxy attributes of access to healthcare and under-five
mortality rates while separating out these confounding health determinants.
Reviewing a Small Set of Relevant Literature External to the Use of DHS:
Examples exist from the WHO Department of Public Health and the Environment that
aim to quantify global disease burden (e.g. Ezzati et al., 2008) and national disease
burden (e.g. Pruss-Ustun et al., 2008) that can be prevented by environmental
interventions. Pruss-Ustun (2007) also identifies many environmental intervention areas
for health, some related to the built environment e.g. housing (effects on respiratory
70


infections) and road design (on road traffic injuries). However, quantitative analyses
linking these multiple infrastructure factors and mortality in specific cities are in most
cases not presented. Other literature external to the DHS has addressed the health effects
of water and sanitation. To name three of many studies, Van Poppel et al. 1997 identifies
several risk factors and confounders specific to child mortality in a Dutch municipality,
Mollah, 2006 reviews water, sanitation and hygiene literature primarily focusing on rural
studies, and Boschi-Pinto, 2008 conducts cross-country reviews addressing child
mortality due to diarhhea.
Based on this review, the contributions of this study are to explore U5M
associated with wealth and literacy conditions, multiple inadequate household
infrastructures and access to healthcare in All-India, Urban India and Delhi, India.
Methods
Analyses for All-India (n=51,555 households), Urban India (n=19,483
households), and Delhi (n=l,150 households) DHS 2005/6 survey data are conducted to
explore the extent to which current U5M outcomes are shaped by multiple infrastructures
while controlling for key confounding social factors including SES and literacy. As
background, the DHS is a national survey administered every five to six years to collect
accurate, nationally representative data on health and population in India; and is used
here for its comprehensive nature allowing for comparisons of health, infrastructure,
socioeconomic conditions, and in that the DHS is required and done frequently.
DHS Survey Data: Figure 1 presents a summary of data availability and some
initial DHS variables of interest as preliminary aggregated data analyses conducted for
Delhi. While these initial explorations of slum versus non-slum areas with a breakdown
71


of presence of basic conditions and under-five / infant mortality rates help depict the
potential of comparisons between such areas, these analyses do not isolate impact of
infrastructure inadequacy from other factors such as literacy and wealth that also affects
health outcomes.
Strengths and Weaknesses of Study Data Analyses: A key data limitation is that
only child mortality data are explored as full population mortality data was unavailable.
While additional analyses of mortality versus access to services in Delhi slum versus
non-slum areas are possible using DHS data (as shown in Figure 1), they are not in the
scope of this study. For example, Figure 1 depicts for slum and non-slum households a
10% variation in terms of access to health insurance and a 28% variation in terms of
having a bank account or formal education, but both health insurance and bank services
data is not further explored in terms of under-five or infant mortality.
Methods: Five steps were taken to reorganize and analyze the DHS data:
1. Data Re-organization: the categorical data were recoded to consistently represent
worst to best conditions (lowest to highest number), and cleaned for non-responses;
2. Identification of Predictor Variables and their interactions: The response variable was
identified as U5MR per 1000 births in past five years computed as Equation 1 below:
U5MR VncidenceofUnder5Deaths_pastfiveyears) ^qqq
(.IncidenceofLiveBirths pastfiveyears)
Equation IV. 1 Computing Under-Five Mortality Rates (U5MR)
The three categories of predictor variables selected for analyses include:
o Socioeconomic conditions (wealth and literacy). First, correlations with these
socioeconomic conditions are explored after which they are controlled for;
72


o Multiple infrastructures (housing, water, toilets, and cooking fuels); and the
o Proxy attribute of access to healthcare (ATH) including costs, transportation
challenges to reach healthcare, and households responding health facility is
not open when needed as indicated by the mother as reasons they did or did
not deliver at a health facility.
Pearson Correlations are then used to identify interactions between predictor variables
indicative of multiple collinearity, and between predictor and response variables.
3. Exploratory Factor Analysis and Principal Component Analysis: were then used to
identify variables of interest, their importance, and to reduce the number of predictor
variables for further computation of relative risk.
4. Relative Risk (RR) was computed using Equation 2:
U 5MR(PopulationWithUnimprovedConditions)
RR =----------------------------------------------------
U5MR(CorrespondingBestInfrastructureCondition)
Equation IV.2 Computing Relative Risk of Under-Five Mortality Rates
U5MR were computed for unimproved versus improved infrastructure conditions
using two approaches. First, we consider only presence (versus absence) of combinations
of three infrastructures water supply on premises, toilet facility on premises, and clean
cooking fuels and categorize in the form of Good: having all three improved conditions
vs. Medium: having two of the three improved conditions vs. Poor: having one or less
of the three improved conditions to explore how just providing multiple infrastructures
together reduces U5MRs and relative risk, without addressing the quality of the
infrastructure provided. Second, we consider aggregate measures that include presence /
absence of basic infrastructure provisions including house type, water supply, toilet
facility type, and cooking fuels along with their quality when present (e.g., having pucca
73


(solid physical structure) home coded as high quality vs. semi-pucca vs. kaccha worse
quality; having taps on-site coded as high quality vs. up to 30 minutes away vs. more
than 30 minutes away worse quality; having flush toilet as high quality vs. pit toilet
vs. no toilet facility; and having electricity, LPG, biogas for cooking fuel as high
quality vs. kerosene vs. solid fuels including wood, charcoal, agricultural crop, animal
dung as worse quality). Together this yields three categories as Bad (25% or less of
high quality infrastructure condition) vs. Average vs. Good (i.e. >75% high quality
infrastructure condition).
Because the U5MR are found to be numerically relatively small, although still of
public health concern i.e., of the order of 2% to 5% (e.g. 20 to 50 deaths per 1000
births), the probability of deaths are relatively low, and therefore the odds ratio
(probability of death / probability of surviving = 2% / 98%) is numerically similar to
U5MR. The logio of U5MR will therefore be similar to the LOGIT model for child
survival to age five. Thus, although only U5MR is shown in this paper, similar
relationships are expected for child survival as well.
5. Comparing relative risks: The RR were computed while controlling for highly
correlated socioeconomic factors (multi-collinearity) and for the infrastructure
combinations previously listed in Step 4 and these were compared with the
conventional case where RR are typically computed assuming each single
infrastructure was independent of the other and of the socioeconomic variables.
It is important to note that other confounding risk factors may also play a role in
shaping under-five mortality such as the birth order of children, as indicated in studies
74


such as in Van Poppel et al., 1997 and suggested in Gunther and Fink, 2010. Such data
were not recorded in the India DHS and hence could not be studied.
Results
Exploring Variable Interactions: Correlations in the All-India Data: Pearson
correlations of Under-5 years of age mortality incidence (U5MI) and the nine predictor
variables, as well as the U5MR (mortality per 1000 live births) versus the percentage
occurrence of the same nine predictor variables were assessed as shown in Table 2 and
Table 3 for the All-India dataset, respectively. In these analysies, a forward causal model
with respect to U5MI and U5MR and the various predictor variables is expected.
Table 2 tabulates correlation coefficients across U5MI and 9 predictor variables
for -51,556 households; note that for n > 1000 pairs of data, any r > 0.05 can be
considered statistically significant (see Clark, 2009). As is shown in Table 2, all pairs of
predictor variables are significantly correlated with each other (in many cases with r>
0.58), which indicates the predictor variables are highly correlated, and hence not
independent. For example, wealth of households (low to high SES) was highly correlated
with incidences of literacy (r=0.58), water supply (r=0.60), toilet facility (r=0.74) and
clean cooking fuel (r=0.62). The incidence of under-five mortality (U5MI; whether a
death occurred or not) was observed to be weakly correlated with wealth (r = 0.07),
literacy (0.07); incidences of water supply (0.06), toilet facility on premises (r=0.07),
quality cooking fuel (r=0.07); and ATH (r=0.06) reflected by the facility being open as
noted in the maternal survey. Correlation of U5MI with house type, shared toilet, and
access to healthcare attributes of costs and too far / no transport were not statistically
significant with r less than |0.05|. The same is seen in Table 3, where the data
75


representing incidences are aggregated by the five wealth quintiles, yielding n=5. In
Table 2, mortality incidences are aggregated within each of the five wealth quintiles and
divided by the number of live births within each quintile to yield the response variable -
U5MRs computed for each wealth quintile. Likewise the % occurrence of the other
predictor variables (best condition) is also computed for the homes in the 5 wealth
quintiles, as shown in Table 3. The percentage occurrence of all predictor variables (best
condition) are shown to be highly correlated with wealth, and with each other, (any r >
0.67 representing significance for n = 5 (see (Clark, 2009)). The mortality rates likewise
are highly correlated with wealth and pairwise with the % occurrence of each of the
predictor variables.
Both Table 2 and Table 3 show that the predictor variables are nat_ independent.
Similar results were seen for the Delhi data (shown in Appendix). This means generalized
linear models or other regression models assuming independent predictor variables
should not be employed. Tables 2 and 3 indicate that the relationship of the response
variable U5MR should be explored with a smaller sub set of key predictor variables,
identified in principal component analysis (PCA) described next, and by controlling for
variable interactions to compute relative risk for different conditions of interest.
Principal Component Analysis /Exploratory Factor Analysis conducted with the
nine predictor variables revealed that five key variables including water supply, toilet
facility, wealth, literacy, and cooking fuels explain 75% of the total variance. Using
Kaisers criteria [21; 22], eigenvalues over 1 are considered stable (wealth, water and
sanitation) and component 4 (literacy) and 5 (cooking fuels) are just under 1, so may also
be significant. Thus relationships of U5MR were explored with these predictor variables.
76


In addition, we also explored the relationship of U5MR with ATH represented by facility
not being open when needed (as reported in the maternal survey), which was observed to
not be significant in Table 4.
Computing RR and Controlling for Confounding Variables: We first demonstrate
in analyses that infrastructure deficiences are common in All Indian cities, and improve
with SES as shown in Figure 2 (along with a summary of the datasets used from within
the DHS child recode dataset). We then explored RR of U5MRs for confounders, SES
and then literacy conditions as shown in Figure 3. By considering U5MR correlations
with these conditions for India, Urban India, and Delhi, analyses show a relative risk of
~2 for both wealth and literacy from lowest to highest (i.e. each of these figures show that
U5MR ranges from about 40 to 20 deaths per 1000 total births for worst to best
conditions, respectively). Therefore, these factors are controlled for in the remaining
analyses.
Controlling for wealth and literacy to explore the impact of infrastructure
provision is illustrated for All-India and Urban India in Figure 4 (with Delhi in the
Appendix). Access to healthcare conditions represented by facility not being open at time
of delivery in the maternal survey is also explored as a potentially important factor. This
factor is identified as a proxy attribute for access to healthcare conditions, with
circumstances relating to facility not being open for mothers during delivery also
potentially preventing children from accessing health care during critical health events.
See Table 4. Analyses show this variable to be insignificant with a relative risk of 1.0 or
1.1 for low SES groups in All-India and Urban India. A larger relative risk exceeding 2.5
for high SES groups in All-India, Urban India, and Delhi; and a relative risk of 10 was
77


found for the Delhi high SES population; but this was not a consistent trend across
population groups as was the case for wealth and literacy. For the Delhi low SES
population, the ATH relative risk could not be evaluated due to the small n of responses
for this attribute in the Delhi dataset.
Following these analyses, we then moved to infrastructure conditions focusing on
presence and absence as well as provision and quality of infrastructures. Figure 5 and
Table 5 shows the high importance of providing multiple high quality infrastructure
provisions in India and Urban India, which becomes visible specifically among the
partially literate population in the lowest socioeconomic status group (Presence vs.
Absence: India RR= 4.9; 95% Cl: 1.2, 19.8] and Urban India RR = 8.6 [95% Cl: 1.2,
64.6]; Provision and Quality: India RR=5.6; 95% Cl: 2.2, 14.2 and Urban India RR=14.5;
95% Cl: 2.9, 70.9, respectively). For the same low SES subgroup and fully literate and
illiterate population, the provision of these multiple infrastructures makes much less
difference (RR = 1 to 1.2, shown in Table 5) likely because fully literate populations may
be able to make other adjustments in face of poor infrastructure, e.g., handwashing
benefits can be realized even without toilet facilities. In contrast, among illiterate
populations, the lack of awareness of prevention strategies and or other best practice
medical recommendation (e.g., oral rehydration) could mask the benefits of infrastructure
provision. However, these results highlight a key insight that the sensitivity of health
outcomes (U5MR) to infrastructure provision can be masked by wealth and literacy.
Among the poorer households with limited or partial literacy (i.e. respondents able to
read and write only parts of a sentence), the significant benefits of providing multiple
infrastructure provision becomes apparent as indicated by the high RR, indicating U5MR
78


to be roughly >500% worse with unimproved infrastructure than with improvements, i.e.,
provision of infrastructure can reduce U5MR by > 80%.
To verify these findings, analyses were repeated for the All-India data set
considering high SES and various literacy conditions and similar results were obtained.
The analyses were repeated with only neonatal deaths, infant death and U5MR-ND (less
neonatal deaths; at least 30 to 50% of neonatal deaths in developing countries have been
shown to be from infections with the most common causes being diarrhea, pneumonia,
tetanus and sepsis ((Moss et al., 2002), (Stoll, 1997), and Bang et al., 1999), which can be
related to infrastructure issues. The infant, neonatal, and U5M split for the conditions
explored are also shown as Figure 7. The finding of high RR for the lack versus the
provision of infrastructure are seen consistently in partial literacy, low SES conditions
(for NMR, IMR, U5MR and U5MR-ND), and are absent in the high SES group. All these
results are shown in Table 6.
We also repeated U5MR analysis for Delhi and found similar results as shown in
Appendix. The paucity of data made the analysis more challenging, as it was not possible
to compare with the best case (which had zero incidences of U5M). However, we still see
two times higher U5MR (RR=2) comparing the worst case with the average case when
controlling for ATH, wealth, and literacy.
The U5MR-ND computations for All-India in Table 6 also show that high RR
exists for certain combinations e.g. across all literacy conditions, with partial literacy
remaining the most sensitive. Such results suggest U5MR-ND as another useful indicator
for assessing infrastructure combinations, as removing neonatal deaths excludes factors
that may be unrelated to infrastructure that can shape neonatal mortality (e.g. preterm
79


birth and asphyxia which accounted for -50% of neonatal deaths in 2000 in a global
study of 193 countries (Lawn et al. 2006).
The relative risk computations in Figure 3 also demonstrate why controlling for
factors like literacy when possible, in addition to wealth, is important for exploring
U5MR associations with inadequate infrastructure condition combinations (e.g. of
housing, water, toilets, and fuels). As shown, the importance of basic infrastructure is
masked by the importance of both wealth and literacy (which is closely related to SES).
Plots are presented in Figure 4 to bring together some of the analyses conducted
and highlight U5MR sensitivity for the partially literate population to infrastructure
provision and condition. In this analysis, having three or more out of four high quality
infrastructure provisions is shown as Good: >75% High Quality Conditions, while one
or less of high quality infrastructures is shown as Bad: < 25% Low Quality Conditions.
To identify which infrastructure provision had the largest impact on U5MR within
the partial literacy group, we repeated the analysis considering only the presence and
absence of various combinations of key infrastructures identified as important in the
EFA, which are water supply on premises, toilet on premises, and clean fuels for cooking.
These results are shown in Figure 4 for India and Urban India (with Delhi in Appendix).
For both All-India and Delhi datasets shown in Figure 6a, all infrastructure
combinations when controlling for confounding variables of SES and literacy -
demonstrate significantly higher relative risks then previously reported in the literature.
For example, U5MR increases by a factor of four or more (400% increase) for all
infrastructure combinations. For the All-India poor and partially illiterate households,
80


Figure 6 indicates the exceptionally high importance of having toilets and taps together
(RR=7.0); as well as having toilets, taps, and clean fuels together (RR=5.9).
Similar results are seen in the Delhi datasets too with the combination of toilets,
taps, and fuels resulting in a RR of 12.4 -after controlling for confounding social
variables where possible (a focus on full literacy households with all wealth groups
included due to small n). In contrast, when not controlling for any of these variables, RR
results are in the range of 1.3 to 1.7, for each individual infrastructure condition shown in
Figure 6b, in line with current observations in the literature. These results indicate
controlling for multi-collinearity unmasks the critical importance of providing multiple
basic infrastructures demonstrated by examining the low SES, partially literate subgroup.
Discussion
Results from this analyses indicate over a 4 times higher U5MR for those low
SES partial literate households without multiple high quality basic infrastructure
provisions, and roughly a 2 times higher U5MR for those households of lower
socioeconomic status and literacy levels. In addition, preliminary exploration indicate up
to a 75.9% reduction in under-five mortality rates for All-India may be achievable in low
socioeconomic status households with limited literacy for those having versus lacking
basic infrastructure provisions.
With United Nations Millenium Development Goal # 4 aiming to reduce by two
thirds the under-five mortality (between 1990 and 2015), strategies focused on multiple
infrastructure and access to healthcare interventions can (presumably) be more successful
then efforts focusing on improving socioeconomic conditions alone. Future DHS surveys
81


could help shed additional light on these associations by exploring the possibility of
larger survey sample sizes for Delhi and other cities.
In using DHS to quantify the benefits of multiple infrastructures, this analysis
presents one of the first to quantitatively separate out access to healthcare, literacy and
socioeconomic factors from multiple infrastructure conditions and to link a variety of
socio-economic and infrastructural determinants of health specifically for U5M. We
emphasize that additional exploration of infrastructure-health associations and causal
relationships through field work and epidemiological study is needed.
The results of this study provide an initial quantitative exploration opening up a
new line of inquiry on urban health improvements specifically of under-five mortality
through changes in infrastructure conditions and socio-economic conditions. Controlling
for confounding factors are important to improving understanding of how multiple
infrastructure provisions shape child mortality.
82


Full Text

PAGE 1

EXPLORING THE NEXUS OF INFRASTRUCTURES, ENVIRONMENT, AND HEALTH IN INDIAN CITIES: INTEGRATING MULTIPLE INFRASTRUCTURES AND SOCIAL FACTORS WITH HEALTH RISKS By JOSHUA B. SPERLING B.S., Civil Engineering, University of Colorado Boulder 2007 M.Eng., Envir onmental and Sustainability Engineering, Univ. of Colorado Denver, 2010 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Philosophy Civil Engineer ing 2014

PAGE 2

ii This thesis for the Doctor of Philosophy degree by Joshua B. Sperling has been approved for the Department of Civil Engineering by Anu Ramaswami, Advisor JoAnn Silverstein, Chair Wes Marshall Debbi Main Siddharth Agarwal May 2, 2014

PAGE 3

iii Sperling, Joshua B. (Ph.D., Civil Engineering) Exploring the Nexus of Infrastructures, Environment, and Health in Indian Cities: Integrating Multiple Infrastructures and Social Factors with Health Risks Thesis directed by Professor Anu Ramaswami. ABSTRA CT The overarching goal of this thesis is to explore and assess infrastructure environment health interactions in Indian cities, addressing social factors such as wealth and literacy, as well as the provision of multiple infrastructures. Five main studie s are conducted. First, exploration of Delhi all cause mortality data and survey of local experts on associations between infrastructures environment, and health outcomes Key findings include: a) that 50% of deaths in Delhi are reported with cause not cl assified (demonstrating the need for bottom up study to supplement hospital data) and b) that ~ 19% of classified deaths by cause in Delhi, India could be related to infrastructure or infrastructure related environmental factors Second, review of epidemio logy studies relating health outcomes to infrastructure and pollution exposure in Indian and Asian cities is conducted to help identify initial evidence and gaps for infrastructure related health effects and quantification of differential risk based on soc ial factors (e.g. low socioeconomic status (SES)). Third, top down analyses using national survey of under age five mortality rates (U5MR) by multiple infrastructure conditions are studied while addressing confounding social factors A key finding is that the relative risk for under five mortality rates are 860% higher in Urban India for those lacking multiple basic infrastructure provisions relative to improved conditions for low SES condition and limited literacy households.

PAGE 4

iv These analyses demonstrate se nsitivity of limited literacy households and importance of considering multiple infrastructures together over single infrastructure improvements. Fourth, bottom up comparative c ommunity study helps characterize infrastructure, environment, extreme weather conditions and local sustainability priorities. A key finding was that households deprived of infrastructure provisions would prioritize that first over pollution or extreme weather conditions. In addition, both low SES communities studied were different i n their coverage of all infrastructures except cooking fuels. In the high SES area, infrastructure conditions were ranked as a highest priority (e.g. drainage) with pollution and climate related extreme weather events still higher priorities than low SES a reas, which selected water supply, parks and open space, and drainage as highest priorities. Multiple dimensions of access to healthcare conditions in the same neighborhoods were explored next with findings indicating the two low SES areas to have similar travel costs to reach care and different abilities to pay for care. The high SES area also had higher accessibility to care yet with quality of care less acceptable relative to low SES areas that had issues with wait times, affordability, and access sugge sting future study should address such factors and effects on health outcomes. Finally, data availability, needs, and challenges are explored for computing health benefits of multiple infrastructure interventions, while also identifying preliminary interv ention scenarios and who may benefit more or less by age, gender, and SES These efforts offer a preliminary approach to helping prioritize future decision making in Asian cities by demonstrating initial methods that can be useful for modeling risks and i nteractions between infrastructure provisions, environment, and health. The form and content of this abstract are approved. I recommend its publication. Approved: Anu Ramaswami

PAGE 5

v To my wife, Ari el family, friends and mentors who m h ave all given me a lifelong appreciation for learning, exploring and ma king a difference in this world.

PAGE 6

vi ACKNOWLEDGMENTS I want to acknowledge my advisor, Prof. Anu Ramaswami and committee members including Prof. Debbi Main, Prof. JoAnn Silverste in, Prof. Wes Marshall, and Dr. Siddharth Agarwal for their insightful guidance, encouragement, and support in building an interdisciplinary and international professional network. I especially want to thank Dr. Ramaswami, my primary advisor, for offering extensive professional training, mentorship, and insightful supervision throughout my doctoral study. I'm very appreciative for all her time, contribution of ideas, and mentorship. I would also like to thank the NSF IGERT Sustainable Urban Infrastructure program at University of Colorado Denver for supporting this work, providing initial funding and the opportunity to conduct international research with the local assistance of Dr. Agarwal, Dr. Tapan Kalita, Dr. Ajay Nagpure and Navneet Baidwan for communit y based fieldwork in India. This research was also supported and influenced by the United States India Fulbright Nehru fellowship and Indo US Science and Technology Forum: Research in Science and Engineering Fellowship in India (thanks to my research hos ts, Dr. Agarwal at Urban Health Resource Centre and Dr. Singhal at TERI University Department of Policy Studies, Delhi, India), the NSF Research Coordination Network on Sustainable Cities and NSF Partnerships for International Research and Education. Each opportunity has provided a unique platform and network for building on acquired knowledge and methods for continued learning and study, locally and globally. The UC Denver IGERT program faculty, coordinators, students as well as colleagues, family, and fri ends in diverse localities also enriched this journey. I hope to build on these various professional and personal relationships for a long time to come.

PAGE 7

vii TABLE OF CONTENTS CHAPTER I. INTRODUCTION ................................ ................................ ................................ ........ 1 Re search Objective ................................ ................................ ................................ 1 Rationale ................................ ................................ ................................ ................. 1 Key Question, Unique Features of Indian and Asian Cities, and Contributions .... 5 Dissertation Organization ................................ ................................ ....................... 8 Key Contributions: Linking Multiple Infrastructure Sectors, Inadequate Infrastructures, and Urban Health ................................ ................................ ......... 10 II. EXPLORING HEALTH OUTCOMES AS A MOTIVATOR FOR LOW CARBON CITY DEVELOPMENT: IMPLICATIONS FOR INFRASTRUCTURE INTERVE NTIONS IN ASIAN CITIES ................................ ................................ ........... 12 Abstract ................................ ................................ ................................ ................. 12 Introduction ................................ ................................ ................................ ........... 13 Literature Review ................................ ................................ ................................ .. 15 Methods ................................ ................................ ................................ ................. 22 Results ................................ ................................ ................................ ................... 26 Discussion: Implications for Infrastructure Interventions ................................ .... 31 Conclusion ................................ ................................ ................................ ............ 34 Acknowledgments ................................ ................................ ................................ 35 Appendix. Chief registrar for NCT of Delhi Re ana lysis of births and deaths 2008 report for Delhi. ................................ ................................ ........................... 36 III. A REVIEW OF EPIDEMIOLOGICAL STUDIES RELATING TO INFRASTRUCTURE AND POLLUTION IN ASIAN CITIES: EVIDENCE AND GAPS FOR INFRASTRUCTURE RELATED HEALTH EFFECT ESTIMATES IN DE LHI 35 Abstract ................................ ................................ ................................ ................. 35

PAGE 8

viii Introduction ................................ ................................ ................................ ........... 36 Methods ................................ ................................ ................................ ................. 37 Literature Review ................................ ................................ ................................ .. 38 Results: Risk Factors and Health Effect Estimates ................................ ............... 38 Discussion: Implications for Infrastructure Related Health Interventions ........... 57 Conclusion ................................ ................................ ................................ ............ 62 Appendix. Health Determinants of a Population and To Which Fac tors Health Effect Estimates are Most Sensitive ................................ ................................ ..... 63 IV. UNMASKING THE ROLE OF MULTIPLE INADEQUATE BASIC INFRASTRUCTURES IN INFLUENCING UNDER FIVE MORTALITY IN ALL INDIA, URBAN INDIA & DELHI ................................ ................................ .................. 64 Intr oduction ................................ ................................ ................................ ........... 65 Literature Review ................................ ................................ ................................ .. 68 Methods ................................ ................................ ................................ ................. 71 Results ................................ ................................ ................................ ................... 75 Discussion ................................ ................................ ................................ ............. 81 Appendix A. Delhi Analyses and All In dia Principal Component Analysis ........ 94 Appendix B. All India DHS Analyses: Progressions by Wealth Quintiles .......... 96 Appendix C. Diarrheal Incidence Analyses ................................ .......................... 97 Appendix D. Ad ditional DHS Re Analyses for Six Indian Cities, DHS Urban vs. Rural India, and Delhi vs. All India ................................ ................................ ...... 98 V. INFRASTRUCTURE CONDITIONS AND SUSTAINABILITY PRIORITIES IN ASIAN CITIES: A COMPARATIVE STUDY OF THREE NEIGHBORHOODS IN DE LHI, INDIA ................................ ................................ ................................ ............... 101 Abstract ................................ ................................ ................................ ............... 101 Introduction: Why Asian Cities, Civil Infrastructures & Local Priorities .......... 102 Objectives ................................ ................................ ................................ ........... 105 Review of Survey s on Global to Local Priorities ................................ ............... 106

PAGE 9

ix Methods ................................ ................................ ................................ ............... 108 Results and Discussion ................................ ................................ ....................... 114 Conclusions ................................ ................................ ................................ ......... 122 Appendix A. Delhi District level Infrastructu re Condition (Census India, 2011) ................................ ................................ ................................ ............................. 123 Appendix B. Preliminary Characterization using Survey, Transects, and a Preliminary Descriptive Matrix of Physical Infrastructure and Social Infrastructure ................................ ................................ ................................ ....... 125 Appendix C. Prioritization Visual Used for Survey ................................ ........... 127 Appendix D. Survey Visuals: Levels of Satisfaction and Inconvenience .......... 128 Appendix E. Assessing Levels of Satistfaction and Inconvenience Exp erienced ................................ ................................ ................................ ............................. 129 Appendix F Initial Analyses Discussion by Individual Infrastructures ............. 130 Appendix G. Household Survey Instrument Used ................................ .............. 147 VI. A FOCUS ON EXPERIENCES WIT H AND BARRIERS TO ACCESS TO URGENT HEALTHCARE IN THREE NEIGHBORHOODS OF DELHI INDIA ...... 174 Abstract ................................ ................................ ................................ ............... 174 Introduction ................................ ................................ ................................ ......... 175 Rationale ................................ ................................ ................................ ............. 176 Lit erature Review and Defining Access ................................ ............................. 177 Methods ................................ ................................ ................................ ............... 180 Results ................................ ................................ ................................ ................. 183 Discussion ................................ ................................ ................................ ........... 191 Conclusions ................................ ................................ ................................ ......... 191 Appendix A. Supplementary Analyses ................................ ............................... 192 Appendix B. Consent Form and Household Survey Instrument Used ............... 200

PAGE 10

x VII. WHAT IS KNOWN AND WHAT IS NEEDED TO ESTIMATE HEALTH BENEFITS OF INFRASTRUCTURE INTERVENTIONS: CASE STUDY OF DELHI, INDIA ................................ ................................ ................................ ............................. 209 Abstract ................................ ................................ ................................ ............... 209 Introduction ................................ ................................ ................................ ......... 210 Literature Review ................................ ................................ ................................ 211 Results of First Order Computation ................................ ................................ .... 217 Discussion: Synergies in Sustainable Infrastructure Development and Urban Health ................................ ................................ ................................ .................. 223 Conclusion ................................ ................................ ................................ .......... 224 VIII. SUMMARY & CONCLUSIONS ................................ ................................ ....... 227 REFERENCES ................................ ................................ ................................ ............... 232 Chapter 1 ................................ ................................ ................................ ..... 232 2 Chapter 2 ................................ ................................ ................................ ....... 233 Chapter 3 ................................ ................................ ................................ ....... 239 Chapter 4 ................................ ................................ ................................ ....... 253 Chapter 5 ................................ ................................ ................................ ....... 257 Chapter 6 ................................ ................................ ................................ ....... 260 Chapter 7 ................................ ................................ ................................ ....... 262

PAGE 11

xi LIST OF TABLES Table II.1 Review of Health Benefits and Risks of Infrastructures. ................................ .......... 16 II.2 Tracing Health Outcomes to Infr astructure and Environment Related Factors. ....... 19 II.3 Short Term (24 hour mean) Environmental Exposure Guidelines and Delhi Compliance. ................................ ................................ ................................ ...................... 21 II.4 Comparative Health Indicators for Delhi, Mumbai, and India. ................................ 27 II.5 Summary of Local Expert Opinion on 2008 Deaths. ................................ ................ 32 III.1 Available Health Effect Estimates for CVD and Total Mortality Risk. ................... 41 III.2 Available Healt h Effect Estimates for Respiratory Mortality. ................................ .. 45 III.3 Available Health Effect Estimates for Airborne Infectious Disease Mortality. ....... 48 III.4 Available Health Effect Estimates for Road Safety. ................................ ................. 49 III.5 Available Health Effect Estimates for Diabetes. ................................ ...................... 50 III.6 Available Health Effect Estimates for Waterborne Disease Pathogens .................... 53 III.7 Available Health Effect Esti mates for Cancer. ................................ ......................... 56 III.8 Study Key Findings and Identified Knowledge Gaps. ................................ .............. 58 IV.1 Review of Studies In Past Ten Years Using the DHS: 2002 2012 .......................... 84 IV.2 Exploring C orrelations: All India Pearson's correlation (r) Matrix With U5M Incidence. ................................ ................................ ................................ .......................... 86 IV.3 Correlation Matrix: Aggregate Analyses where U5MRs are computed for Wealth Quintiles (n = 5) and Plotted Against Changes in Other Variables Du e to Wealth .......... 87 IV.4 Summary of U5MRs and RRs for a Proxy Attribute of Access to Healthcare ......... 87 IV.5 Summary of All India U5MR, IMR, NMR, U5MR ND Analyses for Presence and Absence of Infrastructure Combinations ................................ ................................ .......... 93 V.1 Review of Small Sample of Example Surveys on Global and Local Priorities. ..... 108

PAGE 12

xii V.2 Example Analyses on Time to Fetch Water and Motor Vehicle Ownership ........... 116 V.3 Summary of NM & BJ T Test Results for Key Infrastructure Conditions .............. 118 VI.1 Review of Studies Characterizing Access to Healthcare and Providing Quantitative Findings in Terms of Experiences with Access to Health care. ................................ ...... 179 V.2 Access to Healthcare: Costs to Reach Care and Affordability of Urgent Care ....... 188 VI.3 Comparing Qualitative Responses on Knowledge and Experiences With Health Facilities along with Len gth of Residence ................................ ................................ ...... 189 VII.1 What We Know and What is Needed to Explore Health Benefits of Infrastructure for the Case of Delhi, India ................................ ................................ ............................. 212 VII.2 Literature Review Integrating Multiple Infrastructure Related Health Risk Categories and Health Effect Estimates ................................ ................................ .......... 213 VII.3 Summary of more health effect estimates useful for intervention scenarios ........ 216

PAGE 13

xiii LIST OF FIGURES Figure I.1 Unique Features of Indian Cities and the Case of Delhi, India ................................ ..... 6 I.2 WHO Framework Model on Social Inequalities and Environmental Risks ................. 8 II.1 Mortality by Age: Delhi (2008) ................................ ................................ ................ 27 II.2 Mortality by Cause (if classified) in the NCT of Delhi (2008) ................................ .. 28 II.3 Estimating Air Quality Related Cardiovascular & Respiratory Mortality Reduction in Delhi ................................ ................................ ................................ ................................ .. 30 III.1 E xample Relative Risk Findings for Infant and Child Mortality .............................. 52 III.2 WHO Framework Model on Social Inequalities and Environmental Risks ............. 59 III.3 A preliminary schematic representation / conceptua l diagram for study of the nexus of urban infrastructures, environment and public health ................................ .................. 60 III.4 A revised schematic diagram for studying infrastructures, environment climate, and health with consideration of biophysical and socio biological risk factors ...................... 61 IV.1 Initial Exploration of Infrastructures, Health/Banking/Education, and Infant / <5 Mortality Per 1000 births for Slum vs. Non Slum (DHS/NFHS 3) ................................ 83 IV.2 Infrast ructure deficiencies: common in all Indian cities, improve with SES ........... 84 IV.3 Example U5MRs by Wealth and Literacy from Lowest to Highest (India & Urban India) ................................ ................................ ................................ ................................ 88 IV.4 Flow Charts for India and Urba n India: Controlling for Wealth and Literacy ......... 89 IV.5 India and Urban India Multiple Infrastructure Presence / Absence vs. U5MR ....... 90 IV.6a RR of Infrastructure Condition Combinations: India (top) v. Delhi (bottom) ........ 91 IV.7 Breakdown of All India Neonatal, Infant, and Remaining Under 5 Mortality ........ 92 V.1 % of Births as Institutional Deliveries by Delhi Districts ................................ ........ 110 V.2 Northeast (NE) Delhi Study Neighborhoods on East side of Yamuna .................... 111

PAGE 14

xiv V.3 Closest Nearby Air Quality Station to NE District (West of Yamuna River) ......... 112 VI.4 Socio Economic Factors By Study Area: Avg. Monthly Expenditures ................. 113 V.5 Study Neighborhoods Infrastructure Conditions Snapshot ................................ ... 113 V.6 Summary of Results in Study Areas Compared with NFHS 3, 2006 Indicators ..... 11 7 V.7 Summary of Results from the Three Study Areas ................................ ................... 117 V.8 Percent of Total Votes for Top Three Priorities Among Infrastructures, Environment, and Climate Related Extreme Weather Events by Study Area ................................ ...... 121 VI.1 Geographic Distribution of % Slum population to Total Population (Census, 2001) and Births as Institutional Deliveries by Delhi Districts (DLHS, 2008) ........................ 177 VI.2 Physicians Per 100,000 population by Country and Life Expectancy vs. Physicians Per Capita by Country (UC Atlas of Global Inequity and WHO) ................................ .. 180 VI.3 % of Births as Institutional Deliveries by Delhi Districts ................................ ...... 181 VI.4 Socio Economic Factors By Study Area: Avg. Monthly Expenditures ................. 185 VI.5 Comparing Self Reported Access to Healthcare in Three Neighborhoods ............ 186 VII.1 Example Analysis using Cropper et al. HEE from Thesis Chapter II ................... 218 VI.2 Rajarathnam et al. HEE for NO2 and PM10, with PM10 by Age and Gender ...... 220 VI.3 Thesis Ch. IV HEE Initial Application: Multiple Basic Provisions to Low SES Delhi Households Lacking Basic Provi sions (Using DHS Survey and Delhi Data) ................ 222 VI.4 First Order Computation of Reduced Traffic Fatalities ................................ ......... 223 VIII.1 Schematic Representation of a Preliminary Analytical Framework ................... 229

PAGE 15

xv LIST OF EQUATIONS Equation III.1 Estimating Excess Mortality ................................ ................................ .................... 25 IV.1 Computing Under Five Mortality Rates (U5MR) ................................ ................... 72 IV.2 Computing Relative Risk of Under Five Mortality Rates ................................ ....... 73

PAGE 16

1 CHAPTER I. I. INTRODUCTION Research Objective The overarching goals of this thesis are to explore and assess infrastructure environment health interactions in Indian c ities. More specifically, this thesis seeks to 1) explore the extent to which multiple civil infrastructures (e.g. water, sanitation, energy infrastructures) and related environmental factors (e.g. air and water quality, extreme events) can shape health ou tcomes and 2) begin to understand these interactions while also addressing issues of socioeconomic conditions, literacy, and access to healthcare. Insights are used to begin modeling the multiple risks and linkages between infrastructure provisions, envir onmental conditions, and health. Efforts using the case of Delhi, India offer a preliminary approach to helping inform future engineering, planning, and policy decision making on infrastructure development for health benefits in Indian cities such as Delhi India specifically by presenting an improved knowledge base on how infrastructures can shape health risks and benefits, and, what data is available and needed. Rationale Asian Cities will Dominate Future Urbanization and Global GHG emissions : S even bill ion people now live on this planet (UN, 2011) and over half of humanity now lives in cities ( UN, 2007). In the next 20 years, nearly 60 percent of the world's people will be urban dwellers, with rapid growth in Asian cities. Trends suggest urban population s are expected to soar to five billion from more than three billion today and this will include 60% of China's population (current 46% urban), 41% of India's

PAGE 17

2 population (currently 30%), and 87% of the USA population (currently 82% urban) by 2030 ( Ramaswami & Dhakal, 2011) Meanwhile, these same countries are estimated to contribute ~ 46% of global CO2 emissions from energy use (IEA, 2010) and ~47% of global greenhouse gas (GHG) emissions (UNFCCC, 2010) GHG emissions are being generated from various engineer ed infrastructure sectors serving cities such as transportation, energy generation / distribution, buildings, water supply, waste / wastewater management, telecommunication and industrial production ( Chavez & Ramaswami, 2013; Ramaswami, 2013 ). Inadequate Infrastructures, Urbanization, and Health in Asian Cities: The current infrastructure conditions in Asian cities are often quite poor. In Delhi, India, for example, estimates suggest ~40 50% of the population are living in slums or slum like conditions, wh ich the United Nations defines as households that lack access to improved water, sanitation, sufficient living area, durability of housing, and security of tenure (UN, 2007). Currently, 16% of households lack access to drinking water taps (putting resident s at risk of waterborne illnesses), 6% lack access to latrin es, and 8% use wood, dung, and charcoal for cooking (MoUD, 2009). In slum areas of Delhi, 19% lack water supply on premises, 76% lack an improved private toilet facility, 52% use kerosene or solid fuels, 35% lack a pucca' (solid structure) house, and 48% of these households have overcrowded sleeping areas with more than 5 persons per sleeping room (Census, 2011). As a result, a large proportion of the population may be suffering today due to curre nt local conditions. Infrastructure Environment Interactions and Health in Asian Cities: While civil infrastructures enable economic development in rapidly expanding and newly emerging Asian cities, they also pose risks to the environment (e.g. pollution, resource depletion,

PAGE 18

3 GHG emissions) that can impact public health in various ways. For example, Delhi average pollutant concentrations can be up to four times higher than national outdoor air quality standards for residential areas ( putting residents at ris k of cardiac and respiratory problems), and up to 18 times higher than drinking water quality standards (Sperling and Ramaswami, 2012). In addition, approximately 1.8 million people, primarily in India and China, die prematurely from exposure to black carb on in combustion emissions (WHO, 2010). Environmental pollution concentrations and health risks seem to be trending in the wrong direction with reports highlighting the rise of inadequacy of limited drainage connections to wastewater treatment outlets (Cen sus, 2011), electricity demand, which is set to double from 2009 levels by 2015 leading to large increases in fossil fuel burning infrastructures (Central Electric Authority, 2013), motorization: roughly 1000 vehicles added to Delhi roads daily (CSE, 2012) and health risks, with roughly one third of all medically certified causes of death due to cardiovascular and respiratory illness (NCT of Delhi, 2008). In addition, surveys are suggesting current local human health risks from infrastructure and environme ntal factors may be more important then future (uncertain) climate considerations. Public Health, Infrastructure s and Climate Change : High vulnerability of Asian cities to current and future climate related hazards (flooding, extreme heat), amplified by p oor socioeconomic and infrastructure conditions also exists. In fact, episodes of extreme weather events such as extreme heat, drought, flooding, and infectious disease outbreaks are already taking a toll on urban health and these impacts are expected to b e further exacerbated by global climate change and associated increases in climate variability (Bush et al., 2011). Heat related deaths for Delhi have been estimated to increase overall

PAGE 19

4 deaths by 2% and future CVD related deaths are estimated to rise by 4% per degree Celsius increase in temperature greater than 20 degrees Celsius (95% con fidence interval) (Hajat, 2005). Flooding events are also impacting the lives of many Delhi residents, especially slumdwellers who often live near open sewage drains with the effect of sewage flooding their homes every year during heavy rains, outbreaks of malaria, dengue, and other vector and water borne diseases (Aggarwal, 2013; De et al., 2013). Infrastructure Priorities: A key thesis question is what do Delhi residents in different socioeconomic strata prioritize in terms of infrastructures? Do they prioritize provision of basic infrastructure that can improve health and livelihoods, clean up of polluting infrastructures, or improved management of extreme weather events ? Are there infrastructure interventions that can address all three of these? What are the trade offs? Quantifying Infrastructure Pathways for Healthy and Low Carbon Asian Cities: Measuring both synergistic and antagonistic infrastructure pathways toward low carbon and healthy cities is of importance. For example, providing clean drinking water systems can place new energy demands on cities by requiring new water treatment plant facilities and can result in increases in GHG emissions, meanwhile reducing wa terborne diseases. In addition, infrastructure upgrades such as moving from improperly ventilated cook stoves burning solid fuels (dung, wood, et) to clean cook stoves and from commuting in diesel vehicles to clean mass transit can have both health and cli mate benefits. As such, this thesis defines and explore synergistic pathways achieving GHG emission reduction and health benefits and antagonistic pathways that may increase health benefits, while worsening GHG emissions or vice versa. Continued efforts in this area can help to identify many tradeoffs across infrastructure sectors that have yet to be evaluated.

PAGE 20

5 Evidence Base and Understanding of Priorities: Quantitative knowledge is missing about how multiple inadequate infrastructures and lack of quality b asic services including water supply, sanitation, and access to urgent medical care can have important health impacts. This thesis attempts to improve understanding of such knowledge gaps f or rapidly growing cities in Asia like Delhi, India, through quanti tative data analyses. In addition, local priorities are identified for planning, designing, and constructing of new en gineered infrastructure systems, which can be of significant importance to society as they affect health outcomes today and may still be i n place for many years to come. Key Question, Unique Features of Indian and Asian Cities and Contributions This thesis will assist the beginning of understanding the key research question: what are the interactions between inadequate and polluting infra structures, environmental conditions, and health outcomes in Indian and Asian cities and how are these interactions modified by socioeconomic and access to healthcare conditions? Top down studies using secondary data and bottom up studies using primary da ta collection efforts in different Delhi neighborhoods are used to inform and help answer these questions. Both offer important insights, some of which are used to compute health benefits of infrastructure interventions; with other aspects of these studies used to inform understanding of local conditions and realities. In addressing this research question, several unique characteristics of Indian and Asian cities are identified and explored, four of which are shown in the schematic below that they often a) lack adequate basic infrastructures, b) have high levels of pollution, c) have limited causes of death records, and d) lack access to quality healthcare services.

PAGE 21

6 Figure I 1 Unique Features of Indian Ci ties and the Case of Delhi, India These unique features are important, interrelated, and integral to the thesis research question for several reasons. First, with ~50% of deaths occurring outside institutions where cause of death records are maintained fo r the capital city of Delhi, India, an incomplete understanding of mortality in Asian cities and its' associations with other risk factors will continue to be the case. Improving understanding as to how inadequate infrastructures might inhibit access to qu ality urgent healthcare services may be important if to improve both the care provided for serious health episodes and the mortality records themselves. Second, Asian cities are faced with issues on both proactive and reactive sides of health risk preventi on lack of proactive health risk mitigation to avoid inadequate and polluting infrastructures causing diarrhea, respiratory illness, and road accidents to name a few; and lack of reactive risk mitigation of providing critical care to those facing such se rious illness. Third, exploring current mortality associated with high pollution exposures, lack of infrastructures or climate related hazards can only be assessed to a certain extent if mortality records are inadequate and quality healthcare access such a n important issue that it may confound potential quantitative associations.

PAGE 22

7 In order to address and improve understanding of these unique features, this thesis makes four key contributions (shown from largest to smallest): (1) A first quantitative data an alysis in Delhi and Urban India examining correlation of under five mortality with lack of adequate basic infrastructure s while controlling for the confounding effects of access to healthcare, literacy, and socioeconomic status; (2) Bottom up study of infr astructure conditions, priorities, and preliminary health diary reports in three neighborhoods of Delhi, India; (3) Characterizing experiences with and barriers to access to urgent healthcare in these neighborhoods with the impact on health not yet able to be determined due to differences in infrastructure conditions by neighborhood; and (4) Using the integration of bottom up and top down data analyses, literature review, and collected mortality data for Delhi first order health benefit computations are modeled related to findings on infrastructure provision, environmental conditions, and health. These contributions align with the WHO (2010) framework shown below, which this thesis aims to utilize and build on specifically by exploring the roles of mul tiple civil infrastructures, socioeconomic and environmental conditions in shaping health effects. Characteristics of access to / quality of health services are also explored using top down secondary data re analyses and preliminary bottom up community pri mary data analyses.

PAGE 23

8 Figure I 2 WHO Framework Model on Social Inequalities and Environmental Risks Dissertation Organization This dissertation consists of eight chapters. First, Chapter 1 provides an intro duction to the research objectives, rationale, main research question, contributions, organization, and proposed outputs. Chapter 2 provides a preliminary baseline assessment that explores the extent to which civil infrastructures (i.e., water, sanitation, energy, transport and building infrastructures) and environmental factors (e.g. air and water quality) associated with these infrastructures shape current urban health outcomes in cities in Asia using the case of Delhi, India. Analyses on current mortalit y data and a preliminary survey of local expert opinion indicate up to 19% of all recorded deaths in Delhi, India may be infrastructure related. While preliminary, the findings suggest health outcomes may be a large factor in motivating low carbon developm ent in Asian cities. Chapter 3 includes literature review of existing epidemiology, infrastructure, and environmental exposure studies and models, with emphasis on documentation of the impact of social disparities as they modify health effects estimates a nd outcomes.

PAGE 24

9 Evidence of associations between current urban health outcomes and infrastructure / environment related factors are reviewed to determine the state of evidence and gaps in forming a database of health effect estimates specifically for Delhi, I ndia and Asian cities (while presenting additional studies where useful that are often global or from North America or Europe). The goal of developing such a database is to begin developing first order computations of health benefits from infrastructure in terventions (in Ch. 7). Gaps identified in the literature inform where bottom up primary data collection efforts and community health studies are needed (e.g. access to healthcare impacts on health). Chapter 4 then describes results from a top down' anal ysis of multiple civil infrastructure conditions (including housing, drinking water, toilet facility, cooking fuels), socioeconomic conditions (wealth and literacy), proxy attributes of access to healthcare conditions (affordability and distance / transpor tation to healthcare facilities) and under five mortality rates while separating out confounding health determinants. These analyses were conducted for Delhi, India; five additional Indian cities (Hyderabad, Indore, Chennai, Mumbai, Kolkata), All India, a nd urban vs. rural conditions. Chapters Five and Six then present the results of a bottom up community study in India that aims to fill certain knowledge gaps and form baseline neighborhood level information for preliminary comparative assessment of curre nt health, environment, extreme weather, and access to quality urgent healthcare conditions for people living in three neighborhoods of Delhi having different infrastructure and socioeconomic levels. In Chapter Seven, results and new analyses are utilized for developing a first order assessment of how different infrastructure intervention scenarios can mitigate health (e.g. mortality) and climate (greenhouse gas emissions) risks, offering a

PAGE 25

10 preliminary approach to help prioritize future decision making in Asian cities Finally, Chapter Eight summarizes dissertation findings and proposes future research needs in studying the infrastructure environment climate health nexus in Delhi and other cities. Key Contributions : Linking Multiple Infrastructure Sectors, Inadequate Infrastructures, and Urban Health The broad impacts of this thesis are two fold: 1) quantitative risk assessment linking multiple infrastructures; and 2) using place and evidence based data in cities to study the infrastructure environment heal th nexus. While others have looked quantitatively at energy and fossil burning infrastructures, environment and health at country and global levels (Pruss Ustun and Corvalan, 2007), qualitatively at infrastructure environment health in cities (Campbell Len drum and Corvalan, 2007), few studies have attempted to quantitatively assess health risks of multiple infrastructures (e.g. water, sanitation, energy, transportation, and housing) and in infrastructure deprived areas in Indian cities. Relevant literature and large datasets are analyzed and synthesized across multiple infrastructures and health impacts that have often remained separate in the past. Social and environmental determinants are also linked in Chapter 4 analyses. As a result, this dissertation ma kes three original contributions : (1) An Improved Knowledge Base Addressing Inadequate Infrastructure : study provides quantitative analyses examining associations and correlations of under five mortality and inadequate infrastructure / housing conditions in Indian cities. (2) A Place based S tudy of the Infrastructure Environment Health Nexus: These complicated but important relationships are investigated in top down analyses and bottom up community studies of infrastructures, environment, socioeconomic

PAGE 26

11 condition s, local priorities, and health outcomes within three neighborhoods of the rapidly growing megacity of Delhi, India. (3) New Lines of Inquiry Linking Multiple Urban Infrastructures with Health: methods are developed to explore the key hypothetical question: i f improving health is a primary objective in Asian cities, what levels of health benefits may or may not be achieved with infrastructure related investments that focus on improved health (some of which may also have GHG co benefits)?' This question (or app roach) addresses specific features of Asian cities (e.g. poor infrastructure conditions, high levels of environmental pollution, lack of access to urgent healthcare, and undocumented causes of death), and is partially motivated by the transposed question o f other recent research looking at the public health co benefits of household energy GHG reduction strategies (Wilkinson et al., 2009), transport GHG reduction strategies (Woodcock et al., 2009), low carbon electricity generation (Markandya et al., 2009), food and agriculture GHG reduction strategies (Friel et al., 2009), and reducing short lived greenhouse gas pollutants (Smith et al., 2009). Using thesis findings and other literatures, a preliminary summary of what we know and what is needed to compute or estimate health benefits from potential infrastructure intervention scenarios is described for Delhi, India. Challenges to estimating health benefits for populations and subpopulations by social factors, and linking to first order computation of GHG co be nefits for air pollution alternatives are also described.

PAGE 27

12 CHAPTER II. II. EXPLORING HEALTH OUTCOMES AS A MOTIVATOR FOR LOW CARBON CITY DEVELOPMENT: IMPLICATIONS FOR INFRASTRUCTURE INTERVENTIONS IN ASIAN CITIES Abstract Sustainable urban infrastructure inter ventions can help achieve both public health and low carbon goals in cities. This paper explores the extent to which civil infrastructure (i.e., water, sanitation, energy, transport and building infrastructures) and environmental factors (e.g. air and wate r quality) associated with these infrastructures shape current urban health outcomes in cities in Asia using Delhi, India as a case study. Current mortality data for Delhi are used as context to estimate the extent to which urban health outcomes are shaped by infrastructure and infrastructure related environmental factors, some of which could directly or indirectly reduce mortality through low carbon interventions. Mortality data along with a preliminary survey of expert opinion indicate up to 19 percent of all recorded deaths in Delhi may be infrastructure related. More detailed epidemiology studies and infrastructure models are needed to confirm these initial findings. The findings suggest public health outcomes may be a large factor in motivating low carb on development in Asian cities.

PAGE 28

13 Introduction Over half of humanity now lives in cities. By 2030, nearly 60 per cent of the world's people will be urban dwellers, with 60% of China's population living in cities (currently 46% urban), 41% of India's popul ation (currently 30%), and 87% of the USA population (currently 82% urban) (Ramaswami & Dhakal, 2011). Meanwhile, these same countries are estimated to contribute a total of 46% of global CO2 emissions from energy use (IEA, 2010) and ~47% of global greenho use gas emissions (UNFCCC, 2010). These greenhouse gas (GHG) emissions are being generated from various engineered infrastructure sectors serving cities such as transportation, energy generation / distribution, buildings, water supply, waste / wastewater m anagement, telecommunication, and industrial production (Hillman & Ramaswami, 2011). Engineered infrastructures provide many benefits to society, including economic development. At the same time, they pose risks to the environment (e.g., resource depletio n, groundwater depletion, GHG emissions) and can impact public health in different ways. Infrastructure improvements such as moving from biomass fuel sources to clean cookstoves or installing wastewater treatment facilities can improve public health. On th e other hand, fossil fuel combustion in current infrastructure can pose a risk to health i.e., traffic congestion, air pollution, traffic accidents, and associated mortality. Public opinion surveys in China and the US suggests that local environment an d public health concerns can be a more important motivator for low carbon infrastructure interventions than the abstract goal of mitigating climate change (Lo, 2010; PRC, 2010). A few authors have discussed the relative role of carbon mitigation in the co ntext of local sustainable development priorities. One indicator for this has been the human

PAGE 29

14 development index (HDI), introduced in the early 1990s as a new way of measuring development by combining indicators for health (e.g. life expectancy at birth and under age five mortality), education (expected years of schooling), and living standards (gross national income per capita) using a scale of 0 (worst) to 1 (best). Recent research on HDI in relation to energy and carbon emissions demonstrate trends of high levels of human development being achievable with certain energy and carbon emission minimum thresholds, as well as shifting maximum thresholds. Several countries report stable and high levels of HDI corresponding with both low (e.g., Japan and Costa Rica ) and high (e.g., USA) energy usage and carbon emissions. Other rapidly developing countries like India and China are showing sharp increases in HDI for relatively small increases in energy use and carbon emissions (Steinberger, 2009). These tradeoffs betw een HDI versus energy use and carbon emissions have been evaluated in the context of local sustainable development by Amekudzi (2011). Amekudzi, Ramasw ami, Chan, Lam, and Meng (2011) identify two types of low carbon development objectives with Type 1 as GH G mitigation as the primary objective (with other co benefits as secondary objectives such as water and energy savings); and type 2 as economic development or other sustainable development priorities (e.g., health, reducing childhood mortality, etc) as the primary objective with GHG mitigation as a secondary priority (or not a priority at all). This paper asks the question if improving public health is your primary objective, what level of GHG mitigation may be achieved (or not) as a co benefit?' This appr oach is the transposed question of other recent research looking at the public health co benefits of household energy GHG reduction strategies (Wilkinson et al., 2009), transport GHG reduction strategies (Woodcock et al., 2009), low carbon electricity gene ration

PAGE 30

15 (Markandya et al., 2009), food and agriculture GHG reduction strategies (Friel et al., 2009), and reducing short lived greenhouse gas pollutants (Smith et al., 2009). The objectives of this study are to estimate the proportion of health outcomes th at may be related to infrastructure in order to identify whether urban health outcomes may be a motivator for low carbon infrastructure development in cities. Using Delhi, India as a case study, t he research question identified here is: to what extent doe s civil infrastructure (i.e., water, sanitation, energy, transport and building infrastructures) and environmental factors (e.g. air and water quality) associated with these infrastructures shape current urban health outcomes in cities in Asia ?' The paper is divided into three parts: literature review, preliminary city health data analysis, and qualitative implications for infrastructure interventions. The preliminary data analysis for Delhi includes gathering different types of public health data and local expert opinion on the relationship of mortality to engineered infrastructures and infrastructure related environmental factors. The paper concludes with a preliminary exploration and discussion of how infrastructure interventions can shape both health and low carbon goals. Such analysis provides a rationale and pathway for the twin goals of developing healthy and low carbon cities. Literature Review The literature review addresses two different perspectives: 1) The first summarizes how individual infrastr ucture affects health, examining both direct and indirect health benefits and risks; 2) The second approach traces observed negative global health outcomes to infrastructure and infrastructure related environmental factors.

PAGE 31

16 Table 1 provides a summary of th e first approach including thirty publications describing health benefits and risks (both direct and indirect) of infrastructures. The literature identified is fairly strong quantitatively in terms of environment pollution related health impacts and quite weak with respect to direct and indirect health benefits of infrastructure. The exceptions are the case of water supply and sanitation and in some cases, energy. Measurable improvements in health based on infrastructure interventions provide an important e vidence base for decision making, but in many cases the factors driving health outcomes can be quite complex, and are confounded by factors like socio economic status and accessibility (and transport costs) to hospitals and health services. Table II 1 Review of Health Benefits and Risks of Infrastructures. Infrastructure Sector Literature Review of Health Benefits (Direct and Indirect) Literature Review of Health Risks (Direct and Indirect) Transport Acces s to work, school, & essential health services (Killoran et al., 2011; Babinard & Roberts, 2006); wealth, job creation, economic development (Mohan, 2004) Accidents (WHO, 2004); transport related urban outdoor air pollution affecting asthma exacerbation, a cute & chronic bronchitis, respiratory / cardiovascular illness, & lung cancer (Samet, 2000); lack of mobility as causal factor in maternal & neonatal mortality (Molesworth, 2006) Energy Enables improved standards of living (Pasternak, 2000), extended hou rs & expanded services for hospitals (Hess, 2011; Schwartz et al., 2011); enables cooking, boiling water, space heating, cooling; electricity; transport enabling access to livelihoods; social networks; industrial production; & communication (Wilkinson, 200 7; Saatkamp et al., 2000; McMichael, 1994) Outdoor and indoor air pollution (Listorti, 2004; IEA, 2010), injury risks, and industrial hazards (Venkataraman et al., 2010); cardiovascular disease; respiratory disease; bronchitis; asthma; and eye infections ( Kammen, 2011) Water Supply & Sanitation Reduces waterborne illnesses and prevent spread of animal borne disease pathogens (Butala, 2010); reduces infant mortality (UNW DPAC, 2011) & mosquito related illnesses (Gunther, 2011) Diarrhea (Montgomery, 2007), s chitosomiasis, intestinal helminths; trachoma; trypanosomiasis (Eisenburg et al., 2001); malnutrition (Gleick, 2002); cholera; typhoid (WQHC, 1995); lung, bladder & skin cancer (Smith, 2000) Hazardous Waste Management Protects water, air, & soil by promot ing proper storage & disposal of toxic waste (Guerriero, 2009) Toxic chemicals and waste affect airway diseases and brain, lung, & gastrointestinal cancer (Rushton, 2003)

PAGE 32

17 The second approach involved exploring health outcomes, and how they rela te to poss ible infrastructure and environmental causal relationships The health outcomes that are reported in the global health literature that may relate to infrastructure include: Airborne Pollution and Health Risks: In 2000, urban outdoor air pollution caused 79 9,000 deaths globally, an estimate calculated by considering effects of particulate matter (PM10 at 10 m or less and PM2.5 at 2.5 m or less) on health for all cities with populations over 100,000 (Ezzati, 2006). More recently (2010), World Health Organization (WHO) estimates that every year, urban outdoor air pollution causes 1.3 million deaths worldwide while indoor air pollution from improperly ventilated cook stoves burning solid fuels is responsible for 1.6 million deaths worldwide (WHO, 2011). Waterborne Pollution and Health Risks: In 1995, water borne diseases caused more than five million deaths worldwide. Of these, about four million deaths were of children below age five (Gadgil, 2003). More recently (2008), estimates suggest annual mortality from waterborne diseases is closer to 3.6 million deaths per year (WHO, 2008). While 2.5 billion cases of diarrhea occur each year among children under 5, and estimates suggest overall incidence has remained relatively stable over the past two decades, diarrhea mortality has declined over past two decades from an estimated 5M deaths among under 5 children t o 1.5M deaths in 2004 (Boschi Pinto, 2009). Transport Accidents: In 2002, there was a n estimated 1.2 million deaths from road traffic accidents: an average of 3242 deaths per day (WHO, 2004) More recently (2010), road deaths accounted for 1.3 million de aths worldwide while also causing between 20 million and 50 million non fatal injuries every year (WHO, 2011). In India, where rapid motorization is occurring at a rate of 10% per year, transport accidents has become a growing public health concern, with t he country having experienced an average increase of about 4% per year in total number of traffic fatalities in the period 1997 2003, and the rate having increased

PAGE 33

18 to 8% per year since then. In 2007, 114,000 persons were killed in traffic accidents (Mohan, 2004; WHO, 2004) Cancer: In 2000, cancer was estimated to account for about 7 million deaths (12% of all deaths), with more than 70% of all cancer deaths occurring in low and middle income countries. More recently (2007), cancer accounted for 7.9M death s worldwide Projections suggest an increase to an estimated 12 million deaths in 2030 (WHO, 2011). Diabetes: In 2000, an estimated 959,000 deaths worldwide were caused by diabetes (WHO, 2004). More recently (2011), estimates suggest 346 million people w orldwide have diabetes currently, and projects that diabetes deaths now estimated at 3.8 million deaths worldwide per year will double between 2005 and 2030 (W HO, 2011). In India, 40 million suffer from diabetes, and this fig ure is estimated to go up t o 80 million by 2025. In Delhi, estimates suggest three million suffer from this disease (DFI, 2011). Traffic accidents, air pollution and waterborne disease are more readily linked with infrastructures or infrastructure related environmental factors. Inde ed, protection of public health is the basis on which WHO and India's Central Pollution Control Board have established the following: Air quality standards limiting ambient concentrations in air of particulate matter (PM), ozone (O3), nitrogen dioxide (NO2 ), and sulfur dioxide (SO2). These concentrations are often exceeded in many Asian cities, and levels beyond these standards help estimate the excess mortality from pollution episodes, which typically occur from vehicle emissions. Drinking water and river quality standards addressing pathogen content and limiting concentration of certain carcinogens in drinking water (e.g., arsenic, benzene, nickel, cadmium, etc). Levels above these prescribed standards also often occur in Asian cities as shown in Table 3 In the above cases, adverse health impacts can be expected when infrastructure condition (i.e., traffic, untreated water, etc.) are such that these environmental water or air

PAGE 34

19 quality standards are exceeded. In contrast, other diseases such as diabetes an d cardiovascular disease (CVD) risk are more difficult to link to infrastructure, however, careful qualitative quantitative studies are uncovering such linkages, as summarized in Table 2. As an example: "Development of type 2 diabetes mellitus is influence d by built environment, which is, the environments that are modified by humans, including homes, schools, workplaces, highways, urban sprawls, accessibility to amenities, leisure, and pollution.' Built environment contributes to diabetes through lack of a ccess to physical activity and increased prevalence of obesity in less walkable neighborhoods (Saelens et al., 2003). With globalization, there is a possibility that lifestyles from western countries like the US may be replicated in developing countries su ch as India, "where the underlying genetic predisposition makes them particularly susceptible to diabetes" (Pasala, 2010). Thus, diabetes which is a major health risk category is linked to the built environment infrastructure in this example. Similarly each of the health risk categories in Table 2 below is described broadly in the left column, while the infrastructure relationship is summarized in the middle column. Table II 2 Tracing Health Outcomes to Infrastructure and Environment Related Factors. Health Risk Categories Infrastructure and Environment Related Factors References Heart Diseases & Heart Attacks Fossil energy from transportation, energy generation, industrial production and buildings inf rastructure causes air pollution (among 14 identified triggers for heart attacks, along with alcohol use, anger, physical exertion, and others). Studies on obesity suggest built environments and motorized transport as major contributors to sedentary lifest yles and increased heart disease risks. Infrastructure linkages : energy, transportation, buildings, and industrial production. Nawrot, 2011; Mobley, 2006; Gordon Larsen, 2006; CDC, 2011 Airborne Pollution and Health Risks Fossil energy related indoor and outdoor airborne pollutants (e.g., sulfur dioxide, particulates, carbon monoxide), minimal distance to nearby highways, overcrowded homes, sleeping conditions, and lack of proper sanitation are contributors to serious health issues including tuberculosis, bronchitis, asthma, pneumonia, diphtheria, influenza, whooping cough, and leprosy (with inhaled respiratory droplets being an airborne transmission pathway). Infrastructure linkages: energy, transportation, buildings, sanitation, and industrial production Shaw, 2004; Ezzati, 2006; Fang et al., 2008; WHO, 2011; Agarwal, 2005; Gupta, 2007

PAGE 35

20 Table II .2 (cont'd.) To gether, Tables 1 and 2 suggest infrastructure and infrastructure related environmental factors can have tremendous local and global impact. However, detailed epidemiology study is still needed to quantitatively understand infrastructure and infrastructure related environmental factors at the city scale, particularly for diseases such as diabetes with complex infrastructure dependencies. Further, epidemiological studies comparing households with and without urban infrastructures can address confounding facto rs (e.g., socio economic status and accessibility to hospitals and health services) with regard to resulting health outcomes. Before embarking on such detailed study, we describe preliminary methods for exploring relationships between urban health and infr astructure using a mixed quantitative and qualitative study addressing the case of Delhi. We also present Table 3 as initial context for comparing existing environmental guidelines ( where infrastructure and environment related health associations are well understood) to current Delhi pollutant levels. Waterborne Pollution and Health Risks Unsafe drinking water, insufficient water for hygiene, sanitation conditions, and water rel ated insect vectors are major contributors to diarrhea, dysentery, jaundice, malaria, typhoid, acute poliomyelitis, cholera, food poisoning, and liver disease. Infrastructure linkages : water supply and sanitation (e.g., wastewater and sewage treatment). Ga dgil, 2003; Gleick, 2002; Boschi Pinto, 2009; WHO, 2004 / 2011 Cancer Hazardous waste exposure and environmental contaminants (e.g., methane, benzene, cadmium) are contributors to cancer (influencing the likelihood of developing brain, lung, and gastroin testinal cancer) and central nervous system damage. Arsenic contaminated water also contributes to lung, bladder, and skin cancer. Infrastructure linkages : water treatment and hazardous waste management. Rushton, 2010; Griffith, 1989; Guerriero, 2009

PAGE 36

21 Table II 3 Short Term (24 hour mean) Environmental Exposure Guidelines and Delhi Compliance. Environmental Compliance Standards Pollutant Concentration Guideli nes Delhi Average Pollutant Concentrations were Health Related Rationale for Standard Outdoor Air Quality WHO 2005 Guidelines: PM 10 of 50 g/m3 National Ambient Air Quality Standards (NAAQS) of India: Sensitive Areas: PM 10 of 75 g/m3; Industrial Areas: PM 10 of 150 g/m3; Residential Areas: PM 10 of 100 g/m3 Up to four times higher than the NAAQS residential standard. State of Environment for Delhi 2010 Report: 2008 PM 10 : 400 g/m3 Quote: "PM concentrations have remained (1997 2010) very high compared to the national ambient air quality standard." "Short term exposure to a PM10 concentration of 150 g/m3 would be expected to translate into roughly a 5% increase in daily mortality, an impact that would be of significant concern, and one for which immediate mitigation actions would be recommended" (WHO, 2005). "Major concerns for human health from exposure to PM 10 include: effe cts on breathing & respiratory systems, damage to lung tissue, cancer, and premature death" (US EPA, 2010) Drinking Water Quality WHO 2004 Guidelines: Benzene: 0.01 mg/L Arsenic: 0.01 mg/L India Central Pollution Control Board (CPCB) Standards: Biologic al Oxygen Demand (BOD): 3 mg/l or less (Note: this is CPCB'sC' Class Criteria set for the use of drinking water sources after conventional treatment and disinfection. ) Up to eighteen times higher than CPCB standards. Ministry of Environment, 2008: Rang e of BOD levels: 3 to 55 mg/l Quote: "Municipal Corporation of Delhi (MCD) found 15% of Delhi's water (90 out of 765 samples) to be unfit for drinking. In South Delhi, contamination is highest with 50% of samples declared polluted." (Chandel, 2009). St andards are set with the ultimate goal of no adverse health effects as a result of human uses of water. For example, under the US Safe Drinking Water Act, the US Environmental Protection Agency set standards for ~90 contaminants in drinking water and set l egally enforceable standard limits e.g. maximum contaminant levels (MCL). Water meeting these standards is considered safe to drink.' (Ramaswami et al., 2005).

PAGE 37

22 Methods A case study of Delhi was conducted to explore the extent to which current urban h ealth outcomes are shaped by infrastructure and infrastructure related environmental factors: First, baseline annual mortality data from multiple data sets in Delhi are collected and analyzed to assess data quality and to understand how the various source s of data are different. Next, associations are identified between infrastructure and health through a combination of literature review and local expert opinion on the number of observed deaths in Delhi, India as it relates to infrastructure and environme ntal factors. Local experts were selected as either internationally recognized public health and public health engineering experts based in and around Delhi; senior professionals having in depth knowledge and understanding of public health and infrastructu re development in Delhi; and senior professionals having in depth experience researching and analyzing local health and infrastructure outcomes and associations in Delhi. Third, estimates of excess mortality were computed for the case of air quality where infrastructure/environment associations are well understood to demonstrate the scale of excess deaths related to just a single environmental/infrastructure factor, s imilar to other recent studies estimating excess mortality (NRC, 2008; Jacobson, 2007).

PAGE 38

23 In future studies, similar computations will be needed for all infrastructure and environmental factors to provide cities with a template for becoming both healthy and low carbon. 1. Delhi Mortality Data: For initial exploration of health outcomes, mortalit y is selected over other indicators (e.g., morbidity) due to its' policy relevance, significant / intrinsic importance as a measure of societal well being (Sen, 1998), and the availability of mortality data at city scale. As suggested by Amartya Sen: "The significance of mortality information lies in a combination of considerations, including the intrinsic importance we attach and have reason to attach to living; the fact that many other capabilities that we value are contingent on our being alive; and the further fact that data on age specific mortality can, to some extent, serve as a proxy for associated failures and achievements to which we may attach importance" (Sen, 1998). Other metrics used by health professionals include morbidity, hospitalizatio n, disability adjusted life years (DALY), missed days at work, and height to weight ratios to reflect for malnutrition. Ideally, multiple health indicators should be used including morbidity with DALYs (Murray, 1994) both of which are sensitive to acute e nvironmental conditions (e.g., responses to high pollution episodes). However, public heath experts, recognizing the complexity in relating health outcomes to infrastructure, suggested that this initial study focus on mortality data first. Datasets: Initia lly, three datasets were reviewed to extract mortality data, including the: (1) 2010 National Health Profile by Ministry of Health's Central Bureau of Health Intelligence (CBHI) using State / Urban Territory Data for Delhi,

PAGE 39

24 (2) 2008 Births and Deaths (B&D) Repo rt for the North Capital Territory (NCT) of Delhi by the Office of the Chief Registrar, and (3) 2008 Certified Causes of Death Report for Delhi by the Office of the Chief Registrar. The 2008 Births and Deaths Report for the North Capital Territory (NCT) of D elhi was selected for further analyses as the most comprehensive, consistent, and clearly defined data providing 40 specific medically certified causes of death and providing more comprehensive accounting of cause specific deaths then other databases. For example, data inconsistencies existed in 2009 reports at time of this research and the CBHI National Health Profile presents data for only 29 specific medically certified causes of death and undercounts the total number of deaths for Delhi. The 40 specific causes of deaths as reported by the 2008 B&D Report in Delhi are shown in the Appendix and are aggregated into twelve major health risk categories that are shown in left column of the Appendix table and in the Figure 2 pie charts. Of these twelve categori es, six were found to be infrastructure and environment related based on the literature review shown in Table 2. These six categories are shown hatched in Figure 2 and were used to gather local expert opinion on the relationship to infrastructure. 2. Infr astructure Health Associations: Local expert opinions were gathered to get their best estimate at two phenomena: What percent of deaths could be attributed to infrastructure and infrastructure related environmental factors?

PAGE 40

25 How many of those deaths are e stimated to be attributed to infrastructure and infrastructure related environmental factors, could be avoided by access to health services? Overall, six experts including two government health officials, a clinical physician, a public health professor, a public health engineer, and an internationally recognized development consultant participated in a ten question survey and provided a total of about 60 quantitative replies for Delhi which guided preliminary estimates shared in this paper. As such, this su rvey approach provides a first attempt at estimating the proportion of mortality that is likely infrastructure and environment related in Delhi using local expert knowledge. We emphasize that these results are preliminary and exploratory; untangling infra structure and health relationships are complicated and additional epidemiological field data is much needed. The expert estimates summarized here, and accompanied by sample quantitative analysis, provides a starting point for detailed epidemiological study 3. Sample Estimate of Excess Mortality: The third computation shows how epidemiological data, where available, can in fact help make quantitative associations between health outcomes and varying levels of infrastructure and environmental conditions. For particulate matter concentrations in environment, as one example, the following equation (EPA, 2006) is used: "#$%&'(%)!*+,-.%(#/01#''-%(#/!23&/4+!5!677+.%!68%(9&%+!5!:/.(,+/.+!*&%+!5! 1#;-'&%(#/! Equation II I 1 Estimating Excess Mortality

PAGE 41

26 The data for this computation is based upon a single environmental factor, suspended particulate matter (SPM) smaller than about 10 microns in diameter ( PM 10 ), which is related to emissions from fossil combustion, e.g., in transportation and industrial production which also accounts for over 60% of CO2 emissions in India (IEA, 2011) and almost 80% in Delhi (TERI, 2010; Chavez et al., 2011). Epidemiological studies of mortality observed in Delhi following high PM episode s were gathered by (Cropper et al. 1997) and modeled using the Poisson model. The epidemiological study showed 4.3% and 3.1% excess CVD and respiratory mortality, respectively, associated with an increase of 100 g/m 3 in SPM (Cropper et al., 1997), which i s the effects estimate. Incidence rate refer s to CVD and respiratory mortality observed in the base case (i.e., the current condition). The equation above is applied in this paper to compute avoided mortality if Delhi were to transform their transport and indu strial emissions to meet India's NAAQS standard of 100 g/m3 in residential areas with reductions from the current annual average PM 10 levels of 400 g/m 3 (i.e. a reduction of 300 g/m 3 to be in compliance with NAAQS set by the CPCB). Results 1. Exploring Baseline Mortality, Mortality Rates, Life Expectancy, and Causes of Death In the National Capital Territory of Delhi, the mid year population of Delhi in 2008 was estimated at 17,115,000 (Government of NCT Delhi, 2008a), and the number of total deaths regi stered under the Civil Registration System included 107,600 deaths, of which 57,122 were institutional deaths (53%) and 50,742 were domiciliary (47%). 68,033 deaths were of males and 39,567 were of females. As to mortality distribution by age group, 14% of deaths were of under 5 children, 72% of deaths between the ages of 5 and 69, and 15%

PAGE 42

27 of those greater than 70 (see Figure 1). Annual mortality rates and average life expectancy in India, Delhi, and Mumbai are also presented below: Table II 4 Comparative Health Indicators for Delhi, Mumbai, and India. Health Indicators Delhi Mumbai India Est. Annual Mortality Rate (Per 100,000 population) 629 689 989 Average Life Expectancy 72 71 69 A comparis on of these two health indicators in Table 4 suggest mortality rates and life expectancy may be related. However, life expectancy as an indicator is less sensitive to changing environmental conditions on a periodic basis, e.g., acute pollution or heat ep isodes that may occur a few days a year. For such purposes, mortality is a better metric. Figure II. 1 Mortality by Age: Delhi (2008)

PAGE 43

28 2. Delhi Mortality Outcomes Associated With Infrastructure Related Factors Among the tota l registered deaths in Delhi, over 55% were not classified by cause of death. The remaining 45% (or 48,148 deaths) were grouped into twelve health risk categories of which six that had a literature based relationship to infrastructure are shown hatched in Figure 2. Among the classified deaths, up to 62% of classified deaths could be infrastructure related (as shown in pie chart B). However, this is likely an overestimate of the influence of infrastructure; to address this, we conducted local expert intervie ws. The pie chart A on left shows percentages of total deaths in 2008 for the NCT of Delhi (including unclassified deaths); Pie chart B (right) shows percentages of total classified deaths. The hatched portions represent health outcomes with potential in frastructure related linkages. Figure II. 2 Mortality by Cause (if classified) in the NCT of Delhi (2008) As shown in the figure, the 40 causes of death (presented with reported data in Appendix) are grouped into twelve health risk categories six of which are infrastructure and environment related: 1) heart diseases and heart attacks; 2) airborne disease pathogens (e.g., bronchitis and asthma); 3) waterborne disease pathogens (e.g. cholera, typhoid, diarrhea); 4) diabetes; 5) cancer; and 6) accidents. While epidemiology studies are being designed, local expert interviews were conducted using questions shown in Appendix B to estimate percentage of deaths for these six categories as (to their best guess) related to engineered inf rastructure and the environment:

PAGE 44

29 Heart Diseases & Heart Attacks (14.7% of total deaths): the median of expert responses suggests 25% could be infrastructure related of which 25% could be avoided or delayed with timely access to health services. Airborne Di sease Pathogens (5.7% of total deaths): the median of expert responses suggests 25% could be infrastructure related of which 25% could be avoided or delayed with timely access to health services. This includes deaths from tuberculosis, bronchitis and asthm a, pneumonia, diphtheria, influenza, whooping cough, and leprosy. Diabetes (4.3% of total deaths): the median of expert responses suggests 35% could be infrastructure related of which 15% could be avoided or delayed with timely access to health services. Cancer (2.85% of total deaths): the median of expert responses suggests 25% could be infrastructure related of which 15% could be avoided or delayed with timely access to health services. Transport Accidents (1.4% of total deaths): The expert responses su ggest 100% could be infrastructure related of which 15% could be avoided or delayed with timely access to health services. This includes only transport accident deaths, while the above pie charts showing 3% (Pie Chart A) and 6% (Pie Chart B) include additi onal types of accidents: burns, falls and drowning, accidental poisoning, bites, and others. Waterborne Disease Pathogens (0.7% of total deaths): The expert responses suggest 50% to 100% could be infrastructure related of which 25% could be avoided or dela yed with timely access to health services. This includes deaths from dystentery and diarrhea, jaundice, malaria, typhoid, acute poliomyelities, cholera, and food poisoning. To supplement the expert opinions, and to gain a quantitative understanding of the order of magnitude of deaths that could be attributed to environmental pollution when epidemiological data are available, a sample quantitative analysis is provided for the case of outdoor air quality.

PAGE 45

30 3. Estimating CVD & Respiratory Mortality Reduction Due to Standards Compliance A sample calculation for Delhi mortality reductions expected from compliance with India's Central Pollution Control Board (CPCB) National Ambient Air Quality Standards (NAAQS) is shown in the calculation below based on the key d ata provided in Table 3. The equation utilized is also used in the model, BenMAPP, which estimates health impacts when city populations such as Mumbai, India experience changes in air quality (EPA, 2006). By running these sorts of equations based on enviro nmental exposure models and epidemiological studies, improved understanding of infrastructure related health effects can be achieved. Figure II. 3 Estimating Air Quality Related Cardiovascular & Respiratory Mortality Reductio n in Delhi

PAGE 46

31 This sample calculation shows that quantitative epidemiological models indicate 12% of CVD and respiratory deaths (NCT Govt. of Delhi, 2008b) are pollution related in Delhi. This is due to changes in PM only. Change in levels of other air pollut ants (e.g., SO2, NOx, unburned volatile organic compounds) can also have significant health impacts (Cropper, 1997; Russell, 2008). Discussion: Implications for Infrastructure Interventions Preliminary exploration using 2008 Delhi mortality data and local expert opinion on the extent to which infrastructures and infrastructure related environmental factors shape the number of observed deaths in Delhi is summarized in Table 5 below. In aggregate as shown in Table 5, almost 19% of recorded classified deaths may be associated with infrastructure and environment related causes. It is important to recognize that 55% of the total deaths are not classified, and many of the deaths in developing world cities may not happen in institutions where causes of death recor ds are maintained. If all these unclassified deaths followed similar patterns, infrastructure and infrastructure related environmental factors shaping 19% of health outcomes could be a large factor in motivating low carbon development in Asian cities Whil e this still needs further exploration with rigorous epidemiology data, the preliminary results in this study suggest infrastructures can have a significant impact on health outcomes for developing country cities (estimated at 19%). In contrast, future he at related deaths have been estimated for Delhi to increase overall deaths by 2% and future CVD related deaths by 4% per degree Celsius increase in temperature greater than 20 degrees Celsius (95% confidence interval) (Hajat, 2005).

PAGE 47

32 These results suggest a ddressing public health outcomes at the present time may be a greater motivator for low carbon development then future climate related health risks. Table II 5 Summary of Local Expert Opinion on 2008 Deaths. Infrastructure & Environment Related Health Risk Category (% of classified deaths) Median % of Deaths Attributed to Infrastructure & Environmental Factors by Local Expert Opinion Estimated % of All Classified Deaths in Delhi That May Be Related to Infrast ructure Heart Diseases & Heart Attacks ( 32.9%) 25% 8.2% Airborne Disease Pathogens ( 12.7%) 25% 3.2% Transport Accidents (3.2%) 100% 3.2% Diabetes (7.8%) 35% 2.7% Waterborne Disease Pathogens (1.5%) 50% to 100% 0.35% to 0.7% Cancer (6.4%) 25% 0.7% As shown in Table 5, the highest percentage of mortality related to infrastructure is heart diseases and heart attacks. These preliminary results demonstrate that infrastructure interventions and air quality compliance strategies that decrease fossil en ergy use and GHG emissions can also have significant affects on reducing risks of cardiovascular and respiratory deaths, thereby leveraging both low carbon and healthy cities. As shown in Table 2, heart diseases, airborne pathogen risks, and diabetes heal th outcomes are associated with air pollution and obesity, each of which relate to those infrastructures implicated such as transport energy. Through transportation sector

PAGE 48

33 improvements alone, four of the aggregated health risk categories can be addressed v ia cleaning up of air pollution (e.g., cleaner fuels such as low sulphur diesel, reducing vehicle miles travelled (VMT), making vehicles more efficient), promoting mode shifts that make communities more walkable, physically active, and perhaps more transit dependent thereby reducing road conflicts. Such infrastructure interventions reducing VMT, cleaning up fuel, making vehicles more efficient, and mode shifts to transit modes all can help achieve both health and low carbon city goals. Industrial symbiosis and other clean energy interventions can also clean production, thereby reducing airborne pollution (e.g. SO2, NOx, and PM10). With airborne pathogen risks, tradeoffs exist between house size efficiency (for energy consumption and low carbon goals) and mo rtality (as it relates to overcrowded sleeping conditions affecting tuberculosis and other health outcomes). Passive strategies for building and home design that increase ventilation and provide natural cooling reduce energy loads and indoor air pollution. Measuring both synergistic and antagonistic pathways toward low carbon and healthy cities is of importance. Synergistic pathways achieve both GHG emission reduction and health benefits. Antagonistic pathways may increase health benefits, while worsening G HG emissions or vice versa. For example, providing clean drinking water systems and reducing exposures to cancer causing pollutants through improved hazardous waste management systems may place new energy demands on cities by requiring new water treatment plant facilities and new transport requirements for hazardous waste that can result in increases in GHG emissions, that reduce waterborne diseases and cancer related deaths. Synergistic pathways meanwhile do exist as

PAGE 49

34 demonstrated by Miller et al. who has d one a life cycle assessment of energy use in wastewater treatment in Hyderabad, India, showing that energy investments in wastewater treatment plant infrastructure may actually reduce overall GHG emissions, by reducing emissions of methane and nitrous oxid e from untreated sewage (Miller et al., 2011). Wastewater treatment therefore may be net GHG mitigating. Conclusion "To develop low carbon cities of the future, carbon mitigation potential must be combined with quantitative analysis of other benefits and c o benefits, including energy security and public health, suited to the overall goals of society." A. Ramaswami & S. Dhakal, 2011 Wit h some 1050 cities in the US (US Mayors, 2011), eight cities in China, and 40 cities in India setting low carbon develop ment goals (ICLEI SA, 2009), city scale strategies focused on infrastructure and low carbon interventions tailored to fit the unique local cultural, social, economic, public health and human development aspirations in each city, can (presumably) be more s uccessful then failing international efforts focusing on carbon mitigation as primary objectives. Results from this preliminary exploration indicate up to 19 percent of all classified deaths in Delhi may be infrastructure related, and as shown here, reduci ng mortality through infrastructure interventions provides an important driver for low carbon city development. We emphasize that additional exploration of infrastructure health associations and causal relationships through field work and epidemiological s tudy is needed. The results of this study provide an initial qualitative quantitative exploration that opens up a new line of inquiry. The initial explorations in this paper can be complemented by future epidemiological assessment of the different degree s of health hazards associated with the infrastructure in Delhi from 2008 to 2011. Currently, initial estimates suggest 55% of

PAGE 50

35 households live within 500 meters of roads with high levels of air pollution (putting residents at risk of cardiac and respirator y problems), 16% of households lack access to drinking water taps (putting residents at risk of waterborne illnesses), 6% lack access to latrines, and 8% are using solid fuels (wood, dung, and charcoal) for cooking (Govt. of Delhi, 2009; Govt. of India Min istry of Urban Development SLB Databook, 2009). Additional research and a carefully constructed epidemiological study design could be useful in establishing objective associations on mortality statistics or other health indicators and the provision and upg rading of specific infrastructures, including water, sewage, electricity, transport, and buildings (e.g. upgrades to pucca and kutcha' housing) within the city's geographic areas. Health effect estimates, risk factors, and potential confounding factors fo r relevant infrastructure related health outcomes need to be identified in such a study. For example, is there a measurable improvement in households with piped water and how can confounding factors like age, socio economic status, access to hospitals and health services be accounted for? Acknowledgments This study was supported by an Integrative Graduate Education and Research Traineeship Grant (IGERT; NSF Grant #DGE 0654378) from the United States National Science Foundation (NSF) to the Center for Sust ainable Infrastructure Systems a t University of Colorado Denver

PAGE 51

36 Appendix. Chief registrar for NCT of Delhi Re analysis of births and deaths 2008 report for Delhi. Chief Registrar for NCT of Delhi Births and Deaths 2008 Report for Delhi No. Aggregated Health Risk Categories No. of Deaths % of All Deaths Medically Certified Cause of Death (% of All Deaths By Specific Cause) Recorded Deaths By Cause [Infrastructure Causal Relationship] 1 Heart Diseases & Heart Attacks 15876 14.75% Heart Diseases & Hear t Attacks (14.77%) 15876 [transport, energy, buildings, industry] Tuberculosis (2.45%) 2632 [air pollution and overcrowded conditions] Pneumonia (1.43%) 1539 [E, T&AQ, O] Bronchitis & Asthma (1.22%) 131 6 [E, T&AQ] Diptheria (0.23%) 345 [O & poor sanitation] Influenza (0.15%) 161 [T minimal distance to nearest national highway] Whooping Cough (0.06%) 60 [O] 2 Airborne Disease Pathogens 6114 5.68% Leprosy (0.06%) 61 [O] 3 Diabetes 4626 4.30% Diabetes (4.30%) 4626 [T&LU] 4 Cancer 3070 2.85% Cancer (2.85%) 3070 [HW, T&AQ, E] Transport Accidents (1.43%) 1540 [T&AQ] Accidental Burns (0.62%) 670 Falls & Drowning (0.17%) 187 Accidental Poisoning (0.11%) 122 Bites (0.05%) 49 5 Accidents 2765 2.57% Ot her Accidents (0.18%) 197 Dysentery & Diarrhea (0.19%) 206 [WS] Jaundice (0.13%) 140 [WS] Malaria (0.09%) 96 [WS] Typhoid (0.08%) 90 [WS] Acute Poliomyelities (0.08%) 87 [WS] 6 Waterborne Disease Pathogens 714 0.66% Cholera (0.0 7%) 79 [WS]

PAGE 52

37 Food Poisoning (0.01%) 16 [poor sanitation] Homicide (0.14%) 147 7 Homicides & Suicides 333 0.31% Suicide (0.17%) 186 Birth Related Affecting Child (3.07%) 3303 [H] Cerebral (Paralysis) (0.68%) 7 29 Pregnancy Complications (0.15%) 165 [H] 8 Pregnancy & Birth 4231 3.93% Abortion (0.03%) 34 Tetanus (0.80%) 864 Meningitis (0.61%) 652 Rabies (0.08%) 84 Syphillis (0.07%) 72 9 Other Infectious Diseases 1672 1.55% Measles (0.04%) 38 Chronic Liver Disease & Cirrhosis (1.10%) 1187 [poor sanitation] Appendicitis (0.40%) 427 10 Disease of Digest ive System 1374 1.28% Ulcer of Stomach & Duodenum (0.06%) 63 11 Genetic Blood Related Diseases 625 0.58% Anemia (0.58%) 625 12 Elderly 6545 6.08% Senili ty (6.09%) 6545 A Deaths With Cause Not Classified 59314 55.12% % of Deaths Infrastructure Related 27.93% B Deaths With Cause Classified 48286 44.86% C Total Recorded Deaths 107600 100.00% LEGEND for Infrastructure Related Causes of Death (Informed by Lit. Review): O = Overcrowded housing and sleeping conditions ; WS = Water Supply & Sanitation (e.g., inadequate drinking water or sewerage systems) ; E = Energy (e.g., indoor smoke from solid fuels) ; HW = Hazardous waste (e.g., inadequate storage and disposal of toxic waste) ; T&AP = Transport & associated air pollution (e.g., outdoor pollution, road accidents) ; H = Health Infrastructure (e.g., inadequate accessibility to hospital and/or health services) Notes : Associations of i nfrastructure related health factors to causes of death are shown in [ ] (also see Leg end ) and are informed by the literature below ( full citations available in the references): Shelter / Crowded Sleeping Conditions: tuberculosis (Agarwal, 2005); measles (Agarwal, 2005) Water Supply and Sanitation: cholera (Butala, 2010); typhoid (WHO, 2011); dysentery and diarrhea (Esrey, 1991; Gleick, 2002); liver disease (WHO, 2011) Energy: tuberculosis (Agarwal, 2005), pneumonia (WHO, 2011); cancer (Slaper et al., 1996; WHO Fuel for Life, 2006); Hazardous Waste: cancer (Guerriero, 2009); brain, lung, and gastrointestinal cancer (Rushton, 2009; Griffith, 1989 Transportation: accidents (Mohan, 2004); heart disease (Pande, 2002; Nawrot, 2011); pneumonia / bronchitis/ asthma (Krzyzanowski, 2005); infl uenza (Fang et al., 2008) Health Infrastructure: complications related to pregnancy (Molesworth, 2006); Condition such as birth injuriespremature originating in perinatal period (Aguilera, 2006)

PAGE 53

35 CHAPTER III. III. A REVIEW OF EPIDEMIOLOGICAL STUDIES RELATING T O INFRASTRUCTURE AND POLLUTION IN ASIAN CITIES: EVIDENCE AND GAPS FOR INFRASTRUCTURE RELATED HEALTH EFFECT ESTIMATES IN DELHI Abstract Many Asian cities curr ently lack analytic tools and an evidence base to link infrastructure improvements with health out comes. This paper builds a preliminary foundation to address such challenges by conducting a review and synthesis of literature, models, and field data. Over eighty quantitative epidemiological studies, using literature focused on Asian cities where possib le, are identified and reviewed across seven categories of infrastructure related health outcomes: cardiovascular disease, respiratory, airborne infectious disease s transport accidents, waterborne pollution and health risks, cancer, and diabetes. An initi al database of health effect estimates is created to model mortality risk mitigation options. K nowledge gaps are also identified where further bottom up study may be needed for example, how access to urgent healthcare shapes health outcomes By reviewing epidemiology studies across Asian cities and specifically for Delhi, India this top down study helps to inform areas where bottom up field study are still needed and also to help inform a first order scenario tool that can be used to compute health risk reduction benefits in Asian cities and for Delhi, India in particular.

PAGE 54

36 Introduction Asia now represents half of the world's urban population. Cities in India and China alone make up 28% of world urban population and by 2030, Indian and Chinese cities wil l represent 40% of global urban population. Such urbanization creates huge demands for infrastructures, basic services, and improved environments. In response to such challenges, the World Health Organization (WHO) in 1994 created the Healthy Cities initia tive to improve health and quality of life in global cities. Since then this initiative has taken the form of a long term program aiming to place health high on agendas of decision makers and to promote comprehensive local strategies for health protection and sustainable development. Such goals become increasingly critical with half the world's population now living in cities and ~1/3 of the world's urban population living in what are defined by the United Nations as slum conditions ( UN Habitat 2006). In this study, we analyze current health risks facing Asian cities posed by infrastructures (or lack thereof) and infrastructure related environmental factors. In Delhi, India for example, the current urban population of 18 million is expected to soon reach over 26 million by 2025, creating demands for not only new civil infrastructures (i.e. water, sanitation, energy, transport and building infrastructures), but also infrastructure upgrades that help improve environmental conditions. In Delhi, air quality is currently up to four times worse than national ambient air quality standards and water quality is up to eighteen times worse than the Central Pollution Control Board water quality standards; 55% of households live within 500m of roads with high levels of air pollution (putting residents at risk of cardiac and respiratory problems), 16% of households lack access to drinking water taps (putting residents at risk of waterborne illnesses), 6% lack access to

PAGE 55

37 latrines, and 8% are using solid fuels for cooking (N FHS, 2006) While an understanding of health effects associated with provision of infrastructures and infrastructure related environmental conditions would indeed be beneficial, such a database for Asian cities does not yet exist. Meanwhile, various threa ds of evidence on infrastructure and environment related health impacts in Asian cities have evolved in recent decades. This is particularly the case for outdoor air pollution in Asian cities with now well over 140 studies (most conducted in China) publish ed in peer reviewed literature presenting original estimates of health effects of outdoor air pollution (HEI PAPA, 2004). At the same time, models that connect fossil combustion from transportation, power generation, industrial processes with air quality a nd public health are highly limited in their application. Methods This study conducts literature review of existing epidemiology, infrastructure, and environmental exposure studies and models, with emphasis on documentation of the impact of social dispari ties as they modify health effects estimates and outcomes. Evidence of associations between current urban health outcomes and infrastructure / environment related factors are reviewed to determine the state of evidence and gaps in forming a database of hea lth effect estimates specifically for Delhi, India and Asian cities (while presenting additional studies where useful that are often global or from North America or Europe). The goal of developing such a database is to begin developing first order computat ions of health benefits from infrastructure interventions. Gaps identified in the literature inform where bottom up study is needed.

PAGE 56

38 In a recent study by Sperling & Ramaswami (2012), major categories of urban health outcomes in Delhi are traced to infrast ructure (e.g. water, energy, transport) and infrastructure related environmental factors (e.g. air and water quality). These major categories include cardiovascular disease, respiratory, airborne infectious disease pathogens, waterborne pollution and healt h risks, transport accidents, cancer, and diabetes To build on this work, this review takes the approach of assessing quantitative associations between infrastructure and health using the case of Delhi, India and global cities where relevant studies are a v ailable. The review summarizes ~8 0 studies presenting infrastructure and infrastructure related mortality risk factors and health effect estimates. Literature Review While not all potential risk factors are addressed in this literature review, as this wo uld be nearly impossible, we highlight key risk factors and epidemiological evidence on health effect estimates primarily from cities in India, Asia, and the USA for comparison. In all cases, we note whether the study, in developing its' hea lth effect esti mates, consideres differential risk by age, gender, ethnicity, income levels, education levels, and access to health services. Conclusions reached and study limitations where explicitly mentioned by authors' themselves are presented. Where primarily ur ban health studies from Delhi and Asia are lacking, we identify health effect estimates for infrastructure related risk fact ors from studies in US, Europe and OECD countries Results: Risk Factors and Health Effect Estimates I nfrastructure related mortal ity risk factors and epidemiological HEEs are described below for the identified seven majo r categories of infrastructure related health outcomes

PAGE 57

39 and are presented in the order of cardiovascular disease, respiratory, airborne infectious disease, waterborne pollution and health risks, transport accidents, cancer, and diabetes Cardiovascular Diseases (including Heart Diseases and Heart Attacks): The major mortality risk factor reported from the literature for this health outcome are emis sions of particulate matter which can be ten times more harmful than ozone exposure levels. In fact, more than 500,000 people die every year from diseases related to air pollution, often attributed to particulate matter (PM10) in outdoor air ( WB 2000). One study, relevant to Table 1, notes that PM10 sources vary, and multiple interventions are therefore needed to reduce PM10 concentrations. Garg (2011) finds that 31% of PM10 emissions in Delhi, India are attributable to fossil fuel combustion and 11% to biomass combustion, and 17% from other sources of ambient dust s uch as construction activities. H owever, the dominant sources can vary across cities, particularly the newly industrializing ones. Such source information is essential to estimate the impact of infrastructur e in terventions (e.g. mode shifts to transit) on air quality given that most Asian cities do not typicall y have transport models or industrial emission models. Definition : "Parti culate matter (PM), in the form of PM less than 10 m and 2.5 m in aerodynamic diameter (also referred to as PM10 and PM2.5), is inhalable material emitted directly from motor vehicles, power plants, and other sources or formed in the atmosphere through r eactions with gaseous emissions (eg, nitrogen and sulfur oxides [NOx and SOx] react to form nitrates and sulfates, respectively). Although the health effects of PM have been of concern for many decades, short term and long term epidemiologic studies publis hed in the United States and Europe in the 1990s found associations of PM with increased morbidity and mortality at ambient levels below the national air quality limit values at the time, the basis of action in both the European Union and the United States to establish more stringent standards for PM." HEI, 2004

PAGE 58

40 Studies shown in the below tables have (in some cases) identified (quantitatively) differential risk by controlling for and also differentiating findings across subpopulations, many of whom often have different exposures to hazards based on their different infrastructure conditions (e.g. energy, water, sanitation, transportation, buildings), environmenta l conditions (e.g. air and water quality, seasonal variation, weather and climatic changes) and also biological (e.g. age, gender, ethnicity) and socio economic conditions (e.g. income, education, access to healthcare). These subpopulations may also be mor e susceptible or vulnerable to the resultant health effects (WHO, 2010). For the available health effect estimates for CVD, many of the studies are short term exposure studies with only a few long term studies (e.g. the Pope et al.,1995 American Cancer Soc iety cohort study). In the table, each study identified is characterized by the HEE findings, whether the study was short or long term, the exposure concentrations for the various study localities, and the models used. HEEs in many cases are developed as i t relates to sensitivity by age, SES factors such as income, and whether having asthma or not (e.g., Pope, 2011). Meanwhile, none of the identified studies consider HEEs in terms of sensitivity related to inadequate or delayed access to healthcare often a common issue in Asian cities (Goli et al., 2011). An important limitation of this review is that many additional studies may exist in the literature as air pollution remains a major investigation field and action domain for improving public health in c ities globally. The research articles obtained were through a a search of air pollution and mortality initially focused on only a couple Indian cities, followed by Asian cities, then the US and Europe where larger bodies of evidence exist.

PAGE 59

41 Table III 1 Available Health Effect Estimates for CVD and Total Mortality Risk. Health Hazards Epidemiological Health Effect Estimates / Dose Response Relationships Quan tification of Differential Risk by: Age (A), Gender (G), Ethnicity (E), Education (Ed), Income (I), Season (S) hlth care access (ATH) PM10 (Indian cities) (1) Cropper et al, 1997 (Delhi India ): 0.43% increase in CVD mortality & 0.23% increase in total mortality per 10ug/m^3 increase in PM10 ; Findings base d on 1991 to 1994 study of short term exposure; avg. total suspended particulate (TSP) was 375 ug/m^3 over study period (~5 times higher than WHO annual standard); levels also exceeded WHO 24 hour standard 97% of the time. Model used : Poisson Study Limita tion /s : results based on limited cause specific mortality data representing only 25% of deaths during study period; TSP no longer routinely monitored or considered good marker for health effects (Balakrishnan et al. 2010) (2) HEI short term study by Rajarathna m et al., 2010 (Delhi, India): 0.15% increase in All Natural Cause Mortality per 10 ug/m^3 in PM10 (all natural cause excludes the following causes of death: accidents and suicides/homicides); Study of short term exposures from 2002 2004; annual mean PM10 concentration: 170 ug/m^3 (~3 times higher than Central Pollution Control Board Nat'l. Ambient Air Quality Stndrd. of 60 ug/m^3 for residential areas); Model used: quasi Poisson regression model ; Study Limitation/s: AQ data from 9 of 10 stations available for < 100 days per year ; no monitoring on weekends/holidays; medically certified cause of deaths for only 55% of total deaths (3) HEI short term study by Balakrishnan et al., 2010 (Chennai, India): 0.44% (95% CI: 0.17 0.71) increase in All Natural Cause mortali ty per 10ug/m^3 increase in PM10 concentrations ; Daily avg for PM10 at 92 ug/m^3 ; often exceeded NAAQS 100 ug/m^3 across ind., resid., & commercial zones; Model used: quasi Poisson generalized additive model; Study Limitation/s: stations monitor only 100 120 days / yr; no monitoring on weekends/holidays; HARMONIZATION : Range of 0.15% 0.44% increase in all natural c ause mortality per 10ug/m^3 increase in PM10 The (#) below corresponds to study identified in left column. The following studies only add ress: (1) A (2) A; G (3) A; G; S The studies do not address E, ED, I, & A T H.

PAGE 60

42 Table III .1 (cont'd) PM10 (Asian, US, & Other Global Cities) (4) HEI PAPA 2004 meta analyses of short term studies (Seoul, Honk Kong, Bangkok, Inchon): 0.41% increase in mortality per 10ug/m3 increase; 29 European Cities: 0.6% increase per 5 0ug/m3 increase ; concentrations ranged from 10 to 290ug/m3; Poisson model; (Katsouyanni et al., 2001); 90 US Cities (Samet et al. 2000): 0.41% increase (5) Chen, 2011 (Beijing, Shanghai, an d Shenyang): short term study finding a 10 g/m3 increase in 1 day lagged PM10 2.5 was associated with a 0.25% (95% CI: 0.08 to 0.42) increase in total mortality, 0.25% (95% CI: 0.10 to 0.40) increase in CVD mortality, and 0.48% (95% CI: 0.20 to 0.76) inc rease in respiratory mortality. Avg. d aily concentrations of PM10 2.5: 101 g/m3 (Beijing, 2007 2008), 50 g/m3 (Shanghai, 2004 2008), and 49 g/m3 for Shenyang (2006 2008); GLM Poisson model used (6) Bell et al. 2013 (Global review: 31 European studies (12 in Italy), 24 in Asia, 8 in Canada, 7 in Latin America, 1 in Russia / Australia with some as single and multi city studies): short term exposure studies of subpopulations: consistent evidence that elderly experience higher risk of PM associated hospitalizatio n / death and suggestive evidence that those with lower education, income, or employment status have higher death risk. Meta analysis findings: 0.64% (95% CI: 0.50 0.78) increase in death risk for older populations compared with 0.34% (95% CI: 0.25 0.42) f or younger populations per 10ug/m3 increase in PM10. Women: 0.55% (95% CI: 0.41, 0.70) vs. Men: 0.50% (95% CI: 0.34 0.54) (not statistically significant difference). (7) Yang et al. 2013. (Beijing after the 2008 Olympics): time series analysis using generaliz ed additive model: 1.8% increase (95% CI: 1.21 2.40) for 10 ug/m3 increase in PM10; mean daily average PM10 concentrations were 121.04ug/m3 from 2009 2010; CVD mortality increase per 10ug/m3 for males: 0.96% vs. females: 2.64%; elderly age 65+ : 1.97% vs. 45 and below: 0.53% (8) Ostro, 2004 (Global Cities Short Term Studies Review ): All natural cause mortality % increase per 10ug/m3 change in PM10: 0.8%; 1.7% in CVD mortality (Bangkok: Ostro e t al. 1999); 1.83% (Mexico City: Castillejo s et al., 2000); 1.1% (Sa ntiago : Ost ro et al., 1996); 0.8% (Incheon : Hong et al., 1999); 1.6% (Brisbane: Simpson et al. 1997); 0.95% (Sydney: Morgan et al., 1998). (9) Schwartz, 1994 short term studies; All Natural Cause Mortality: 0.4% (e.g. Santa Clara) 0.9% (Tennessee) increase in total mortality per 10 ug/m^3 increase in PM10 (study includes Philadelphia, Detroit, Minneapolis, Utah Valley, St Louis, etc) ; average concentrations of TSP ranged from 56 to 111 g/m3; Model used: Poisson (10) Dominici & Samet et al., 2002 (88 US Cities): 5% increases in total mortality per 10ug/m3 increase in PM10 (i.e. Denver, CO: ~0.5%) ; Short term exposure effect studies using the National Morbidity, Mortality, and Air Pollution Study (NMMAPS) data base from 1987 1994; hierarchial linear model used (11) Dock ery at al. 1993: Harvard Six Cities (Steubenville, St. Louis, Harriman, Watertown, Topeka, Portage) long term cohort study with 16 year follow up of 8111 adults: All cause of death risk ratio of 1.26 (95% CI: 1.08 1.47) for most vs. least polluted city; M odel used: Cox Proportional Hazards Survival model (12) Pope et al. 1995 (151 U.S. Metro areas): American Cancer Society long term cohort study (the original ACS study) from 1982 to 1989 of 552,138 adults: All cause of death risk ratio of 1.17 (95% CI: 1.09 1.1 7); HARMONIZATION : Range: 0.25% 0.9% increase in All Natural Cause mortality per 10ug/m^3 increase in PM10 (4) Only A is addressed in this study (5) Only A is addressed in this study (6) A, G, Ed, I, and Employment (7) A; G (8) A / E are both addressed; A T H is acknowledg ed as important, but not addressed explicitly. (9) DR not addressed in this study (10) DR not addressed in this study (11) A; G; E; smoking; occupational exposure; BMI (12) E and smoking addressed

PAGE 61

43 Table III .1 (cont'd) PM2.5 (13) Ostro et al., 2006: % incr ease in total mortality per 10ug/m3 change in PM2.5: 0.6% (nine California counties short term exposures study from 1999 2002; mean daily PM2.5 concentrations: 14 to 29 ug/m^3; Model used: Poison multiple regression model using natural or penalized spline s; Study Limitation/s: exposure measurement error ) and 1.4% (Mexico City short term study by Borja Aburto et al., 1998) ; 0.6% in CVD mortality and 2.2% in respiratory mortality respectively (nine CA counties) (14) Krewski et al. 2009 (US nationwide, NYC & LA): % mortality increase per 10ug/m3 change in PM2.5 (1979 1983 : mean concentrations: 21.2 ug/m3 ; ): All Natural Cause Mortality: 0.04%; Ischemic Heart Disease (I HD): 0.18%; Cardiopulmonary : 1.09%; (1999 2000 ; mean concentrations: 14.02 ): All Causes: 0.06%; IHD: .24%; CP: 0.14% ; random effects Cox model used (15) Laden et al. 2006 (Extended Follow up of Harvard Six US Cities Study): % increase per 10 ug/m^3 change: All Natural Cause Mortality: 0.16%; CVD: 0.28 1.0% (16) Miller et al. 2007 (Long term Cohort study separa te from ACS cohort: 65,893 postmenopausal women without preexisting CVD in 36 U.S metropolitan areas: Difference of 10 g/m3 PM2.5 associated with 24% increase in risk of a cardiovascular event (relative risk 1.24; 95% CI 1.09 1.41) and 76% increase in risk of CVD mortality (RR 1.76; 95% CI 1.25 2.47) HARMONIZATION : Range of 0.04 % 1.4% increase in All Natural Cause mortality per 10ug/m^3 increase in PM2.5 (13) A; G; E; Ed; and diabetic subpopulation. I, AH, S not addressed (14 ) E (e.g. black, white, hispanic); I (e.g. unemployment, median HH income; income disparity; Ed (e.g. < high school education). AH is not addressed. (15 ) TBD (16) A; G; E; diabetes NO2 (2 ) Rajarathnam et al., 2010: 0.84% increase in total mortality per 10ug/m^3 change (7) Yang et al. 2013: 2.63% increase in nonaccidental mortality per 10ug/m3 change (2) A; G (7) A; G NOx (17) Greenbaum, 2011: 0.65% increase in mortality per 10ug/m^3 increase (17 ) DR not addressed SO2 (2) Rajarathnam, 2010: Delhi: no significant effect for SO2 concentration changes (2) DR not addressed Ozone (18 ) Smith, 2009: CVD: 0.22 % change in mortality per 1 ug/m^3 change in pollutant (18 ) DR not addressed Ozone (19 ) WHO, 2011(Europe): heart disease: .3 .4% mortality increase per10 g/m3 change (19 ) DR not addressed Sulphate (18 ) Smith, 2009: % change per 1.0 ug/m^3 change in p article sulphate concentrations: All Cause: 0.12 0.88; CVD: 0.09 0.36; Cardiopulmonary: 1.01; Respiratory: 0.37 0.70 (18 ) DR not addressed Solid Fuel s (17 ) Smith, 2000 (India): CVD: 0.287% increase in mortality for children <5 exposed to solid fuels (17 ) DR not addressed

PAGE 62

44 Table III .1 (cont'd) Note: Many estimates are based on specific thresholds of environmental pollution exposures, with estimates presented as relevant to cardiovascular disease (CVD), cardiopulmonary, coronary heart disease (CHD), and all natural cause mortality. Black Carbon (19) Gan, 2011 (Vancouver) : An elevation in black carbon (0.97 10 5/m) was associated with a 4% increase in CHD mortality. (19 ) DR not addressed Diet (20)WHO GBD, 2009: Insufficie nt intake of fruit & vegetables (prevalence measure of 5 servings / day) is esti mated to cause ~ ~11% of ischaemic heart disease deaths worldwide. (20) G, I

PAGE 63

45 Airborne Pollution Risks for Respiratory Mortality : The major mortality risk factors and health effect estima tes reported from literature for respiratory deaths (e.g. tuberculosis, bronchitis, and asthma including ICD8 460 519, and excluding 463, 464, and 474) are similar to those for CVD. Outdoor air pollutants such as PM, ozone, and sulphate are all important risk factors. Other risk factors include indoor air pollutants from use of solid fuels for cooking and as sociated black carbon emissions. The HEEs in Table 2 are presented as relevant to respiratory mortality, which includes deaths due to acute respirator y infection, asthma, pneumonia, and bronchitis. In most cases, age is considered and in some cases gender, education, and seasonal variation is addressed. Table III 2 Available Health Effect Estimates for Re spiratory Mortality. Health Hazards Epidemiological Health Effect Estimates / Dose Response Relationships Differential Risk (DR) by: Age (A), Gender (G), Ethnicity (E), Education (Ed), Income (I ), Access to Healthcare (AH), Seasonal (S) PM10 (Indian citi es) (1) Cropper et al, 1997 (Delhi): 0.31% increase in respiratory mortality per 10ug/m^3 increase ; in PM10 ; Findings based on 1991 to 1994 study of short term exposure; avg. total suspended particulate (TSP) was 375 ug/m^3 over study period. Model used : Poisson model (1) A PM10 (Asian, US, & Global Cities) (2) Ostro et al, 1999 (Bangkok): 3 6% increase in respiratory mortality and a 1 2% increase in all natural cause mortality per 10ug/m^3 increase in PM10 (3) Tellez Rojo et al, 2000 (Mexico City): 2.9 4.1% increase in resp mortality per 10ug/m^3 PM10 increase (2) A (3) A; AH E, ED, I are not addressed in these studies

PAGE 64

46 Table III .2 (cont'd.) PM2.5 (4) Laden et al 2006 ( Extended Long Term' Follow up of Harvard 6 US Cities): R espiratory mortality increases by 0.08% per 10ug/m^3 change in PM2.5 exposure levels; all natural cause mortality by 0.16% ; 1979 1997 annual avg. PM2.5 concentrations ranged from 10 to 40 u g/m^3 ; model used: Cox proportional hazards regression model (5) F ranklin et al 2007 (27 US Communities short term case crossover study from 1997 to 2002 ): 1.21% increase in all cause mortality; 1.78% increase in respiratory mortality; and 1.03% increase in stroke related mortality with a 10 ug/m3 i ncrease in previous da y's PM2.5; 9.3 in Palm Beach, FL to 28.5 in Riverside, CA with mean concentration across all communities:15.7ug/m3; model used: conditional logistic regression model (6) Ostro 2006 (9 California Counties short term exposure study from 1999 to 2002 ): 10 ug/ m3 change in 2 day average PM2.5 conc corresponded to 0.6% increase in all natural cause mortality, 0.3% in CVD; 1.3% in respiratory. Mean daily PM2.5 levels ranged from 14 g/m3 in Sacramento and Contra Costa Counties to 29 g/m3 in Riverside County, exce eding the U.S. EPA annual average PM2.5 standard of 15 g/m3 in six of the nine counties; Models Used: random effects model / penalized and natural spline model (7) Englert, 2004 (Re analysis of three long term exposure cohort studies: Harvard 6 US Cities Study; ACS study, and AHSMOG study): key findings: difference in lung cancer estimates by gender is considerable (RR = 2.14 (male) vs. 1.20 (female) per 10ug/m3 increase in PM2.5; highest effects also linked with lowest education,especially for lung cancer (4) A; G; Ed (5) A; G (6) A; G; E; Ed; S ; diabetic (7) A; G; Ed; I; and smokers Ozone (8 ) Lancet Short Lived GHG Smith, 2009 (US); 0.57% change in resp. mortality per 1 ug/m^3 change in pollutant (9 ) Ji, 2010; Emergency hospitalizations for total res piratory disease increased by 4.47% per 10 ppb 24h ozone among elderly w/out adjustment for publctn bias (2.97% w adjustment) (10 ) Bell et al, 2004 (US Natl Avg: 95 large US urban communities, including ~40% of the total US population): A 10 ppb increase i n the previous week's ozone was associated with a 0.52% increase in daily mortality and a 0.64% increase in cardiovascular and respiratory mortality. (8 ) DR Not Addressed (9 ) Study Addresses A&S. G, E, Ed, I, and AH not addressed (10 ) Study Addresses A &S. G, E, Ed, I, and AH not addressed Solid Fuel Use (11 ) Smith, 2000 (India): ARI: 0.381% increase in mortality for children <5 exposed to household use of solid fuels; TB: 0.366% increase for children <5; Asthma: 0.315% increase (11 ) A Particle Sulphat e (12 ) Smith, 2009 (Multi city study in 6 Californian counties): 0.7 % change in resp. mortality per 1ug/m^3 change in pollutant (13 ) Dockery et al.,1993 (US): 1.26% change in all natural cause mortality for sulfate particles exposure between the between t he most polluted city and the least polluted city (range: 4.8 to 12.8 g per cubic meter) (12 ) DR not addressed (13 ) Only Ed addressed. Malnutriti on (14 ) James, 1972 (San Jose, Costa Rica cited by CPCB, 2008 Delhi Study on Ambient Air Quality, Resp. Sym ptoms & Lung Function of Children): Malnourished children experience 2.7 times more bronchitis & 19 times more pneumonia than normal weight properly nourished children (14 ) Only age (and weight) addressed

PAGE 65

47 Airborne Infectious Disease Risks : HEEs have bee n identified in Table 3 specifically for the infrastructure related risk factor of overcrowded sleeping conditions. The estimates are presented as relevant to airborne infectious diseases, which includes the following diseases: tuberculosis, influenza, dip theria, whooping cough, and meningitis. As shown in the Table below, the only health effect estimates identified so far for a global south city in terms of overcrowding and its impacts on TB mortality was for Sao Paolo, Brazil (Antunes et al., 2001). Addi tional HEEs were developed for Canadian first nation communities (Clark et al., 2002), in the Bronx, New York and for boroughs of London in the UK (Lienhardt, 2001). While associations between overcrowding and physical health have been well documented qual itatively including over 40 studies (UK Deputy Prime Minister Office, 2004), limited evidence exists on quantitative health effect estimates, particularly for Asian cities. The literature reviewed to date suggests an association between overcrowding and TB in both children and adults, with overcrowding in childhood affecting aspects of adult health.

PAGE 66

48 Table III 3 Available Health Effect Estimates for Airborne Infectious Disease Mortality. Health Hazards Healt h Effect Estimates Differential Risk (DR) by: Age (A), Gender (G), Ethnicity (E), Education (Ed), Income (I), Access to Healthcare (AH), & Seasonal (S) Crowded Household Sleeping areas (1) Antunes & Waldman 2001 ((Sao Paolo, Brazil): estimate a 14% incre ase in tuberculosis (TB) mortality over a five year period due to an average increase of one additional dweller per bedroom (over one person) in the household. (2) Lienhardt, 2001 (Bronx, New York): under age 5 children living in severely crowded areas we re about five times more likely to develop tuberculosis (adjusted for HIV status) than children living in areas with limited or no crowding. From 1982 to 1991, tuberculosis notification rates in London boroughs (United Kingdom) increased by 12 % for each % increase in the number of persons living in overcrowded accommodations (3) Clark, Riben et al. 2002 (Canadian first nation communities): Morbidity rate was 18.9 per 100, 000 in communities with average of 0.4 0.6 persons per room (ppr), while communities with 1.0 1.2 ppr had a rate of 113.0 p er 100 000. In crease of 0.1 ppr in a community was associated with a 40% increase in risk of 2 TB cases occurring, while an increase of $10 000 in community household income associated with 0.25 the risk, and being an isolated community increased risk by 2.5 times. (4) Baker, 2008 (New Zealand): for every 1% increase in the average crowding level of a Census Area Unit (CAU) there would be a 5% increase in the expected TB count. (1) A; I; Other: foreigners / migrants significantly higher mortality risk than persons born in Sao Paolo (88.57% increase) (2) A; G; E; Ed; I; & AH (3) I, E (4) A, I, E, Other: migrants (the total TB cases notified in the 2000 2004 period were 1898. About 60% of these cases occurred in mi grants born in high incidence countries). Transport Crashes : Road fatalities are considered to be an emerging epidemic, and are listed in the top ten causes of global mortality (GBD, 2011). Recently, there's been a trend towards the term crashes' replac ing accidents' in the transport literature and is therefore the term used in the remainder of this paper. WHO (2004) divides the main risk factors for road fatalities into the following major categories: Factors influencing exposure to risk (e.g. economic factors, land use planning practices which influence the length of a trip or travel mode choice, high speed motorized traffic mixed with vulnerable road users, insufficient attention to road layout and design, etc)

PAGE 67

49 Factors influencing crash involvement ( e.g., inappropriate or excessive speed, presence of alcohol, medicinal, or recreational drugs; fatigue; being a young male; travelling in darkness; defects in road design, layout and maintenance which can also lead to unsafe road user behavior; vehicle fac tors such as braking, handling and maintenance; inadequate visibility due to environmental factors, etc), Factors influencing crash severity (e.g. inappropriate or excessive speed; seat belts and child restrains not worn; roadside objects not crash protec tive; etc) Factors influencing severity of post crash injuries (e.g., lack of appropriate pre ho spital care, appropriate care once in hospital emergency rooms, diffic ulty evacuating people, hazardous materials leakage fire presence from collision, etc) Fo r two wheeler vehicle users, one major risk factor identified is crash h elmets not being worn by users; for transport fatalities, gender and age play a large role (Arnett, 2002; De Hartog, 2010; Fischbeck et al., 2012). In first row of table X, the risk fa ctor of mode choice is analyzed with respect to avoiding exposure to risk by shifting from road use to rail use. The health effect estimates due to mode choice are developed using the National Crime Records Bureau (NCRB) transport fatalities 2008 data for Delhi India. Table III 4 Available Health Effect Estimates for Road Safety. Health Hazards Health Effect Estimates Differential Risk Mode Choice / Road Safety (1) NCRB, 2008 (Delhi): From Auto to Rail: 9.7 deaths / 1% mode shift; From 2 Wheeelers to Rail: 20.6 deaths / 1% shift; 5 deaths / 1% reduction in ped fatality risk via shift to rail (1) DR not addressed Lack of Helmets (2) Ichikawa, 2003 (Municipality of Khon Kaen Province, Thailand): After enforce ment of the helmet act, helmet wearers increased five fold; head injuries decreased by 41.4% and deaths by 20.8%. (2) by helmet wearing Lack of Seatbelts (3) Mohan, 2004 (India): Use of seat belts, child seats and airbag equipped cars can reduce car occup ant fatalities by over 30% (3) A & G ; by seatbelts Lack of Seatbelts (4 ) WHO, 2009 GHR p.33 (Global): Seatbelts, when used correctly, estimated to reduce risk of death in crash by 61% (5) by seatbelts

PAGE 68

50 Table III .5 (cont'd.) Lighting (5 ) Elvik, 2006 (meta analysis): Using daytime running lights (DRLs) on cars shows that such use reduces the number of multi party daytime accidents by about 10 15% for cars using DRLs; Similar results for DRLs by motorcyclists in Serembam, Malaysia (Radin 2003) (4) DR not addressed Vehicle Crashes (6) Novoa et al., 2009 (Spain): Graduated licensing system (31% road traffic injury reduction); vehicle electronic stability control system (2 to 4% reduction); area wide traffic calming (0 to 20% reduction); a nd speed cameras (7 to 30% reduction) (6) DR not addressed Diabetes : The major infrastructure related health effect estimates reported from the literature for this infrastructure related health risk category includes levels of physical activity, obesity, and diet. Most studies identified to date, are not Delhi or South Asia specific and further epidemiological study is therefore needed. Table III 5 Available Health Effect Estimates for Diabetes Health Haz ards Epidemiological Health Effect Estimates Differential Risk by: PM2.5 (1) Pearson, 2010 (US): short term exposure effect study from 2004 2005 finding a 1% increase in diabetes prevalence seen with a 10 g/m3 increase in PM2.5 exposure; comparing US co unties with highest (annual mean concentration quartile of 13.8 17.1 ug/m3; EPA limit is 15ug/m3) compared to lowest levels (annual mean concentration quartile of 2.5 7.7ug/m3) of PM2.5 exposure: resulted in a >20% increase in diabetes prevalence (1) A, E I, Ed, Obesity rates (BMI>30kg/m^2; physical activity), population and fast food establishment density, health insurance Inadequate Physical Activity (2) De Hartog, 2010 (Hu et al, 2004 Finland): 25% reduction in diabetes mortality from walking/cycling to work (unadjusted for other domains of phy act); (Matthews, 2007 Shanghai) 21% 34% reduction in diabetes mortality for Chinese women cohort (adjusted for other phy act) (2) A, G ; cyclists vs. car commuters with tradeoffs in DR by exposure to air poll ution, traffic accidents, higher physical activity Obesity (3) (WHO, 2000 (Obesity: Preventing & Managing the GlobalEpidemic): Diabetes Mellitus: ~64% of male and 74% of female cases of diabetes could have been prevented if no one had had a BMI over 25 (3) A, G, I, genetic susceptibility Diet (4) Carter et al., 2010 (UK): Summary estimates showed that greater intake of green leafy vegetables was associated with a 14% reduction in risk of type 2 diabetes. (4) Only A, G, E, and Ed

PAGE 69

51 Waterborne Poll ution, Pathogens and Health Risks : The major infrastructure related mortality risk factors reported from the literature for water and sanitation related health outcome s (e.g. diarrhea, malaria, and malnutri tion related mortality) include water quality (and removal of waterborne disease pathogens), sanitation, water supply, household water storage, drainage, management of solid waste and specific thresholds of en vironmental pollution exposures HEEs hav e been a challenge to identify for many of these risk f actors ( not all ) due to the existence of many other confounding risk factors (see blue text box on right). Additional risk factors studied include house design, wastewater treatment, and landfill management (a more specific aspect of solid waste managemen t). A study by Van Poppel et al. 1997 has identified several other risk factors specific to infant and child mortality in the Dutch municipality of Tilburg. While water supply was not found to have an impact on the mortality of children, their study prese nts A Key Challenge to Developing this Evidence Base: Addressing Confounding Risk Factors in Water Sanitation Studies : Inadeqaute control of confounding variables is a major problem in all be a fe w of the [water and sanitation related health impact] studies The factors most likely to confound results in studies of water or santitation are housing (pathogen survival), crowding (hand to hand contamination), age (acquired immunity), sex (activity and contact with environment), breastfeeding (exposure to pathogens and pathogen viability), health education (use of services), seasonality (availability of other water sources), rural urban (exposure to pathogens), migration (exposure to new or more pathoge ns), income (better food and medical care), diet (quality and quantity), other infections (e.g. malaria causing poor nutrition status), and distance to medical care (whether proper attention is sought) (Esrey and Habicht 1985:73 74). Completely controlling for the large number of confounding variables that might influence the various health indicators is an impossible task in historical research[in addition], many studies assume that the presence of a particular water supply is synonymous with use of that facility" Van Poppel et al. 1997

PAGE 70

52 interesting findings (per their relative risk table below) on risk factors such as sex of child, mother's age at childbirth, child's order of birth in family (e.g. firstborn child), and season of birth. As shown, male children, children with higher bir th order, children who were quickly followed by the birth of another child, children born in the summer period, and children born to very young and old mothers had higher relative mortality risks: Figure III. 1 Example Relat ive Risk Findings for Infant and Child Mortality

PAGE 71

53 Table III 6 Available Health Effect Estimates for Waterborne Disease Pathogens Health Hazards Epidemiological Health Effect Estimates Differential Risk (DR) by: Age (A), Gender (G), Ethnicity (E), Education (Ed), Income (I ), Seasonal (S) etc Water Quality (1) Cairncross, 2010 (meta analysis): Diarrhea mo rbidity risk reductions of 17% associated with improved water quality (with water supply from pu blic sourc e) see Tbl 41.7; 63% diarrhea risk reduction associated with house piped water connection (Tbl 41.7) (4a) Esrey et al. 1991 (adapted by Cairncross, 2006): Water Quality Only Intervention: 15% median reduction in diarrheal morbidity (6) Fewtrell et al. 200 5 (synthesis of five urban sutides): Water treatment at point of use (incl. chemical treatment, boiling, pasteurization, and solar disinfection) produced a relative risk of 0.74 (0.65 0.85) HARMONIZATION : 15% to 26% reduction in relative risk of diarrhea l morbidity Sanitation (1) Cairncross, 2010(meta analysis): improved exreta disposal : 36% reduction in d iarrhea morbidity Water Supply (2) Cairncross, 2006: a 50% reduction expected from excellent water supply alone (3) Van Poppel et al., 1997: child mortality RR of 1.01 for piped water versus other (relatively no impact) (1) A (children) (2) A, G (3) A, G, I, S, socioeconomic group, birth order of child, dwelling type (6) A; E; urban / rural Water, Sanitation and Hygiene (WASH) (4 a) Esrey et al. 1991 (adapted by Cairncross, 2006): Water quantity only intervention: 20% median reduction in diarrheal morbidity; Water quantity & quality intervention: 17% median reduction in diarrheal morbidity; Sanitation Only: 36% reduction; Water & Sanitation: 30% overall & 55% reduction in child mortality; Hygiene Promotion: 33% reduction (4 b) Esrey et al. 1985: R eview of 67 studies on water supply & sanitation interventions / their impact on diarrheal diseases. For < 5 years of age, improved water supply & sanita tion demonstrated a median reduction of 27% and 30% in morbidity and mortality rates, respectively, when studies of better quality were used. (5 ) Gunther, 2011 (Global): average mortality reduction achievable by investment in full household coverage with water and sanitation infrastructure is 25 deaths per 1000 children born across countries (6) Fewtrell and Kaufmann, 2005 (Meta analysis of 46 studies): water, sanitation, and hygiene interventions had a similar degree of impact on diarrheal illness, with r elative risk estimates from the overall meta analyses ranging between 0.63 and 0.75; (7) Waddington et al. 2009 (Meta analysis of 71 studies): water supply: RR of 0.73 to 0.75; water quality: RR of 0.65 1.09; Sanitation: RR of 0.68 to 0.78; Hygiene: 0.53 to 0.68; Multiple interventions: 0.67 (8) Gunther and Fink (2010): water and sanitation infrastructure lowers the odds of children suffer ing from diarrhea by 7 17% and reduces mortality risk for children under age five by about 5 20 %(findings based on rev iew of 172 DHS reports from 70 countries; all are short term effects studies); Models used: Ordinary least squares / logit regression model for child diarrhea; Weibull survival model for under five mortality HARMONIZATION : Having water and sanitation inf rastructure results in a 5% to 30% reduction in the relative risk of under five diarrheal mortality (4a) A; G; Ed (literacy); breastfed or not; (4 b) A (5) A; Ed; G; Proxy of Income; Marital Status; Household Size; Urban & Rural; birth order / spacing; (6) A; Ed; urban / rural (7) A; Ed; urban / rural; seasonal (8) A (e.g. age of mother); Ed; I; household size; marital status; urban / rural; child vaccinations

PAGE 72

54 Table III .6 (cont'd.) Unsafe Wat S an (9) WHO, 2009 GHR (Global): Unsafe water a nd sanitation e stimated to cause 88% of diarrheal deaths Water Supply (8) Gunther and Fink (2010): Private acce ss to water supply decreases relative likelihood of diarrhea by ~ 14%, whereas the odds of diarrhea are reduced by onl y about 7% if the household uses a public pipe and/or a public well/borehole. (10 ) White, Bradley, and White, 1972: Typhoid fever: 80% reduction expected from excellent water supply; Trachoma: 60% reduction expected from excellent water supply; Scabies: 8 0% reduction expected from excellent water supply (11) Esrey and Habitch, 1986: "among the illiterate group an infant was 1.36 times more likely to die if no piped water was available compared with having piped water; while among the literate group the rel ative odds of an infrant dying was more than doubled if no piped water was available" Sanitation (12) Messou et al. 1997 (Ivory Coast; rural setting): Diarrhea Mortality Risk Reduction: 85% (13) Esrey et al. (1991) found water supply and sanitation reduc ed prevalence of ascariasis by a median of 28 percent (range 0 to 83 percent) and of hookworm infection by 4 percent (0 to 100 percent). Those reductions are likely caused by sanitation rather than by the water supply improvements. Indeed, three of the nin e positive studies of ascariasis and three of the five positive studies of hookworm involved sanitation alone. It is also likely the effect of excreta disposal on Trichuris infection is similar to that on ascariasis (Henry 1981). (14) Esrey (1996): Diarrhe a reduction of 13 44% for flush toilets and 8.5% for latrines (findings based on eight Demographic and Health Surveys from Bolivia, Burundi, Ghana, Guatemala, Morocco, Sri Lanka, Togo, Uganda); Limitation: selected only 8 out of the 63 DHS surveys that wer e available in 1995 (8) 13% reduction in diarrhea for having flush toilet vs. open defecation; for households that share toilets with several other households: no significant impact of access to a public latrine/flush toilet on child diarrhea, and the effe ct on child mortality is only 3.3%; but private sanitation facilities reduce the odds of diarrhea by 10% and the likelihood of d ying before the age of 5 by 13% (findings from 172 DHS studies) Managem ent of Solid Waste (15) Fobil et al. 2011 (Accra, Ghana ): On basis of "population and waste generation" sub component, strong evidence of a difference in risk of urban malaria mortality between least deteriorated zone (mean fraction = 0.029, 95% CI: 0.011 0.045) and moderately deteriorated zone (mean fraction = 0.051, 95% CI: 0.04 0.05 9) with a relative mortality [RR] = 1.77 between the two zones; also has e stimates for modifying housing design / poor water supply/sanitation Water insect vectors (16) Insecticide treated bed nets (Glennon et al. Sub Saharan Africa): Mortality RR: 0.85, 95% CI .76 .89 (7) I; A (8) See above (9) A; Ed; In; vitamin and mineral deficiencies (10) Population density; Climatic conditions; (11) A, I, Ed (literacy) (12) A (13/14) A, G, I, Ed (15) House type; waste generated; private or public facilities (16) Having insecticide treated bed nets

PAGE 73

55 Additional systematic literature reviews by WHO at a global scale indicate deaths from diarrhea of children aged less than 5 years represent approximately 19% of total child deaths (Boschi Pinto, 2008). Such estimates are useful for quantifying future health scenarios at city scale if the only data available is under 5 mortality counts with cause not specified (e.g. DHS / NFHS 3 data in India and Delhi). Additional relaven t literature for developing a databas e of infrastructure related health effects estimates includes developing estimates for population as a whole, not just under 5 diarrheal mortality. A synthesis by Walker and Black, 2010 highlights the need for improved diarrhea specific mortality and morbi dity data for different age groups including older children and adults (as deaths for this age group can also be quite high). Lamberti et al. 2012 also identifies the need for improved understanding of diarrhea duration and severity across age groups due t o significant global diarrhea morbidity, and improved treatments leading to decreases in diarrhea mortality. A study by Boschi Pinto (2008) indicates deaths from diarrhea of children aged less than 5 represent approximately 19% of total ch ild deaths (Bosch i Pinto, 2008). For the case of Delhi, India, analyses of death records for 2006 indicate the percentage of diarrheal mortality relative to total deaths reported by cause for under 5 populations was 0.6% (likely indicating underreporting) and for adult pop ulations was 0.1% (NCT of Delhi Annual Births and Deaths Report, 2006). Although not addressed in the table above of health effect estimates, the Central Pollution Control Board of India, the US Environmnetal Protection Agency, WHO, and others also provide drinking water quality guidelines for hundreds of contaminants with water meeting these standards safe to drink, and water unfit being harmful to health.

PAGE 74

56 Cancer: I nfrastructure related health effect estimates for cancer mortality are shown in the Table below Our review found health effect estimates mostly available for lung cancer as it relates to both indoor and outdoor air pollution. In U.S. counties with hazardous waste sites and ground water pollution, as identified by the Environmental Protection A gency, estimates were also made for gastrointestinal cancers. Other cancer risks include radon, asbestos, and othe r hazardous chemical exposures. Findings from primarily European studies suggest diet, specifically fruit and vegetable intake, and physical a ctivity can help reduce the risk of cancer. Table III 7 Available Health Effect Estimates for Cancer Health Hazards Epidemiological Health Effect Estimates Differential Risk PM2.5 (1) Krewski et al. 2009 ( US nationwide, NYC and LA): A 10ug/m^3 change in PM2.5 exposure levels: (Study Period 1979 1983):Lung Cancer mortality increases by 0.09%; (Study Period: 1999 2000): Lung cancer mortality increases by 0.14% (2) Laden et al 2006 (Six US Cities): Lung cance r mortality increases by 0.27 % per 10ug/m^3 change in PM2.5 exposure (3) Pope et al 2002 (US): Each 10 g/m3 elevation in PM2.5 was associated with approximately a 4%, 6%, & 8% increased risk of all cause, cardiopulmonary, &lung cancer mortality, respect ively HARMONIZATION : Min: 0.14% to Max: 8% increased risk of lung cancer mortality per 10ug/m3 change in PM2.5 exposure (1) E; I; Ed are addressed. AH is not addressed. (2) TBD (3) TBD SO2 (1) Krewski et al. Minimal effect: A 10ug/m3 change in SO2 conce ntrations over 15 yr time window as modified by education: (Less than High School): Lung Cancer: 0.02%; (High School or More): Lung Cancer: 0.05% (1) See above Solid Fuels for Cooking (4) Gupta 2001 (Chandigarh, India): Cumulative exposure of > 45 age women to indoor air pollution from use of coal or wood for cooking or heating showed 0.43% incr ease in developing lung cancer (v ery few subjects were employed in high risk occupations ) (4) A; smoking vs. non smoking; occupations; rural vs. urban Radon Ex posure (5) Pavia et al., 2003 (meta analysis of 17 case control studies): Residential exposure to radon at 150 Bq/m3 increases risk of lung cancer by 0.24% (5) G; I; time spent at home Other chemicals (e.g. pesticides; petrochemi cals) (6) Parent et al., 2009 (Montreal, Canada): Evidence of a two fold excess risk of prostate cancer among farmers with substantial exposure to pesticides [odds ratio (OR)=2.3, 95% confidence interval (CI) 1.1 5.1], as compared to unexposed farmers. Also some increased risks a mong farmers ever exposed to diesel engine emissions (OR=5.7, 95% CI 1.2 26.5). (7) Liu et al., 2008 (Taiwan): significantly higher risk of developing brain cancer for those living in municipalities characterized by highest vs low levels of petrochemical a ir pollution (OR = 1.65, 95% CI = 1.00 2.73) (6) A, G, E, Ed, I, occupation histories and exposures (7) A,G, municipalities characterized by high and low levels of exposure

PAGE 75

57 Table III .7 (cont'd.) While the majority of these studies are focused on Europe and North America, o ne rural versus urban study identified in Chandigarh, India presents important lung cancer evidence for reducing indoor air pollution exposures of women age 45 and older that are due to the use of solid fuels for cooking. It was noted in this study that r esidence in urban areas did not entail an increased risk for developing lung cancer (Gupta, 2001) Discussion: Implications for Infrastructure Related Health Interventions To the best of our knowledge, knowledge gaps remain for applying HEEs in Delhi, In dia specifically associated with access to urgent healthcare, overcrowding, and urban water supply and sanitation Cancer and environmental health effect studies are also primarily Europe focused. The table below summarizes relevant HEEs for Delhi and urb an health knowledge gaps for Delhi and more generally, for Indian cities. Diet (8) Key, 2011 (Europe): For cancers of the oral cavity, pharynx and larynx, Freedman et al (2008a) reported that, compared with people who consumed ~1.5 portions of fruit and vegetables each day, people with intakes of ~5.8 portions per day had a relative risk of 0.71 (95% CI 0.55 0.92), an d in a study in Europe of squamous cell cancers of the oral cavity, pharynx, larynx (and oesophagus), Boeing et al (2006) reported that people with a high intake of fruit and vegetables (~7.7 portions per day) had a relative risk of 0.60 (95% CI 0.37 0.99) compared with those with a relatively low intake (~2.5 portions per day) (9) WHO GBD, 2009: Insufficient intake of fruit and vegetables (prevalence measure of 5 servings / day) is esti mated to cause ~ 14% of gastrointestinal can cer deaths, ~11% of ischa emic heart disease deaths and ~ 9% of stroke deaths worldwide. (8) G, fruit and vegatable intake; smokers vs non smokers; obesity; alcohol intake; (9) G, I (PDF P.46) Physical Activity (10 ) Schnohr et al., 2006 (Copenhagen): Adjusted relative risks for cancer for moderate activity 0.77: and high activity: 0.73; and for all cause mortality, moderate: 0.78 and high: 0.75 for both sexes combined. (9) G ; physical activity levels

PAGE 76

58 Table III 8 Study Key Findings and Identified Knowledge Gaps Interventions Addressing... In Exposure Concentration Health Effect Estimate (HEE) HEE Reference (Study Location) Outdoor Air Quality & CVD Particulate Matter (PM): 300 g/m3 0.43% increase in CVD mortality per 10 g/m3 change in PM10 Cropper, 1997 (Delhi, India) Outdoor Air Quality & Respiratory PM: 300 g/m3 0.31% in crease in respiratory mortality associated with a change in PM10 level of 10 g/m3 Cropper, 1997 (Delhi, India) Airborne Infectious Disease N/A Sao Paolo, Brazil: a 14% increase in tuberculosis (TB) mortality with an average increase of one additional dw eller in the household. Antunes & Waldman, 2001 (Sao Paolo, Brazil) Traffic Accidents 17% mode shift of auto & two wheelers to rail From Auto to Rail: 9.7 deaths / 1% mode shift; From 2 wheeler to Rail: 20.6 deaths / 1% shift NCRB, 2008 (Delhi, India) Water and Sanitation N/A Global: having water & sanitation results in a 5% to 30% reduction in relative risk of <5 diarrheal mortality (no urban HEE identified) Cancer Indoor air pollution from use of coal or wood for cooking or heating Exposure of > 45 age women to IAP showed a 0.43% increase in developing lung cancer. Gupta, 2001 (Chandigarh, India) Access to Urgent Healthcare N/A * Refers to a knowledge gap where quantitative health effect estimates are unknown for Delhi, India. This review o f HEEs addresses some, but not all of the key elements in the WHO framework model shown below: e.g. environmental and socioeconomic conditions in shaping differential exposure, individual susceptibility, and health effects. However, this review can help in form first order computations of health risk reduction benefits integrating equity concerns by exploring health effects due to inequalities in conditions and services.

PAGE 77

59 Figure III. 2 WHO Framework Model on Social Inequalities and Environmental Risks Taking into consideration this WHO model and the literature described in this study, the conceptual diagram below is developed for future study of urban infrastructure, environment and health. More specifically, the diagram demons trates how exploring both biophysical system characteristics (e.g. infrastructure and environmental conditions) and human social system characteristics (e.g. socio economic and biological) can be useful as both of which are known to have important implicat ions for health outcomes. Exploring the linkages in this diagram betw een infrastructure, environment climate and health more specifically, between exposures to hazards (left) health effect estimates (middle) and health outcomes (right) can be useful t owards assessing key risk factors to inform quantitative decision making on alternative health risk reduction scenarios for cities.

PAGE 78

60 Figure III. 3 A preliminary schematic representation / conceptual diagram for study of the nexus of urban infrastructures, environment and public health In the above conceptual diagram, this review has zoomed in on the health effect estimates (HEEs), and explored linkages to the other boxes in three ways : (1) literature review of quantitative mortality studies of infrastructure and environmental conditions that identify differential exposures to hazards such as air pollution, and differential susceptibility to health effects and final mortality outcomes (2) preliminary harmonization of HEEs f rom identified studies addressing risk factors and health effects in terms of mortality outcomes; and (3) discuss ion of challenges to developing HEEs (e.g. addressing confounding factors) and identification of knowledge gaps. The efforts in this study hel p develop preliminary HEEs primarily for biophysical (e.g. infrastructure and environment related ) interventions with a somewhat smaller

PAGE 79

61 focus on social system interventions relevant to Delhi, India However, by identifying differential exposure and risks to vulnerable subpopulations as related to socioeconomic and biological risk factors where possible important features in the above conceptual diagram for an improved health effects evidence base in Asian cities. Future development of this diagram could explore behavior changes, coping strategies, and response capacity as social system elements. The below revised version is shown to help distinguish the biophysical (in green boxes) and human social (in red boxes) system characteristics. Figure III. 4 A revised schematic diagram for studying infrastructures, environment climate, and health with consideration of biophysical and socio biological risk factors

PAGE 80

62 Conclusion By reviewing quantitative epidemiology studies on civil i nfrastructure, pollution, and seven related health risk categories, this study develops an initial database of infrastructure related health effect estimates and identifies areas where further inquiry and field study of infrastructure and environment relat ed h ealth impacts are still needed. Identified knowledge gaps include the associations between urban mortality outcomes and inadequate access to quality urgent healthcare services overcrowded housing, and water supply/sanitation Methods for filling such knowledge gaps through epidemiologic al and community based study could be beneficial so that additional infrastructure related health improvement scenarios can be assessed for supporting decision making and prioritization of actions toward healthier Asian cities.

PAGE 81

63 Appendix. Health Determinants of a Population and To Which Factors Health Effect Estimates are Most Sensitive Scientists generally recognize five health determinants of a population: Genes and biology (e.g. sex and age) Health behavior (e.g. alc ohol use, injection drug use, unprotected sex, smoking) Social environment or social characteristics (e.g. income, gender, discrimination) Physical environment or total ecology (e.g. where a person lives, basic infrastructure services, and crowding conditi ons) Health services or medical care (e.g. access to quality health care and having or not having insurance / ability to pay for care) The figure below presents estimates of how each of the se five major determina nts influence population health (Tarlov, 199 9). As shown, genes / biology and health behaviors make up 25% of population health, with the remainder of health most sensitive to social factors including social environment, physical environment and health services. Figure 1. (By Tarlov, 1999) As an example from this chapter's literature review, the case of outdoor air pollution exposure related premature mortality in Delhi demonstrates how women and elderly are most sensitive based on HEEs in the study by Rajarathnam et al, 2010.

PAGE 82

64 CHAPTER IV. IV. UNMASKI NG THE ROLE OF MULTIPLE INADEQUATE BASIC INFRASTRUCTURES IN INFLUENCING UNDER FIVE MORTALITY IN ALL INDIA, URBAN INDIA & DELHI Abstract Prior studies typically have shown infrastructure provision to have small to modest improvements to under five mortali ty rates (U5MR) (e.g., a 17% reduction in U5MR for a relative risk of 1.2), under the assumption of independence of each infrastructure provision from the other and without accounting for confounding effects of wealth and literacy. This study focuses on Al l India, Urban India and Delhi to explore importance as well as interactions among social factors (wealth and literacy), infrastructural factors, i.e., the provision of multiple basic infrastructures (piped water, sanitation, clean cooking fuels, etc.), an d proxy attributes that represent access to urgent care health (ATH) facilities. Key findings include: 1) all response variables highly correlated with each other requiring large datasets to control for socioeconomic status (SES) and literacy to unmask the role of infrastructure; 2) the partial literacy subgroup is particularly sensitive to lack of infrastructure with relative risks exceeding 4.5 for unimproved versus improved conditions; and 3) multiple infrastructure provisions together when controlling f or confounders has larger relative risks compared to single infrastructures. Results indicate 4.9 [95% CI: 1.2, 19.8 ] and 8.6 [95% CI: 1.2, 64.6 ] times higher U5MR for All India and Urban India, respectively, for limited literacy low SES populations with a bsence versus presence of toilets / taps on premises and clean fuels

PAGE 83

65 Introduction This study focuses on All India, Urban India and Delhi to explore importance as well as interactions among several factors influencing under age five mortality rates (U5MR ) in India and Indian cities, including: social factors (socioeconomic status (SES) and literacy), infrastructural factors, i.e., the provision of multiple basic infrastructures (piped water, sanitation, clean cooking fuels, etc.), and proxy attributes tha t represent access to urgent care health (ATH) facilities. Mortality is selected as the response variable over other health indicators (e.g. morbidity) due to its' policy relevance, critical importance as a measure of societal well being (Sen, 1998), and t he availability of under age five mortality data at city scale. Under Five years' mortality rate (U5MR), i.e., deaths of children aged 5 years or less reported over the past 5 years per 1000 live births over the same period is also an important human devel opment indicator recognized globally in the UN Millennium Development Goals (UN, 2013) U5MR in India is important because India is home to 20% of the world's and three quarters of South Asia's under five population respectively (UNICEF, 2008) and has his torically reported U5MR in the bottom 50 countries, globally (World Bank, 2013). While significant progress has been made over the past decade to reduce U5MR in India by about 25% (from 88 to 61 under 5 deaths per 1000 live births from 2000 to 2010, respec tively) the current 2010 U5MR of 61/1000 remains high with one child in every 16 dying prior to age five and the U5MRs still ten times higher than high income OECD countries (61.0 vs. 5.6) (World Bank, 2013) Thus, understanding complex interactions that shape under 5 mortality (U5M) is particularly important for India, and by extension the world. Understanding urban U5MR in India is particularly important given that urban

PAGE 84

66 U5MR is declining slower than rural U5MR (Claeson, 2000), and, urbanization is lea ding to rapid growth of slums in many Indian and Asian cities (Agarwal and Taneja, 2005) where inequalities in socioeconomic conditions, infrastructure provision and public health services are high. More than ten cities (e.g. Mumbai, Hyderabad, Agra, Delhi ) among the more than fifty cities in India with population over one million continue to report a high % population (>30%) living in slum areas that do not have many basic infrastructure provisions for example, 43% of slum households in India lack access to drinking water taps, 34% lack access to latrines and 33% are using solid fuels for cooking (Chandramouli, 2013). Slum residents in India are among the lowest by socioeconomic status (SES) (Gupta et al., 2009) and report significant proportions of the p opulation (~39%) who are illiterate or partially literate (able to read only parts of a sentence') (National Family Health Survey, 2006). Further, one third of the world's urban population currently lives in infrastructure deprived slum conditions, 66% of which are in cities of Asia and Africa where slum populations are projected to increase annually at 2.2% and 4.5%, respectively (UN Habitat, 2010; UN Habitat, 2006). Consequently, an improved understanding of how under age five mortality (U5M) is shaped b y the provision of multiple infrastructures (or lack thereof) in urban India and the interaction of infrastructure provision with social factors delineated in large Demographic Health Survey s (DHS) can contribute significant insights important for improv ing health and human development in other cities of the developing world Prior studies exploring child health in India and the South Asia region have considered the wealth health' link (Mohanty, 2009) urban v ersus rural health comparisons (Bharati et al., 2008) and the role of water and sanitation (Mollah and

PAGE 85

67 Aramaki, 2010) in shaping U5M In addition, prior studies of urban areas specifically using the DHS data have considered single infrastructures such as percentage of households having improved so urce of drinking water, improved sanitation, electricity and cooking gas (Goli et al., 2011) o r slum versus non slum housing (Govt. of India, 2010) ; and a few have also looked at mul tiple infrastructure variables (Das and Gautam, 2013) However, these stud ies have not identified or explored the confounding effects of socioeconomic status literacy, and access to healthcare. Most studies have typically assumed infrastructure provision to be an independent variable and consequently report relative risks (RR, i.e., the ratio of U5MR in unimproved versus improved conditions) of 1.7 for urban poor versus non poor (Agarwal and Srivasava, 2009) and up to 1.2 for having water and sanitation versus not (Gunther and Fink, 2010). Such studies suggest small to modest im provements to U5MR (e.g., a 17% reduction in U5MR for a RR of 1.2) can occur through providing basic infrastructures, under the assumption of independence of each infrastructure provision from the other. A more recent study suggests that the providing mult iple infrastructures together may be more beneficial having synergistic multiplier effects (e.g., (Fink et al., 2011), yet few studies directly examine the improvements in U5MR from multiple basic infrastructure provisions while controlling for key confo unders such as socioeconomic status (SES) and literacy. To address these gaps, the three key objectives of this paper are as follows: (1) Conduct factor analyses to extract key variables affecting the incidence of under five years' mortality (U5M) as reported in India and Urban India DHS data.

PAGE 86

68 (2) To explore associations of under 5 years' mortality rates represented by U5MR (under age five mortality rates expressed as deaths per 1000 live births ) as the response variable with three ca tegories of predictor variab les, while controlling for confounding interactions among: Socioeconomic factors (wealth and literacy levels); Proxy attributes of access to healthcare (A TH) conditions; and I nfrastructure related factors (housing, water, toilets, and fuels for cooking); (3) Developing relative risk (RR) computations representing the mortality rates in unimproved versus improved condition, for various combinations of basic infrastructure provisions (water, sanitation, and cooking fuels), while controlling for wealth and liter acy (as ATH found not to be significant). We explore improvements arising from simply the presence and absence of various combinations of basic infrastructures and then the quality of basic infrastructures Literature Review In this study, two forms of lit erature review are conducted. The first approach looks only at prior studies of under five mortality using the Demographic Health Survey. The second approach reviews a small set of infrastructure environment health studies external to DHS that may be relev ant to this study. Importantly, these studies do not explore U5MRs associated with groups of multiple infrastructure conditions (including housing, drinking water, toilet facility, cooking fuels), socioeconomic conditions (wealth and literacy), proxy attri butes of access to healthcare conditions (affordability, distance / transport to healthcare facilities, facilities being open when needed) while also separating out confounding health determinants at both city and national levels. The two types of

PAGE 87

69 studies reviewed are described below with some providing valuable insights for the methods in this study and ideas for future research and analyses using DHS datasets. DHS Focused Literature Review: Since October 2003, 1265 peer reviewed journal articles have been published including focus on 68 different countries, with research based entirely or primarily on DHS data. For India, this includes 397 papers. Per T able 1 below, several studies have looked at socioeconomic conditions and health care utilization in India (e.g. Mohanty, 2009 Bhargava, 2011 ) while few have considered inadequate civil infrastructure provision, access to healthcare, and under five mortality The most pertinent studies to this paper include Agarwal (2009) which e xplores tuberculosis outcomes with overcrowded sleeping conditions, Das et al (2012) which consider s household living conditions such as housing, toilet facility, drinking water, cooking fuels and its effect on child morbidity (including diarrhea, fever, and cough) in Madhya Pradesh and the Ministry of Health (2009) study that considers infrastructure conditions assoc iated with socioeconomic status, separately the overall U5MRs across eight cities, and nutritional status associated with slum v ersu s non sl um areas Goli et al. 2011 exp lores disease conditions, immunization coverage, and separately the deficits in basic amenities within selected Indian cities, but does not assess under five mortality associated with infrastructure and a ccess to healthcare conditions (e.g. health facility open when needed, affordability, distance / transportation) within the cities, or across All India The Gunther et al. study also aims to comprehensively address water and sanitation across countries using DHS and they observe the highest positive effect of such infrastructure on children's health in urban areas The variables Gunther et al. control for include mother's age, child's order of birth, and ownership of durable goods as a proxy of

PAGE 88

70 household's long term income. However, they do not control for levels of literacy. One additional noteworthy infrastructure and health quantitative analysis using DHS India data was a recent review of tuberculosis (TB) outcomes relative to living space density. Agarwal (2013) finds almost 50% of population in the poo rest wealth quartile are living in houses where 5 persons or more share one sleeping room, with TB incidence up to twice as high in homes with 5 persons or more per seeping room versus only 4 persons per sleeping room (Incidence of 423 versus 268 TB cases per 100,000 population). Similarly, TB incidence for those homes without separate cooking space can be 2.2 times higher than those that have separate cooking space (494 versus 223 TB cases per 100,000 population). An extension of this prior DHS India re an alyses is shown in our results focusing on diarrheal incidence associated with members per household and water and sanitation conditions, to compare with findings in studies by Gunther and Fink. To the best of our knowledge, no article in this first approa ch for literature review has looked at under five mortality associated with multiples infrastructures (including housing, drinking water, toilet facility, cooking fuels), socioeconomic conditions (including wealth and literacy), proxy attributes of access to healthcare and under five mortality rates while separating out these confounding health determinants Reviewing a Small Set of Relevant Literature External to the Use of DHS: Examples exist from the WHO Department of Public Health and the Environment t hat aim to quantify global disease burden (e.g. Ezzati et al., 2008) and national disease burden (e.g. Pruss Ustun et al., 2008) that can be prevented by environmental interventions. Pruss Ustun (2007) also identifies many environmental intervention areas for health, some related to the built environment e.g. housing (effects on respiratory

PAGE 89

71 infections) and road design (on road traffic injuries). However, quantitative analyses linking these multiple infrastructure factors and mortality in specific cities a re in most cases not presented. Other literature external to the DHS has addressed the health effects of water and sanitation. To name three of many studies, Van Poppel et al. 1997 identifies several risk factors and confounders specific to child mortality in a Dutch municipality, Mollah, 2006 reviews water, sanitation and hygiene literature primarily focusing on rural studies, and Boschi Pinto, 2008 conducts cross country reviews addressing child mortality due to diarhhea. Based on this review, the contri butions of this study are to explore U5M associated with wealth and literacy conditions, multiple inadequate household infrastructures and access to healthcare in All India, Urban India and Delhi, India. Methods Analyses for All India (n=51,555 households) Urban India (n=19,483 households), and Delhi (n=1,150 households) DHS 2005/6 survey data are conducted to explore the extent to which current U5M outcomes are shaped by multiple inf rastructures while controlling for key confounding social factors includ ing SES and literacy As background, the DHS is a national survey administered every five to six years to collect accurate, nationally representative data on health and population in India ; and is used here for its comprehensive nature allowing for compari sons of health, infrastructure, s ocioeconomic conditions, and in that the DHS i s required and done frequently. DHS Survey Data : Figure 1 presents a summary of data availability and some initial DHS variables of interest as preliminary aggregated data analy ses conducted for Delhi. While these initial explorations of slum versus non slum areas with a breakdown

PAGE 90

72 of presence of basic conditions and under five / infant mortality rates help depict the potential of comparisons between such areas, these analyses do not isolate impact of infrastructure inadequacy from other factors such as literacy and wealth that also affects health outcomes. Strengths and Weaknesses of Study Data Analyses: A key data limitation is that only child mortality data are explored as full population mortality data was unavailable. While additional analyses of mortality versus access to services in Delhi slum versus non slum areas are possible using DHS data (as shown in Figure 1), they are not in the scope of this study. For example, Figur e 1 depicts for slum and non slum households a 10% variation in terms of access to health insurance and a 28% variation in terms of having a bank account or formal education, but both health insurance and bank services data is not further explored in terms of under five or infant mortality. Methods: Five steps were taken to reorganize and analyze the DHS data : 1. Data Re organization: the categorical data were recoded to consistently represent worst to best conditions (lowest to highest number), and cleaned for non responses; 2. Identification of Predictor Variables and their interactions: The response variable was identified as U5MR per 1000 births in past five years computed as Equation 1 below: U 5 MR = ( IncidenceofUnder 5 Deaths pastfiveyears ) ( IncidenceofLiveBirths pastfiveyears ) 1000 Equation IV 1 Computing Under Five Mortality Rates (U5MR) The three categories of predictor variables selected for analyses include: o S ocioeconomic conditions (wealth and literacy) F irst, correlations with these socioeconomic conditions are explored after which they are controlled for;

PAGE 91

73 o M ultiple infrastructures (housing, water, toile ts, and cooking fuels); and the o P roxy attribute of access to healthcare (A T H) including costs, transportation challenges to reach healthcare, and households responding health facility is n ot open when needed as indicated by the mother as reasons they did or did not deliver at a health facility Pearson Correlations are then used to identify interactions between predictor variables indicative of multiple collinearity, and between predictor and response variables. 3. Exploratory Factor Analysis and Principal Component Analysis: were then used to identify variables of interest, their importance, and to reduce the number of predictor variables for further computation of relative risk. 4. Relative Ri sk (RR) was computed using Equation 2: RR = U 5 MR ( PopulationWithUnimprovedConditions ) U 5 MR ( CorrespondingBestInfrastructureCondition ) Equation IV 2 Computing Relative Risk of Under Five Mortality Rates U5MR were computed for unimproved versus improved infrastructure conditions using two approac hes. First we consider only presence (versus absence) of combinations of three infrastructures water supply on premises, toilet facility on premises, and clean cooking fuels and categorize in the form of Good: having all three improved conditions' vs. Medium: having two of the three improved conditions' vs. Poor: having one or less of the three improved conditions' to explore how just providing multiple infrastructures together reduces U5MRs and relative risk, without addressing the quality of the inf rastructure provided. Second, we consider aggregate measures that include presence / absence of basic infrastructure pro visions including house type, water supply, toilet facility type and cooking fuels along with their quality when present (e.g., havin g pucca

PAGE 92

74 (solid physical structure) home coded as "high quality" vs. semi pucca vs. kaccha "worse quality"; having taps on site coded as "high quality" vs. up to 30 minutes away vs. more than 30 minutes away "worse quality"; having flush toilet as "high qua lity" vs. pit toilet vs. no toilet facility; and having electricity, LPG, biogas for cooking fuel as "high quality vs. kerosene vs. solid fuels including wood, charcoal, agricultural crop, animal dung as "worse quality"). Together this yields three catego ries as Bad (25% or less of high quality infrastructure condition)' vs. Average' vs. Good (i.e. > 75% high quality infrastructure condition) '. Because the U5MR are found to be numerically relatively small although still of public health concern i.e., of the order of 2% to 5% (e.g. 20 to 50 deaths per 1000 births), the probability of deaths are relatively low, and therefore the odds ratio (probability of death / probability of surviving = 2% / 98%) is numeri cally similar to U5MR. The log 10 of U5MR will therefore be similar to the LOGIT model for child survival to age five. Thus, although only U5MR is shown in this paper, similar relationships are expected for child survival as well. 5. Comparing relative risks : The RR were computed while controlling for h ighly correlated socioeconomic factors (multi collinearity) and for the infrastructure combinations previously listed in Step 4 and these were compared with the conventional case where RR are typically computed assuming each single infrastructure was indep endent of the other and of the socioeconomic variables. It is important to note that other confounding risk factors may also play a role in shaping under five mortality such as the birth order of children, as indicated in studies

PAGE 93

75 such as in Van Poppel et al., 1997 and suggested in Gunther and Fink, 2010. Such data were not recorded in the India DHS and hence could not be studied. Results Exploring Variable Interactions: Correlation s in the All India Data : Pearson correlations of Under 5 years of age morta lity incidence (U5MI) and the nine predictor variables, as well as the U5MR (mortality per 1000 live births) versus the percentage occurrence of the same nine predictor variables were assessed as shown in Table 2 and Table 3 for the All India dataset, resp ectively. In these analysies, a forward causal model with respect to U5MI and U5MR and the various predictor variables is expected. Table 2 tabulates correlation coefficients across U5MI and 9 predictor variables for ~51,556 households; note that for n 1 000 pairs of data, any r 0.05 can be considered statistically significant (see Clark, 2009). As is shown in Table 2, all pairs of predictor variables are significantly correlated with each other (in many cases with r> 0.58), which indicates the predictor variables are highly correlated, and hence not independent. For example, wealth of households (low to high SES) was highly correlated with incidences of literacy (r=0.5 8 ), water supply (r=0.60), toilet fac ility (r=0.74) and clean cooking fuel (r=0. 62 ). T he incidence of under five mortality (U5MI; whether a death occurred or not) was observed to be weakly cor related with wealth (r = 0.07), literacy (0.0 7) ; incidences of water supply ( 0.0 6), toilet facility on premises (r= 0.0 7) quality cooking fuel (r=0.07 ) ; and ATH (r=0.06) reflected by the facility being o pen as noted in the maternal survey Correlation of U5MI with house type, shared toilet, and access to healthcare attributes of costs' and too far / no transport' were not statistically significant wit h r less than |0.05| The same is seen in Table 3, where the data

PAGE 94

76 representing incidences are aggregated by the five wealth quintiles, yielding n=5. In Table 2, mortality incidences are aggregated within each of the five wealth quintiles and divided by the number of live births within each quintile to yield the response variable U5MRs computed for each wealth quintile. Likewise the % occurrence of the other predictor variables (best condition) is also computed for the homes in the 5 wealth q uintiles, as s hown in Table 3. The percentage occurrence of all predictor variables (best condition) are shown to be highly correlated with wealth, and with each other, (any r 0.67 representing significance for n = 5 (see (Clark, 2009)). The mortality rates likewise a re highly correlated with wealth and pairwise with the % occurrence of each of the predictor variables. Both Table 2 and Table 3 show that the predictor variables are not independent. Similar results were seen for the Delhi data (s hown in Appendix ). This means generalized linear models or other regression models assuming independent predictor variables should not be employed Tables 2 and 3 indicate that the relationship of the response variable U5MR should be explored with a smaller sub set of key predic tor variables, identified in principal component analysis (PCA) described next, and by controlling for variable interactions to compute relative risk for different conditions of interest. Principal Component Analysis / Exploratory Factor Analysis conducted with the nine predictor variables revealed that five key variables including water supply, toilet facility, wealth, literacy, and cooking fuels explain 75% of the total variance. Using Kaiser's criteria [21; 22], eigenvalues over 1 are considered stab le (wealth, water and sanitation) and component 4 (literacy) and 5 (cooking fuels) are just under 1, so may also be significant. Thus relationships of U5MR were explored with these predictor variables.

PAGE 95

77 In addition, we also explored the relationship of U5MR with ATH represented by facility not being open when needed (as reported in the maternal survey), which was observed to not be significant in Table 4. Computing RR and Controlling for Confounding Variables: We first demonstrate in analyses that infrastru cture deficiences are common in All Indian cities, and improve with SES as shown in Figure 2 (along with a summary of the datasets used from within the DHS child recode dataset). We then explored RR of U5MRs for confounders, SES and then literacy conditio ns as shown in Figure 3 By considering U5MR correlations with these conditions for India, Urban India, and Delhi, analyses show a relative risk of ~2 for both wealth and literacy from lowest to highest (i.e. each of these figures show that U5MR ranges fro m about 40 to 20 deaths per 1000 total births for worst to best conditions, respectively). Therefore, these factors are controlled for in the remaining analyses. Controlling for wealth and literacy to explore the impact of infrastructure provision is illus trated for All India and Urban India in Figure 4 (with Delhi in the Appendix) A ccess to healthcare conditions represented by facility not being open at time of delivery in the maternal survey is also explored as a potentially important factor This factor is identified as a proxy attribute for access to healthcare conditions, with circumstances relating to facility not being open for mother s during delivery also potentially prevent ing children from accessing health care during cri tical health events. See T able 4 Analyses show this variable to be insignificant with a relative risk of 1.0 or 1.1 for low SES groups in All India and Urban India. A larger relative risk exceeding 2.5 for high SES groups in All India, Urban India, and Delhi; and a relative risk o f 10 was

PAGE 96

78 found for the Delhi high SES populat ion; but this was not a consistent trend across population groups as was the case for wealth and literacy. For the Delhi low SES population, the ATH relative risk could not be evaluated due to the small n of res ponses for this attribute in the Delhi dataset. Following these analyses we then moved to infrastructure conditions focusing on presence and absence as well as provision and quality of infrastructures. Figure 5 and Table 5 shows the high importance of pr oviding multiple high quality infrastructure provisions in India and Urban India, which becomes visible specifically among the partially literate population in the lowest socioeconomic status group (Presence vs. Absence: India RR= 4.9; 95% CI: 1.2, 19.8] a nd Urban India RR = 8.6 [95% CI: 1.2, 64.6 ] ; Provision and Quality: India RR=5.6; 95% CI: 2.2, 14.2 and Urban India RR=14.5; 95% CI: 2.9, 70.9, respectively). For the same low SES subgroup and fully literate and illiterate population, the provision of thes e multiple infrastructures makes much less difference (RR = 1 to 1.2, shown in Table 5) likely because fully literate populations may be able to make other adjustments in face of poor infrastructure, e.g., handwashing benefits can be realized even without toilet facilities. In contrast, among illiterate populations, the lack of awareness of prevention strategies and or other best practice medical recommendation (e.g., oral rehydration) could mask the benefits of infrastructure provision. However, t hese resu lts highlight a key insight that the sensitivity of health outcomes (U5MR) to infrastructure provision can be masked by wealth and literacy. Among the poorer households with limited or partial literacy (i.e. respondents able to read and write only parts of a sentence') the significant benefits of providing multiple infrastructure provision becomes apparent as indicated by the high RR, indicating U5MR

PAGE 97

79 to be roughly >500% worse with unimproved infrastructure than with improvements, i.e., provision of infr astructure can reduce U5MR by > 80%. To verify these findings, analyses were repeated for the All India data set considering high SES and various literacy conditions and similar results were obtained. The analyses were repeated with only neonatal deaths, infant death and U5MR ND (less neonatal deaths; at least 30 to 50% of neonatal deaths in developing countries have been shown to be from infections with the most common causes being diarrhea, pneumonia, tetanus and sepsis ((Moss et al., 2002), (Stoll, 1997 ), and Bang et al., 1999), which can be related to infrastructure issues. The infant, neonatal, and U5M split for the conditions explored are also shown as Figure 7. The finding of high RR for the lack versus the provision of infrastructure are seen consis tently in partial literacy, low SES conditions (for NMR, IMR, U5MR and U5MR ND), and are absent in the high SES group. All these results are shown in Table 6. We also repeated U5MR analysis for Delhi and found similar results as shown in Appendix. The pauc ity of data made the analysis more challenging, as it was not possible to compare with the best case (which had zero incidences of U5M). However, we still see two times higher U5MR (RR=2) comparing the worst case with the average case when controlling for ATH, wealth, and literacy. The U5MR ND computations for All India in Table 6 also show that high RR exists for certain combinations e.g. across all literacy conditions, with partial literacy remaining the most sensitive. Such results suggest U5MR ND as another useful indicator for assessing infrastructure combinations, as removing neonatal deaths excludes factors that may be unrelated to infrastructure that can shape neonatal mortality (e.g. preterm

PAGE 98

80 birth and asphyxia which accounted for ~50% of neonat al deaths in 2000 in a global study of 193 countries (Lawn et al. 2006). The relative risk computations in Figure 3 also demonstrate why c ontrolling for factors like literacy when possible, in addition to wealth, is important for exploring U5MR associatio ns with inadequate infrastructure condition combinations (e.g. of housing, water, toilets, and fuels). As shown, the importance of basic infrastructure is masked by the importance of both wealth and literacy (which is closely related to SES). Plots are pre sented in Figure 4 to bring together some of t he analyses conducted and highlight U5MR sensitivity for the partially literate population to infrastructure provision and condition In this analysis, having three or more out of four high quality infrastructu re provisions is shown as Good: 75% High Quality Conditions', while one or less of high quality infrastructures is shown as Bad: 25% Low Quality Conditions'. To identify which infrastructure provision had the largest impact on U5MR within the partial literacy group, we repeated the analysis considering only the presence and absence of various combinations of key infrastructures identified as important in the EFA, which are water supply on premises, toilet on premises, and clean fuels for cooking. These results are shown in Figure 4 for India and Urban India (with Delhi in Appendix). For both All India and Delhi datasets shown in Figure 6a, all infrastructure combinations when controlling for confounding variables of SES and literacy demonstrate sign ificantly higher relative risks then previously reported in the literature. For example, U5MR increases by a factor of four or more (400% increase) for all infrastructure combinations. Fo r the All India poor and partially illiterate households

PAGE 99

81 Figure 6 in dicates the exceptionally high importance of having toilets and taps together (RR=7.0); as well as having toilets, taps, and clean fuels together (RR=5.9) Similar results are seen in the Delhi datasets too with the combination of toilets, taps, and fuels resulting in a RR of 12.4 after controlling for confounding social variables where possible (a focus on full literacy households with all wealth groups included due to small n). In contrast, when not controlling for any of these variables, RR results are in the range of 1.3 to 1.7, for each individual infrastructure condition shown in Figure 6b, in line with current observations in the literature. These results indicate controlling for multi collinearity unmasks the critical importance of providing multip le basic infrastructures demonstrated by examining the low SES, partially literate subgroup. Discussion Results from this analyses indicate over a 4 times higher U5MR for those low SES partial literate households without multiple high quality basic infra structure provisions, and roughly a 2 times higher U5MR for those households of lower socioeconomic status and literacy levels. In addition, preliminary exploration indicate up to a 75.9% reduction in under five mortality rates for All India may be achieva ble in low socioeconomic status households with limited literacy for those having versus lacking basic infrastructure provisions. Wit h United Nations Millenium Development Goal # 4 aiming to reduce by two thirds the under five mortality (between 1990 and 2015), strategies focused on multiple infrastructure and access to healthcare interventions can (presumably) be more successful then efforts focusing on improving socioeconomic conditions alone Future DHS surveys

PAGE 100

82 could help shed additional light on these associations by exploring the possibility of larger survey sample sizes for Delhi and other cities. In using DHS to quantify the benefits of multiple infrastructures, this analysi s present s one of the first to quantitatively separate out access to healthc are, literacy and socioeconomic factors from multiple infrastructure conditions and to link a variety of socio economic and infrastructural determinants of health specifi cally for U5M We emphasize that additional exploration of infrastructure health assoc iations and cau sal relationships through field work and epidemiological study is needed. The results of this study provide an initial quantitative exploration opening up a new line of inquiry on urban health improvements specifically of under five mortali ty through changes in infrastructure conditions and socio economic conditions. Controlling for confounding factors are important to improving understanding of how multiple infrastructure provisions shape child mortality.

PAGE 101

83 Tables and Figures Figure IV. 1 Initial Exploration of Infrastructures, Health/Banking/Education, and Infant / <5 Mortality Per 1000 births for Slum vs. Non Slum (DHS/NFHS 3)

PAGE 102

84 Figure IV. 2 Infrastructure deficiencies: comm on in all Indian cities, improve with SES Table IV 1 Review of Studies In Past Ten Years Using the DHS: 2002 2012 Author, Year, (study location) Title { Additional Description of Interest } Lauridsen, 2011 (India) Socio economic inequality of immunization coverage in India Pathak et al., 2011 (India) Malnutrition among children in India: Growing inequalities across different economic groups Mohanty, 2009 (India) Alternative wealth indices and health estima tes in India. { This study compares estimates for reproductive and child health services by two alternative wealth indices in India. } Bhan et al., 2012 (India) Are socioeconomic disparities in tobacco consumption increasing in India? A repeated cross sectional multilevel analysis { Tobacco Use by wealth, living environment & caste } Subramanian, 2006 (India) Indigenous health & socioeconomic statu s in India {assesses health inequalities between indigenous and non indigenous groups in India; and importance of socioeconomic status, regardless of indigeneity }

PAGE 103

85 Table IV .1 (cont'd.) Bharati, 2008 (India) Chronic Energy Deficiency Am ong Indian Women by Residential Status { Assesses population BMI, CED, and obesity; SES and nutritional status; and rural and urban household differences by different socio economic variables associated with outcomes of urban CED women vs rural CED women. S pecific cases of infrastructure condition are not included. } Mishra et al., 2005 (India) Cooking smoke and tobacco smoke as risk factors for stillbirth { assesses environmental health risk linkages to still births } Ghosh, 2003 (India) Demographic and soci oeconomic correlates of neonatal, post neonatal, and childhood mortality Bhargava, 2011 (India) Health care utilization, socioeconomic factors and child health in India { explores household living conditions in terms of food supply / consumption patterns & its effect on child morbidity } Das, 2012 (India) Household Environmental Factors & its Effects on Child Morbidity in Madhya Pradesh { explores household environmental factors housing, toilet facility, drinking water, cooking fuels, housing density & i ts effect on incidence of child diarrhea, fever, and cough } Ministry of Health, 2009 (India) Health & Living Conditions in 8 Indian Cities { Infrastructure conditions in slum versus non slum areas including housing type, residential crowding, drinking wa ter, toilet facility, solid fuels for cooking, means of transportation; overall U5MRs across 8 cities; nutritional status / anaemia by slum vs non slum areas; no results shown in study on under five mortality associated with infrastructure and access to he althcare conditions in cities, or across All India } Goli et al., 2011 (India) Living and health conditions of selected cities in India: Setting priorities for the National Urban Health Mission { Explores disease conditions fever, asthma, tuberculosis, ac ute respiratory disease and separately the deficits in basic amenities, including shelter, drinking water, cooking gas connections, improved sanitation and electricity for slum and non slum conditions in cities; no results shown in this study on under fiv e mortality associated with infrastructure and access to healthcare conditions specifically affordability and transportation availability within the cities, or across All India } Hazarika, 2010 (India) Women's Reproductive Health in Slum Populations in India: Evidence from NFHS 3 Agarwal, 2009 (India) Social determinants of children's health in urban areas in India { assesses demographic and social correlates of child health in urban areas including poverty, gender, caste status, religion, mother's educa tional attainment, parents' occupational status, women's autonomy in household, and overcrowded sleeping conditions and tuberculosis incidence } Gunther and Fink, 2010 Water, Sanitation, and Children's Health: Evidence from 172 DHS Surveys: Findings from 1 72 DHS studies across multiple countries on water and sanitation, including a 13% reduction in diarrhea for having flush toilet vs. open defecation; for households that share toilets with several other households: no significant impact of access to a publi c latrine/flush toilet on child diarrhea, and the effect on child mortality is only 3.3%; but private sanitation facilities reduce the odds of diarrhea by 10% and the likelihood of dying before the age of 5 by 13% Fink, Gunther, and Hill, 2011 The effect of water and sanition on child health: evidence from the demogaphic and health surveys 1986 2007: Re analyses of 171 surveys exploring the e ffects on child mortality from water and sanitation when controlling for parental characteristics including educati on and other household characteristics (e.g. no. of household members above age five and ownership of assets) Gunther, 2012 (Africa) Deadly Cities? Spatial Inequalities in Mortality in sub Saharan Africa. { assesses inequality, "urban mortality penalty" i.e.differences in child and adult (estimated) mortality between rural and urban areas and urban slums and urban formal settlements } Fotso et al., 2005 (Africa) SES in health research in developing countries: should we be focusing on households, communit ies, or both? { assesses inequality, mortality, nutrition tests a set of measures of SES for predicting health status in developing countries. SE indexes that capture both household and community attributes; applications to data from DHS fielded in five A frican countries (1990s).}

PAGE 104

86 Table IV 2 Exploring Correlations: All India Pearson's correlation (r) Matrix With U5M Incidence [ Note: a forward causal model (or one way relation) is expected with respect to U5MI and predictor variables ]

PAGE 105

87 Table IV 3 Correlation Matrix: Aggregate Analyses where U5MRs are computed for Wealth Quintiles (n = 5) and Plotted Against Changes in Other Variables Due to Wealth [Note: m atrix shows that the changes in Aggregate U5MR and in the % coverage for different variables are all also highly correlated with wealth (matrix assumes wealth as independent variable and is not indicating occurrences correlated, but rather indicates need f or controlling)] Table IV 4 Summary of U5MRs and RRs for a Proxy Attribute of Access to Healthcare [Note: when cell left as ', this i ndicates U5MR and RR could not be evaluated [due to no under five dea th incidences and small number of responses for this attribute in Delhi dataset i.e. facility not open (low SES n=8, high SES n=4)]

PAGE 106

88 Figure IV. 3 Example U5MRs by Wealth and Literacy from Lowest to Highest (India & Urban I ndia )

PAGE 107

89 Figure IV. 4 Flow Charts for India and Urban India: Controlling for Wealth and Literacy

PAGE 108

90 Figure IV. 5 India and Urban India Multiple Infrastructure Presence / Absence vs. U5MR [No te: Same results for Provisions and Quality of Conditions are shown in the Appendix]

PAGE 109

91 Figure IV. 6 a R R of Infrastructure Condition Combinations: India (top) v. Delhi (bottom) Figure IV.6b U5MRs and RRs of Individual I nfrastructures Assuming Independence

PAGE 110

92 Figure IV. 7 Breakdown of All India Neonatal, Infant, and Remaining Under 5 Mortality

PAGE 111

93 Table IV 5 Summary of All India U5MR, IMR, NMR, U5MR ND Analyses for Presence and Absence of Infrastructure Combinations Note: empty cells indicate mortality rates and RR could not be evaluated [due to no death incidences and small n of responses for this attribute]. For High SES, the partial literacy condition could not be evaluated for these same reasons and is therefore not included in above summary table.

PAGE 112

94 Appendix A. Delhi Analyses and All India Principal Component Analysis Table 1. Exploring Correlations: Delhi Matrix With U5M Incidence Analyses [ Note: a forward causal model (or one way relation) is expected with respect to U5MI and predictor variables ] Figure 1. U5MRs by Wealth and Literacy from Lowest to Highest (Delhi)

PAGE 113

95 Figure 2. Flow Chart Summary of Analyses and Findings for Delhi, I ndia Table 2 All India Principal Component Analysis: Variance Accounted For By Each Component Total Variance Explained Initial Eigenvalues Extraction Sums of Squared Loadings Component Total % of Variance Cumulative % Total % of Variance Cumulative % 1 Time to get to water source (minutes) 2.759 27.591 27.591 2.759 27.591 27.591 2 Type of toilet facility 1.805 18.047 45.637 1.805 18.047 45.637 3 Wealth Index 1.124 11.237 56.874 1.124 11.237 56.874 4 Literacy .988 9.880 66.754 5 Type of cooking fuel .924 9.237 75.991 6 House type (as defined in NFHS 2) .896 8.963 84.954 7 Reason didn't deliver at health facility: cost too much .774 7.737 92.690 8 Reason didn't deliver at hea lth facility: facility not open .552 5.515 98.205 9 Reason didn't deliver at health facility: too far / no transportation .121 1.805 100.000 These three factors e xplain 56.9% of the total variance in the items Extraction Method: Principal Component Analysis.

PAGE 114

96 Appendix B. All India DHS Analyses: Progressions by Wealth Qu intiles In the table below, analyses are shown that summarize some of the predictor and response variables considered across the five wealth quintiles for the case of All India. Table 1 Re Analyses of Households Showing Progression in Infrastructures, AT H conditions, U5MRs, and LOGIT by Wealth Quintiles for All India As shown, ~18% of population resides in the poorest wealth quintile of which ~85% are illiterate. For comparison, 21% of surveyed population resides in the richest wealth quintile of which 91% are having full literacy. U5MRs per 1000 live births are also shown for Low vs High SES populations as 40.6 vs. 20.2, for a relative risk of 2.

PAGE 115

97 Appendix C. Diarrheal Incidence Analyses Figure 1. Diarrheal Incidence Per 1000 Population vs. SES and Lit eracy in Delhi & India: Note: Diarrhea incidence is observed to be relatively even across All India wealth quintiles. If to compare Delhi vs. All India analyses by wealth and literacy, Delhi India by wealth (or SES) shows the largest difference in diarrhe al incidence per 1000 population.

PAGE 116

98 Appendix D Additional DHS Re Analyses for Six Indian Cities, DHS Urban vs. Rural India, and Delhi vs. All India Almost all improved conditions appear to steadily increase with sociaoeconomic status (SES), as shown belo w. This observation is tested through analyses of % of households having infrastructures and proxy attributes of access to healthcare (ATH) by wealth across six Indian cities (Delhi, Hyderabad, Indore, Kolkata, Mumbai, Chennai), Urban vs Rural India and De lhi vs. All India. Figure 1. Conditions by SES across Six Indian Cities: Average & Standard Deviation

PAGE 117

99 Figure 1. Comparisons of All India Urban vs. All India Rural by SES

PAGE 118

100 Figure 2. Comparisons of Delhi vs. All India by SES Appendix E FAQ : How are mortality rates related to infrastructure condition and what types of data would you need? In order to compute mortality rates associated with infrastructure conditions, there is a need for mortality, birth, and infrastructure data for the same populatio ns over a given time period. More specifically, data would be needed on the number of deaths per 1000 births in the past X years for households known to have presence or absence of improved infrastructure conditions during that period (e.g. water supply on premises, toilets on premises, and clean cooking fuels). Data for the same populations on potential confounding factors such as wealth and literacy that are important to control for would also be needed.

PAGE 119

101 CHAPTER V. V. INFRASTRUCTURE CONDITIONS AND SUSTAINABI LITY PRIORITIES IN ASIAN CITIES: A COMPARATIVE STUDY OF THREE NEIGHBORHOODS IN DELHI, INDIA Abstract For rapidly growing cities in Asia like Delhi, India, the planning, designing, and constructing of new engineered infrastructure systems are of significant importance to society as they affect health outcomes today and may still be in place for many years to come. More than one third of urban inhabitants in Asia and Delhi live in slums where populations are often at higher risk due to weaker structures, less safe locations, and the inability of infrastructure to withstand extreme weather if poorly designed / constructed This chapter assess es how current infrastructure and environmental conditions shape sustainable development priorities in Asian cities, spec ifically within two low income and one mid to high income locality in Delhi, India. Multiple infrastructure conditions are characterized and future priorities are assessed in ways that may offer new opportunities for focusing efforts and actions A key f inding was that households deprived of infrastructure provisions would prioritize that first over pollution or extreme weather events such as heat and drought, rain and flooding In addition, both low SES communities were different in their coverage of all infrastructures except cooking fuels. Finally, results indicate access to improved drainage as a higher priority for all The f indings demonstrate how infrastructure, environment, and extreme weather are af fecting human settlements today and how experienc e of risk often varies by locality

PAGE 120

102 Introduction: Why Asian Cities, Civil Infrastructures & Local Priorities Over half of humanity now lives in cities and in the next 20 years, nearly 60% of the world's people will be urban dwellers (UN, 2012). Trends sug gest urban populations are expected to soar to five billion from more than three billion today and this will include 60% of China's population living in cities (currently 46% urban), 41% of India's population (currently 30%), and 87% of US population (curr ently 82% urban) by 2030 (WUP, 2011). Meanwhile, almost one third of urban inhabitants, especially in Asia, currently live in slums (UN Habitat, 2011), where populations are often at higher risk due to weaker structures, less safe locations, and the inabil ity of infrastructure to withstand extreme weather if poorly designed / constructed (Costello, 2009). With lack of adequate infrastructure and high environmental pollution an everyday experience for some urban inhabitants, such conditions may be large moti vators for local sustainability priorities. In Delhi, India, for example, estimates suggest ~40 50% of the population are living in slums or slum like conditions, which the United Nations defines as households that lack access to improved water, sanitation sufficient living area, durability of housing, and security of tenure (UN, 2007). Furthermore, the National Capital Terri tory of Delhi, India which hosted a population of over 18 million in 2010 is projected to soon reach 24 million by 2021 and 28 milli on by 2026 (NCT, 2011). The estimated population of the North Capital Territory of Delhi, India is currently estimated at ~22 million. Between 1975 and 2005, Delhi grew from 8 million to 15 million, and represented the third fastest growing megacity in the world (just after Dhaka, Bangladesh and Lagos, Nigeria). Such rapid population growth creates huge new demands for civil infrastructures (i.e. e nergy s upply ; water s upply ; sanitation and w aste ; transportation and

PAGE 121

103 communications; building materials/h ousing ; f ood ; parks/open s paces ) and upgrades to existing infrastructures in ways that address multiple city goals: economy, environment, and public health. Currently, ~1000 new vehicles are added to roads in Delhi daily, 55% of households live within 500 meters of roads with high levels of air pollution (putting residents at risk of cardiac & respiratory problems), 16% of households lack access to drinking water taps (putting residents at risk of waterborne illnesses), 6% lack access to latrin es, and 8% use wood dung, and charcoal for cooking (MoUD, 2009). For rapidly growing Asian cities like Delhi the planning, designing, and constructing of new engineered infrastructure systems today are of significant importance to society as they affect health outcomes tod ay (Sperling & Ramaswami, 2013) and may still be in place for many years to come. Upgrading of existing infrastr uctures is also of importance (e.g., new wate r storage, building ventilation or cooling systems, transport safety, and flood ready drainage can all have implications for current local risks such as dehydration, heat related mortality, road crashes and infectious disease s often present in Asian cities (WHO, 2010; Rydin et al., 2012; Hales et al., 2003) At the same time, with over one thousand citi es worldwide engaged in infrastructure development as part of sustainability and climate action plans (ICLEI, 2011), and initiatives like the Asian Cities Climate Change Resilience Network (ACCCRN, 2013) ramping up it's efforts to reach over 100 cities acr oss Asia, it is important to understand that households and cities continue to care about diverse priorities, not only the abstract concept of greenhouse gas (GHG) mitigation or environmental sustainability. This study, therefore, assesses how basic infras tructures across seven sectors and environmental conditions shape local priorities in diverse areas.

PAGE 122

104 Infrastructure transitions in Asian cities are rapid. Infrastructure is providing both many direct and indirect benefits and risks (Sperling and Ramaswami, 2013). Benefits include access to employment, improved housing/living conditions, reduced sewage, etc that improve health these can all promote human development. At the same time, congestion and polluting infrastructures are creating several risks: air pollution within and outside homes, water pollution from inadequate infrastructure, and climate extremes. Poor access to transportation options may also contribute to poor access to health care. While cities may have various sustainability plans aimed a t GHG mitigation and climate adaptation, disadvantaged residents such as listed and unlisted slum dwellers may have more pressing priorities of just getting adequate and equitable access to basic infrastructures. Although many studies have been conducted o n slum and non slum households (Goli et al., 2011; Gunther, 2012), few to our knowledge have characterized inter slum variation and this was a motive for our study's objectives to compare two slum neighborhoods versus an average neighborhood. In this stu dy, we explore the proposition that sustainability priorities in neighborhoods of Delhi, India are predicated by the existing infrastructure conditions although the local air may be highly polluted, residents will likely have provision of basic infrastru cture as their main priority. As wealth increases, other factors may become priorities. Also, there are literatures on familiarity with risks (e.g., being used to extreme weather events such as heat and flooding) that can make residents more accustomed to such events in summer or monsoon seasons for example (Brebbia et al., 2005; Jiang et al., 2011; De Hoog, 2011; Kanti Paul, 2011). We emphasize this is a descriptive study

PAGE 123

105 with the objectives of comparing two slums versus an average neighorbood and that we are testing only the proposition that people without infrastructure may prioritize that first. Objectives The overall goal of this paper is to quantify inter slum variation in infrastructure provisions, and then assess how current infrastructure and envi ronmental conditions shape sustainable development priorities in Asian cities, using a representative megacity case study including three diverse localities within Delhi, India. This is achieved in two steps 1) Characterizing baseline infrastructure and environment conditions in different neighborhoods; 2) then exploring what are the local priorities in these diverse localities both of which may offer new opportunities for prioritizing efforts and actions especially within Asian cities The three loca lities include: 1) a low income, densely populated neighborhood with low literacy ra tes on interior (away from main road); 2) a low to middle income neighborhood with improved literacy rates, less density (on exterior closer to main road); and 3) one avera ge middle inco me neighborhood with much higher literacy rates. By improving understanding of existing localities e.g., what are the current infrastructure conditions, the priorities for potential neighborhood improvements, and how important current condi tions are for shaping priorities this paper provides a platform to discuss the potential for future sustainability and/or climate action planning at local levels. The following sections review existing surveys on priorities, assesses infrastructure cond itions across seven infrastructure sectors including energy supply; water supply; sanitation and waste; transportation; building materials/housing; food; and parks/open spaces (t he same that are studied in the recently developed infrastructure supply cha in

PAGE 124

106 GHG footpri nts by Chavez and Ramaswami, 2013; Ramaswami, 2013), and explores the potential for local priority driven processes for three different localities within an Asian megacity such as Delhi, India. Review of Surveys on Global to Local Priorities While many surveys exist on global to local priorities, this review addresses three types of surveys on sustainability priorities: one recently being implemented to address global priorities by the UN (2013) and two other specific surveys, one by Gallup (Y u and Pugliese, 2012) on environmental versus economic priorities and another on local priorities in China by the Chinese Academy of Social Sciences (CASS; Lo, 2010). Our review covers five major themes in these surveys, including: 1) geographic focus; 2) types of priorities covered related to local cultural, social, economic, public health or human development aspirations; 3) satisfaction levels; 4) differential priorities based on socio economic status; and 5) scale. Based on this review, we carry out our own survey that includes improvements in three areas: capturing priorities of localities with different socio economic status, measuring satisfaction levels with current conditions, and thereby assessing how current conditions shape local priorities. This approach aligns well with other recent literature: avoiding global distractions, pursuing local priorities" (McGranahan, 2007) and "developing sustainable and inclusive urban infrastructure services" (MoUD, 2011). Existing Surveys on Sustainability Prio rities: This section provides three useful examples of existing initiatives for surveying sustainability priorities in terms of most pressing problems. Table 1 outlines the survey structure and compares them. In the China focused survey by CASS (Lo, 2011), findings suggest local environmental issues

PAGE 125

107 such as air pollution and waste problems are highe r priority then climate change. T he global survey (UN, 2013) finds a good education better healthcare, better job opportunities, an honest and responsive govern ment, access to clean water and sanitation, affordable and nutritious food, protection against crime and violence, better transport and roads, and reliable energy at home to be highe r priority than climate change. In fact, the results to date for this UN M yWorld2015 survey suggest action on climate change as one of the lowest of all sixteen listed priorities in the survey, with the exception of respondents from countries listed as very high in the human development index (HDI). The Gallop survey, although n ot addressing climate change directly, finds that a large proportion and in some cases the majority of respondents in BRIC countries Brazil (83% of respondents), Russia (60%), India (45%) and China (57%), respectively prioritize protection of the environ ment even at the risk of curbing economic growth Compared to the urban infrastructure environment (UIE) survey utilized in this study, the existing surveys have yet to capture differences on priorities in terms of different socio economic strata (with th e exception of the UN survey which addresses low, medium, high, and very high HDI countries) and in some cases satisfaction levels with current conditions. Perhaps more importantly, the priorities looked at do not consider infrastructure, environmental po llution, and climate change all together. The UIE survey approach in this paper addresses these gaps, but also has its' own limitations for example, cultural, social, economic priorities are not considered in our study, yet they are in the UN study. The UIE survey methods and findings are described next.

PAGE 126

108 Table V. 1 Review of Small Sample of Example Surveys on Global and Local Priorities. Example Priorities Surveys My World 2015 Survey UN, 2013 Gallup Survey Yu & Pugliese, 2 012 CASS Survey Lo, 2010 UIE Survey Sperling & Ramaswami, 2013 Geographic Focus: Online Global Community Brazil, Russia, India, China (BRIC) China Delhi, India Priorities Covered: Freedom from discrimination; water & sanitation; climate change; protect ing environment; safety; support for people who can't work; transport and roads; education; affordable food; energy; responsive government; equality; better job opportunities; healthcare Economic Growth vs. Environment Greenhouse gas effect, climate change air pollution, waste, and others Upgrades across multiple infrastructure sectors; improving environmental conditions; and managing multiple extreme weather events such as cold, heat and drought, rain and flooding) Considers Satisfaction Levels With Curr ent Conditions: Not covered Covered (i.e. satisfaction with quality of air, quality of water, and efforts to preserve environment) Not covered Covered (i.e. satisfaction with current infrastructure services and related environmental conditions) Responses by gender (G), age (A), location (L), education (E), and socio economic status (SES) G, A, E, L, and HDI all covered L covered None G, A, E, L, & SE (i.e. priorities for lower vs. middle income households) all covered Scale: Global Only BRIC; China Rural vs Urban; Chinese Cities vs China Overall National Only Neighborhoods Within City Methods To assess how current infrastructure and environmental conditions shape sustainable development priorities in Asian cities, three diverse localities within the nort heast district of Delhi, India are studied. This study characterizes the different neighborhoods in terms of infrastructure and environmental conditions and then explores

PAGE 127

109 the local priorities in these diverse localities both of which may offer new opport unities for prioritizing efforts and actions especially within Asian cities Four methods are used: First, neighborhoods are characterized based on identified criteria (e.g. geography, socio economic conditions, infrastructures), transect walks, and initi al surveys broadly identifying the different conditions, exposures and risks; Second, detailed baseline assessment of the infrastructure conditions and levels of satisfaction with access to infrastructure services in the selected neighborhoods are documen ted. This includes all the infrastructure services supporting life in cities (Chavez and Ramaswami, 2013) drinking water supply, housing, electricity, gas, transportation, food, sanitation, drainage, open space and parks. In addition levels of inconveni ence experienced with current environmental conditions such as exposures to outdoor air and water pollution and severe weather events (e.g. extreme cold, heat / drought, rain / flooding) are assessed; Third a comparison of local priorities when given the options of upgrading infrastructure conditions versus improving environmental conditions versus improving management of extreme weather events (i.e. extreme cold, heat, flooding) are identified in each locality. These priorities are identified within the context of improving health and well being (e.g. livelihood and quality of life). Such baseline a ssessments and comparisons of priorities on infrastr ucture and environment can be useful features of current and futrue local priority driven processes for pla nning and policy decision making within cities and communities

PAGE 128

110 Study Area Selection : The localities selected for comparative analyses are shown below in Figure 2 Selection was based on four main criteria, a transect walk, and baseline condition survey F our main criteria are used for selection including : 1. Neighborhoods where access to healthcare may be an issue: the figure below depicts the area in pink, the NE district having least % of institutional deliveries (less than 60%), indicating the potentia l lack of access to healthcare facilities for such purposes. This area is where the three study neighborhoods are located. Figure V. 1 % of Births as Institutional Deliveries by Delhi Districts 2. Geography n eighborhoods within one district of Delhi, India : Neighborhoods within the Northeast (NE) District are selected. The NE is also selected as one of the most densely populated and underserved district s in Delhi, India in terms of infrastructures, access to healthcare, an d socioeconomic conditions. According to the

PAGE 129

111 District Level Househol d Survey (DLHS 2008 ) more than 40% of births in this district are occurring as non institutional deliveries and 16% of population illiterate of those age 7+. In addition, the NE district has a population density three times higher than that of the Capital and is often characterized by a lack of basic infrastructures including roads, drains, and waste disposal facilities (Census, 2011); with studies noting that the government has taken lim ited interest in developing th is area to date (Lalchandani, 2011). In addition, by observing online GoogleMaps, real time traffic data is largely available except for in the NE district A similar situation exists for available daily air quality measuremen ts the Delhi Pollution Control Committee website presents measurements throughout the city meanwhile, the closest monitoring station is Civil Lines ( quite far from the North East district on opposite side of the Yamuna River). This is shown in Figure 2 below. Figure V. 2 Northeast (NE) Delhi Study Neighborhoods on East side of Yamuna

PAGE 130

112 Figure V. 3 Closest Nearby Air Quality Station to NE District (West of Yamuna River) 3. Socio economic Conditi ons neighborhoods wit h different socio economic st atus and education/literacy levels: New Mustafabad (NM) was chosen as an underserved, low income neighborhood in Shahadra area of NE Delhi with lower levels of education, literacy, mostly informal jobs a nd ended up being predominantly Muslim; the nearby neighborhood of Brijpuri (BJ) was chosen as an improved but still low income neighborhood with many slum like conditions still present (was predominantly Hindu ) ; and Dilshad Gardens (DG) was chosen as a re presentation of a non slum average Delhi neighborhood DG households represent average middle to high income families and often highly educated (included diverse religious backgrounds i.e. Hindu, Muslim, Sikh). highly educated, and of diverse religious bac kgrounds (i.e. Hindu, Muslim, Sikh). Figure 3 below presents the average monthly expenditures for the three areas as one measure of socioeconomic condtions. In the past, studies have look ed at income and asset ownership to determine socioeconomic conditio ns (O'Donnell et al., 2008; Falkingham

PAGE 131

113 and Namzie 2001; Dinsa et al., 2012), with income a sensitive question that can result in untruthful or non responses. Instead of using asset ownership (i.e. Po and Subramanian, 2011), which may not be a good marker for wealth (e.g. televisions are often now in most homes and possessions do not always vary among different types of families), we focus on multiple expenditures incl uding those related to health that can be calibrated w ith other surveys ( e.g. consumer ex p enditure surveys (Govt. of NCT of Delhi, 2008)). Figure VI. 4 Socio Economic Factors By Study Area: Avg. Monthly Expenditures 4. Neighborhoods with different infrastructure conditions : Figure 3 below presents photo visuals of the three study areas and their conditions from unpaved roads, open sewage waterways, lack of waste facilities, and public hand pumps in NM; to both paved/unpaved roads, some greenery, motorbikes, water piped in taps, and partially covered drains in B J; to toy stores, cars, air conditioning units, parks, and so on in DG. For district level GIS maps showing the inadequacy of infrastructure conditions in the Northeast District relative to the other districts, see Appendix A. New Mustafabad Brijpuri Dilshad Gardens Figure V. 5 Study Neighborhoods Infrastructure Conditions Snapshot

PAGE 132

114 Results and Discussion The three study areas selected for assessment and comparisons are shown to be quite differen t in terms of the varying levels of infrastructure, environmental pollution, and extreme weather conditions faced Where comparisons of study areas can be made with the 2006 National Family Health Survey in terms of average Delhi non slum households and a verage slum households, we find that the study area of New Mustafabad (NM) is shown to be as bad or worse than average slum households. Brijpuri (BJ) is shown to be better off than slum households but worse off than the average Delhi non slum household ( 2006) and Dilshad Gardens (DG) is shown to often have conditions better or equal to the average Delhi non slum household. A more in depth view of these results is presented next. The results demonst rate a comparison of more than 115 households, ~48 in New Mustafabad and ~48 in Brijpuri, and ~20 in Dilshad Gardens, depending on surveys with non responses. The results are presented as four parts, in a similar manner to how they are described in methods. First, neighborhoods are selected and characterized base d on identified criteria, transect walks, and initial surveys to broady identify and compare the different current conditions, exposures, and risks. Second, a detailed baseline assessment compares the infrastructure conditions and levels of satisfaction wi th access to infrastructure services across the three study areas. Third satisfaction levels with current conditions and levels of inconvenience experienced with exposure to environmental pollution and extreme weather events are assessed Finally, compari sons are made on whether local priorities for upgrading infrastructure conditions, improving environmental conditions versus improving management of extreme weather events is most important

PAGE 133

115 toward broad goals of improving well being (i.e. health, livelihoo d, standard of living). These analyses in Figure 8 (results) also show an internal validity check to ensure response consistency between priorities and levels of satistisfaction / inconvenience 1. Preliminary Characterization of Differen tial Vulnerability & Risk In The Study Areas Upon selection of the three study areas, both desktop studies of conditions in the selected neighborhoods and field studies including transect walk and preliminary surveys were conducted. Transect walks were done by an engineer, architect, and doctor all of whom were involved in preliminary assessment of conditions and health risks in the three neighborhoods. The initial characterization tool used is summarized in the Appendix B After selection of neighborhoods and preliminary ch aracterization households were randomly sel ected for full survey participation (administered in Hindi ) to conduct baseline characterization or assessment of infrastructure and environmental conditions. 2. Comparative Assessment of Infrastructure Condition s in Three Neighborhoods Figures 5 and 6 below present baseline assessment comparing infrastructure conditions across seven infrastructure sectors: water supply; building materials / housing; energy supply; food; transportat ion; sanitation and waste ; and p arks/open spaces (as is studied in the recently developed infrastructure supply chain GHG footpri nts by Chavez and Ramaswami, 2013; Ramaswami, 2013) How neighborhoods compare with average Delhi slum and non slum conditions (assessed in the NFHS 3 survey i n 2006) is shown. Testing Statistical Significance Using T tests: While most conditions are shown in the figures to be different or similar for the two low SES neighborhoods, T tests to test significance (Gertsman, 2014) can be useful to determine if the data and results indicate the average infrastructure conditions to be similar (research hypothesis) or not (null

PAGE 134

116 hypothesis). To test the research hypothesis that NM and BJ have similar infrastructure conditions, and are different from DG we explore each o f the infrastructure factors characterized via survey responses by using statistical tests to determine if each condition is similar or different. Using the factor of traveling >30 minutes round trip to get water to test this hypothesis as an example, a t test is conducted at a significance level, p < 0.1 (based on internal error in surveys of ~6%). We find the average number of households who have to travel >30 mins for NM is statistically higher than for BJ (t=2.08, df = 91, p<0.1), so the null hypothesis (NM and BJ have different infrastructure conditions) is accepted in this case. Table 2 summarizes the data on which this t test is based, w here t must be > or = 1.664 to i ndicate statistical significance (Gershman, 2014). Table V. 2 Example Analyses on Time to Fetch Water and Motor Vehicle Ownership T Test #1: > 30 minutes round trip to fetch water (1= Yes; 2 = No) T Test #2 (NM & BJ) and #3 (BJ & DG) : Motor Vehicle Ownership (1=Yes; 2= No) Data Summary New Mustafabad (NM) Brij puri (BJ) New Mustafabad (NM) Brijpuri (BJ) Dilshad Gardens (DG) # of observations 47 46 47 46 20 Mean 1.7234 1.8913 1.89362 1.71739 1.3 Standard Deviation 0.45215 0.3147 0.31166 0.45524 0.47016 Variance 0.20009 0.096881 0.09506 0.20274 0.21 Test #1: t=2.08, df = 91, p< 0.1, where t > or = 1.664 indicate s statistically different Test #2: t=2.17, df = 91, p< 0.1, where t > or = 1.664 indicate s statistically different Test #3: t =3.35, df = 64, p<0.1, where t > or = 1.671 indicate s statistically differen t In Table 2, the hypothesis of having improved infrastructure conditions in Dilshad Gardens relative to both low SES areas on most accounts are also tested by comparing DG with BJ and NM for the survey results for having a motor vehicle. Results indicat e that DG does have statistically higher levels of vehicle ownership than BJ (t=3.35, df = 91), and that BJ has statistically higher vehicle ownership than NM (t=2.17, df=64). After Figures 6 and 7, all t test results are summarized below for p < 0.10 in Table 3.

PAGE 135

117 Figure V. 6 Summary of Results in Study Areas Compared with NFHS 3, 2006 Indicators Figure V. 7 Summary of Results from the Three Study Areas

PAGE 136

118 As a summary of the multiple stastical sig nificance t tests conducted for the different infrastructure factors of interest, see Table 3 below. Table V. 3 Summary of NM & BJ T Test Results for Key Infrastructure Conditions T Test Summary for p < 0.10 Piped Water Water >30mins Toilet Facility Clean Fuels Pucca' House Type Waste Disposal Parks Closed Drainage Computed t 4.5 2.08 2.00 1.14 1.74 5.96 2.08 2.82 Critical t value 1.664 1.664 1.664 1.664 1.664 1.664 1.664 1.664 BJ & NM: Statistically Similar (S) v. Differ ent (D) D D D S D D D D Table 3 results show NM and BJ were different in their coverage of almost all infrastructures including water, toilet, house type, waste disposal, access to parks, and drainage conditions (at p < 0.10 level of significance), with the exception of use of clean cooking fuels which was statistically similar T tests were also conducted showing BJ and NM to be lower than DG average conditions (e.g. for closed drainage test between DG and BJ, t=2.82, df=64, where t > or = 1.671 indicate s statistically different) and Delhi average non slum homes where such data is available (see Figure 6). Finally, after exploring levels of satisfaction / inconvenience (results in Appendix D), respondents are asked to "select three top priorities (i n order of first, second and third) for improving your health, livelihood, and well being." These results are shown next. 3. Comparing Priorities Findings on whether local priorities for upgrading infrastructure conditions, improving environmental condi tions versus improving management of extreme weather events is most important toward broad goals of improving well being (i.e. health, livelihood, standard of living ) are shown below in Figure 8. This figure presents the

PAGE 137

119 percentage of total votes (includin g votes for first, second, and third highest priorities) for specific improvements to managing infrastructures, environment and extreme weather events. The results show that the top infrastructure priorities (with over 20% votes) by area are as follows: Ne w Mustafabad (NM): water supply, waste, and roads / drainage Brijpuri (BJ): parks and open space, water supply, drainage, and electricity Dilshad Gardens (DG): none above 20%, yet DG did have infrastructure priorities > 10 % including for gas / cooking fue ls, electricity, water supply, and drainage. In addition, outdoor air pollution and managing extreme heat came up as higher priorities in DG than for NM and BJ, but still at lower levels (<10%). Drainage was rated high in all three communities (among othe r priorities) as >30% for BJ, >20% for NM, and >10% for DG residents. A key conclusion from these results is that households deprived of infrastructure provisions will prioritize that first over pollution or extreme weather conditions. In addition, e ven t h ough expenditures are relatively similar between the neighborhoods, results in this chapter demontrate household infr astructure conditions are widely different and that priorities are also widely different. Internal validation for the priorities results fr om survey response is addressed by computing the relative % error for the priorities response results relative to the survey responses on satisfaction levels (to ensure the same condition mentioned as priority is also having low levels of satisfaction per Appendix E). Given that the error is in range of 6%, and that households in Dilshad Gardens did prioritize climate related extreme heat close to 10%, it's possible this could be a current priority closer to 4% or 15% based on range of error. Overall, the findings show it to be

PAGE 138

120 invalid to assume from income, asset ownership, or expenditures what local priorities may be. Many other factors beside socioeconomic conditions can likely shape local priorities.

PAGE 139

121 Figure V. 8 Percent of Total Votes for Top Three Priorities Among Infrastructures, Environment, and Climate Related Extreme Weather Events by Study Area

PAGE 140

122 Conclusions The broad impacts of this study include two key analytical outputs that showcase methods for bottom up city a nd community based participatory research on characterizing infrastructure conditions and local priorities. First, this study quantified variation across multiple infrastructure, environment, and climate related extreme weather conditions. Then, local prio rities in terms of improved well being are analyzed and indicate that priorities are often based on deficits. In other words, areas that did not have infrastructures, air pollution and climate change were less important. This study characterizes infrastru cture and environmental conditions, exposures to hazards, and also household respondent's suggestions for improvements. A weakness of this study is the limited characterization of environmental pollution conditions, and future research could supplement cur rent findings comparing the three study neighborhoods through deployment of air (outdoor and indoor) and water (e.g. drinking water) quality sampling. Further comparative assessment of these neighborhoods to explore relevant socioeonomic and biological fac tors shaping health outcomes, specifically the experiences with and barriers to accessing healthcare, are conducted next in Chapter Six. We conclude based on the findings that for these Delhi neighborhoods where lack of adequate infrastructures and high le vels of environmental pollution are everyday experiences, strategies for improving current infrastructures are often high priority.

PAGE 141

123 Appendix A. Delhi District level Infrastructure Condition (Census India, 2011)

PAGE 142

124

PAGE 143

125 Appendix B. Preliminary Charact erization using Survey, Transects, and a Preliminary Descriptive Matrix of Physical Infrastructure and Social Infrastructure Table 1. Preliminary Comparisons in Survey of the Three Study Neighborhoods Comparative Assessment of Three Neighborhoods in a Meg acity: Key Indicators and Initial Characterization of Differential Exposure and Vulnerability to Risk Key Indicators by Study Area: New Mustafabad (NM) Brijpuri (BJ) Dilshad Gardens (DG) Population (1) Official; (2) Non Official 37250 Not Available Not A vailable No. of Households 7450 3025 Not Available Estimated dwelling units / acre (field observations and using Google Earth) > 85 >60 >35 Total Population (In Survey) 323 294 106 Total # of Households (In Survey) 42 48 20 Avg People / Household (Sur vey) 7.7 6.1 5.3 Table 2. Initial Descriptive Matrix Tool for Assessing Conditions in Areas Character izing Differential At Risk' Conditions based on Physical and Social Infrastructures Most At Risk Condition Moderately At Risk Condition Least At Risk Condition URBAN INFRASTRUCTURE SERVICES: I. Drinking Water Supply II. Electricity III. Gas IV. Toilet / Sanitation V. Drainage VI. Streets / Roads VII. Open Space & Parks I. Households without piped water into home II. Supply unreliable daily blackouts & houses cut off for not bei ng able to pay bills III. Solid fuels use (wood/dung) for cooking IV. Toilets and defecation in open and / or multiple households sharing one toilet V. No drains, and / or mostly open drains / sewage channels where drains often clogged VI. Unpaved roads V II. No green ery or access to public park or playground areas on neighborhood streets causing urban heat island effect Study Area/s: NM / BJ I. Some households without piped water into dwelling II. More reliable supply unaffordable / erratic bills ; have metered connections through paying landlord III. Most households using LPG, few using solid fuels IV. Common or in home toilets often to open drain V. Some open drains kuccha or pucca; narrow and cemented sewage channels; sometimes clogged drains VI. Both paved / unpaved roads VII. Parks and/or playground areas reachable, but often limited access; limited greenery and urban heat island Study Areas/s: BJ I. Most households having piped water and multiple taps II. Reliable supply metered individual el ectricity connections; still some complaints about high bills III. All households using LPG IV. Private flush toilets to sewer line V. Majority of area has underground drains VI. Paved Roads VII. Parks, greenery providing shade and open space nearby Stu dy Area/s: BJ / DG

PAGE 144

126 Table 2 (cont'd.) Table adapted by Sperling, Ramaswami, & Agarwal from Agarwal & Taneja, 200 5 Character istics of Differential Exposure & Vulnerability to Risk Most At Risk / Vulnerable Condition Moderately At Risk / Vulnerable Condition Least At Risk / Vulnerable Condition LAND STATUS & HOUSING: I. Security of Tenure II. House Physical Structure III. House/Sleeping Room Density IV. Natural Ventilation / Lighting V. Plinth Level ( housing surface relative to street surface; this is important for avoiding flood risk ) I. Some households protected against force d eviction from house or land lived on often unauthorized settlements, slums that are not recognized, or squatters on private or government land i.e. along railways II. Kuccha' house with weak structure; III. High housing density: > 7 ppl / home & > 5 p ersons / sleeping room; IV. Few with separate room for cooking / kitchen fan V. Limited to no ventilation / natural light VI. Low plinth level (on average 32 cm) despite high flooding risks Study Area/s: NM / BJ I. Most protected against forced eviction land / housing often belonging to local authorities or landlords; some having legalized land tenure via lease / rental agreement II. Semi pucca mud / brick walls &/or plastic / thatch roof III. Medium housing density: > 5 ppl / home & > 3 persons / slee ping room IV. Some with separate room for cooking and kitchen fan V. Limited ventilation / light VI. Higher plinth level (on average 47 cm) due to higher flooding risks Study Areas/s: BJ I. All households protected against forced eviction. Most own plot / house or have legal rental agreement. II. Pucca home with permanent structure III. Lower housing density: >4 ppl / home & on average 2.5 persons/ sleeping room IV. Most homes with separate room for cooking / kitchen fan V. Home with natural ventilation a nd lighting present VI. Plinth level a small concern with covered drainage in place (on average 36cm) Study Area/s: BJ / DG HEALTH STATUS I. Morbidity II. Service III. Access to Health Facilities EDUCATION & EMPLOYMENT I. Education of Children II. Education of Adults III. Inc ome Covers Needs IV. Type of Employment HEALTH I. Malnourished children seen; high incidence of water and foodborne illness; reported cases of child mortality; higher no. of reported hospitalizations II. Many children not immunized; home deliveries by untrain ed dais' (or midwives) III. None or limited public facilities within 2 3 km; population often visits quacks or stores EDUCATION & EMPLOYMENT: many children out of school and working; illiteracy in adult population; income not often enough to cover basic needs; irregular daily wage earners Study Area/s: NM HEALTH I. Some incidence of waterborne illnesses; lower no. of reported hospitalizations II. Irregular immunizations, majority institutional deliveries III. More access to public & private facilities; visit qualified doctors; use govt facilities for prolonged illnesses EDUCATION & EMPLOYMENT: most children go to school but with some drop outs; adults with functional literacy; income most often covers basic needs; daily wagers, more self employment (i.e as vendors, odd jobs) Study Area/s: BJ HEALTH I. Fewer incidence of illnesses; low no. of reported hospitalizations II. Majority having immunizations; majority of births are institutional deliveries EDUCATION & EMPLOYMENT: All children / most adults wi th formal education (finishing elementary education); income covers basic needs / some savings maintained; majority service class, some in government or private sector jobs Study Area/s: DJ EXTERNAL & INTERNAL SUPPORT Govt & non govt support ; Communit y Based Organizations (CBO) I. Limited government and non government program coverage II. Limited community based efforts Study Area/s: NM I. Presence of government and non government programs II. CBOs weak Study Area/s: BJ I. Higher levels of suppor t by government & NGO efforts II. Active CBOs i.e. Welfare Associations Study Area/s: DG GENDER STATUS Low gender status, incidence of domestic violence, limited choices on fertility / healthcare {not evaluated} Improved gender status, women have som e ownership over household health decisions {not evaluated} A more equitable gender status, some women are self employed and owning household decisions (not evaluated) IDENTITY STATUS Many witho u t documentation including proof of residence, ration or vote r ID cards Some with identity proof Study Areas: NM Most having identity proof Study Area/s: BJ / DG

PAGE 145

127 Appendix C. Prioritization Visual Used for Survey

PAGE 146

128 Appendix D Survey Visuals: Levels of Satisfaction and Inconvenience

PAGE 147

129 Appendix E Assessing Levels of Satistfaction and Inconvenience Experienced F indings on s atisfaction levels experienced with conditions in all seven infrastructure sectors and levels of inconv enience experienced with exposure to environmental pollution and extre me weather events are presented below in Figure 1. As shown, the comparative assessment includes average levels of satisfaction in the neighborhoods f or each infrastructure service on a scale of 1 ( very poor ) to 5 ( very good) as well as environmental conditions (including outdoor and indoor air pollution) and extreme weather condition which are both on scale of 1 ( minor inconvenience ) to 3 ( causing serious disruptions). Photo visuals sh owing the scale of 1 to 5 and the scale of 1 to 3 were utilized when conducting this surve y. These are shown in Appendix C and D Figure 1. Levels of Satisfaction and Inconvenience Experienced by Neighborhood

PAGE 148

130 Appendix F Initial Analyses Disc ussion by Individual Infrastructures The analyses shown below are based on initial pilot survey analyses (n=60) with ~20 households in each study area. Key figures can be revised to include full dataset in future. A. WATER SUPPLY: Water supply conditions acr oss the three neighborhoods are presented in the Figure 1 below. More specifically, the findings include: Percentage of households having piped water (top left): Dilshad Gardens (DG) with 100% coverage and Brijpuri (BJ) are both better off than the 200 6 Average Non Slum Household in Delhi, India (NFHS 3, 2006) However, New Mustafabad water supply conditions remains significantly below the conditions for the 2006 Average Slum Household in Delhi, India (NFHS 3, 2006) ; Avg. hours of daily water supply an d number of taps in home (top right) : Few survey respondents in New Mustafabad have taps within their household and many also mention water supply only being available for a few hours per day; Main sources of drinking water (middle left): pie charts presen t a breakdown of the water supply situation differences. As shown, only NM residents rely on government tanker truck water and over 80% rely on public water supply. DG has less reliance on piped water than BJ, with 43% of DG households opting for mineral w ater; Estimated time to get water as return trip if water is outside home and household w ater treatment practices (middle left) : ~ 40% of NM residents spend >30 minutes fetching water and less than 18 % of NM households treat their drinking water; Self r eported daily water consumption (bottom right): Gleick et al (2000) estimates human basic water requirements at 50 liters per person per day and the very small sample (n=10) of self reported guesses suggested less consumption in the three

PAGE 149

131 neighborhoods (on e fifth of that for the case of NM). While such results appear interesting, caution must be exercised in placing weight on these findings. In fact, the survey question leading to the responses shown was discarded early on in survey administration due to th e difficulty of validating self reported estimates of daily water consumption. Further study could attempt real time water availability monitoring, if such resources are available, which may prove useful for benchmarking and improving the water supply cond itions. In Table 1 below, qualitative responses to survey questions are presented for further assessing and comparing water supply conditions, in terms of a) how it has affected the household (HH) in terms of health and livelihoods, b) water supply factor s appreciated most, c) most commonly occurring water supply issues and current responses to such issues, and d) important changes that could be useful to address the current situation. It is worth noting that wh ile residents of DG also mention irregularity as an issue, they also have higher aspirations of constant supply. Figure 1. Preliminary comparative assessment findings for water supply conditions

PAGE 150

132 Note: responses to question on self reported water consumption estimates appeared inaccurate / difficult for respondents, so question removed after initial pilot. Figure 1. (cont'd.) Initial comparative assessment findings for water supply conditions

PAGE 151

133 Table 1. Qualitative Assessment of Water Supply Conditions: Initial Sample Responses Question s: New Mustafabad Brijpuri Dilshad Gardens A) How supply/ lacking services affects HH in terms of health and livelihood? NM1: Had jaundice (8 months back); NM2: When tanker water was not available stomach illness BP2: Had two episodes recently of 15 day s each where needed to fetch water from neighbor's borehold and carry for 50 100m DG1: missed work few times due to no water; DG3: recurring infection of stomach resulting in arriving late to office B) Water supply factors appreciated most? No responses B P2: tastes good and amount sufficient; BP10: quite happy with water supply no improvements necessary DG1: clean water; DG2: good quality; DG3: comes on time C) Most commonly occurring water supply issues? (responses to such issues)? NM1: don't have water when need it; availability issue as taps few in number and far away (response: water storage in containers) NM3: water is dirty (look for a different source) NM5: don't have own tubewell for drinking water (storage of water in containers) NM6: high costs of water NM8: irregular and dirty water (storage) NM9: water supply currently too far away (conserving / storing) NM10: dirty; irregular supply NM11: don't have water when needed; source is very far NM12/13: not regular / dirty (conserve and store) NM14: d irty water NM15: "no issues usually, but whenever strange color or smell, we don't drink it" BP1: don't have water when needed (conserving and storing water in containers) BP2: the water is dirty (when storing water, allow dirt to settle down) BP3: having water when needed (nothing) BP5: dirty water BP7: dirty; don't have water when needed BP8: irregular supply (storage in container) BP9: irregular, dirty, and high bills BP11: don't have when when needed; dirty (no community petition has been made in respo nse; just storing water in containers) BP12: supply timing is less than optimal (storage) BP13: irregular and dirty not due to leakage of drain water, but because of stagnancy DG3: high costs of water (alarm system cost of 600 Rs to detect overflow of water in tanks) DG4: high costs of water (conserving and storing water in containers) DG6: 3 to 4 days a month water is dirty DG7: dirty and high costs DG8: dirty and irregular DG9: dirty DG10: high costs DG11: dirty and high cost DG13: irregular, sometim es dirty, high $ DG18: often dirty D) Important changes to address the current situation? NM1: pipelines / public taps in every Gali (block) NM2: water supply taps in Gali ; drinking water is main issue; more covered water storage; currently no household filter NM3: increased no.of taps in gali NM5: no tap on road or in gali / doesn't like to go to neighbor's house to get water (this is current situation) NM6: reduce water bill NM7: government water supply should be available NM8: increased timing of supp ly and cleaner water NM9: more supply and closer NM10: should be available within house premises NM11: govt tanker truck is quite far but most HHs nearby go for this water which is said to be better quality; we get piped water across nala in Brijpuri (als o far); source should be nearer NM13: regular water daily; should come twice daily both regularly and on time NM14: wants water to be cleaner NM16: bill should be regular and less; more clean water NM17: should be cheaper NM18: piped water should be th ere BP1: more taps; subsizided rates for personal connection; installment systems for payment BP3: more regular supply, more clean water BP5: provision of clean water BP7: regular and timely water supply; bill should be less and "proper" BP8: more time for water supply BP9: regular supply, less bills BP11: timely and clean supply; repair of pipeline so there's improved pressure for water to reach 2nd / 3rd floor (where brother's family stays) with equal force BP12: timing to be increased; more & cleaner wa ter BP15: better quality; 2 times water daily rather than 1 BP16: should be cleaner and cheaper BP17: regular and clean water BP18: water supply piped to home and better quality DG2: timing of water supply should be regular; cost should be less DG4: better water quality and costs should be less DG6: clean water, bill should be less DG7: clean water, all the time; less costs DG9: constant supply of clean water DG11: should be clean and cheaper DG18: increased timing of supply; clean { Note: other respondents have no complaints or improvement suggestions )

PAGE 152

134 B. HOUSING: Housing conditions across the three neighborhoods are presented in Figure 2 below. More specifically, the key findings are as follows: House Types (top left) : this figure shows the % in semi pucca (fairly permanent but with mud/brick wa lls with plastic or thatch roof) versus pucca (a permanent structure) housing. In the survey, there were no ka ccha house types (Note: kaccha refers to slum houses without permanent structure' and semi pucca houses co nsidered marginally better than ka ccha ) The semi pucca and pucca house types breakdown are presented versus average Delhi Slum and Non Slum Conditions. Security of Tenure (top right): roughly one in three and one in ten households lack security of tenure in NM and BJ, respectively. Housi ng Density and Presence of C rowded Sleeping Conditions (middle left): the avg. no. of people / home and per sleeping room increases from DG to BJ to NM. Rental Housing and Owned Accomodation (middle right) : almost half of DG residents own, while no more than 15% and 35% own in BJ and NM, respectively. ~40% of households in DG and NM state rent is affordable (i.e. no problems with paying the rent each month) while BJ homes did not respond; Average Rent p er Month (bottom left ): was highest for DG (BJ again did not respond to this question as similar to previous question on affordable housing); Average plinth level : Plinth level was determined by measuring the home surface relative to street surface and can be a useful indicato r for sewage / flood risks The table below also identifies factors appreciate d most about current situation, common issues faced and responses to issues, and what they'd like to change about current situation.

PAGE 153

135 Figure 2. Initial comparative asse ssment of housing conditions

PAGE 154

136 Figure 2. cont'd. Initial comparative assessment of housing conditions

PAGE 155

137 Table 2. Qualitative Assessment of Housing Conditions: Initial Sample Responses Questions New Mustafabad Brijpuri Dilshad Gardens A) How hous ing or lack of this service has affected HH in terms of health and livelihood? NM2: water used to seep in from toilets (toilet blockage during rains; rent 3000Rs/mo most of time facing problem paying rent NM9: too crowded of a home for 11 people BP2: Sin ce home is on top floor, during summer last year children fell ill due to extreme heat No responses B) Water supply factors appreciated most? NM5: open skylight over roof allowing for natural light BP2: open space in front of house; there is extra sleepi ng space within house to host guests BP8: high plinth level (barrier between road & house) DG1: big, spacious veranda DG3: location is good, ventilation, natural light C) What are your major housing issues? (responses to such issues)? NM2: toilet water se eping in and rent NM3: water damage from flooding; poor lighting (sit on roof) NM6: poor ventilation in summer; poor lighting / dimmed low voltage lighting NM9: too crowded NM12: poor ventilation in summer (sit down outside); poor insulation in winter NM1 3: water damage from flooding when heavy rains, water also comes thru the roof (response: put plastic sheets over roof) NM14: poor insulation in winter (plastic hanging in roof door)/ poor lighting NM15: when there are heavy rains ; a little overcrowded NM16: rent NM18: it's very hot in summer BP2: leakage from rooftop BP6: poor lighting;water seeping thru roof BP9: poor ventilation in summer; poor lighting (sit outside) BP11: poor indoor ventilation / lighting (response: use candles, emergency light) BP 13: no cross ventilation (use of Jaali') BP14: structural damage to home BP15: garbage dump in front of house BP16: congested for three persons DG2: noise pollution / vibration sources: vehicle traffic, street dogs, and motor of water (prefers staying h ome, closed door) DG8: garbage dump just by side of the house (complained to the Resident's Welfare Association) DG13: water damage from flooding (damp and paint gets damaged DG15/16: poor ventilation (sit outside most times; open doors and stay out) DG18: poor ventilation; leakage of water thru roof D) Important changes to address the current situation? NM1: increased room space NM3: more ventilation NM8: cementing of floor NM12: less rent, more ventilation NM13: more space, address leaking roof NM14: to get house built completely NM15: make house more waterproof during monsoon NM16: rent shold be less BP2: repair for leakage of rooftop BP3: ground water depletion BP9/13: more ventilation BP11: better ventilation /lighting BP16: should be more spacious / more indoor lighting BP18: want to make plint level higher DG1: more space currently congested DG2: currently less space for kitchen; poor lighting (both of which creates difficulty and can be depressing at times); reduce noise pollution DG8: garbage dumpb ehind home should be removed (slum of ~250 householsd next door with lot of garbage being thrown in an area just at the back of their house which gives bad smell); noise pollution in slums next door should be reduced too DG9: cleaner backyard DG12: entranc e is small; needs to be widened; DG13: address the regular dampness DG15: more ventilation; safety (area prone to theft) DG16: proper ventilation DG18: better facilities

PAGE 156

138 C. ELECTRICITY & COOKING FUELS : Survey results for household energy supply conditions are presented below as it relates to electricity and cooking fuels. Figure 3 presents a preliminary comparison of electricity supply availability and average daily duration of supply across the three neighborhoods; and Figure 4 explores the main sources of cooking fuels used. Figure 3. Electricity Supply Availability and Average Daily Duration of Supply Figure 4. Main Sources of Cooking Fuels: Liquid Petroleum Gas or Solid Fuels Next, in Table 3, we present responses regarding the current situation o f how household energy supply conditions has affected health and livelihood, what the household appreciate s about their current situation, most common issues faced (how they respond), and the improvements to address current situation.

PAGE 157

139 Table 3. Qualitativ e Assessment of Household Energy Conditions: Sample Responses Questions New Mustafabad Brijpuri Dilshad Gardens A) How electricity or cooking fuels supply has affected HH in terms of health and livelihood? NM3: in winter, power cuts; scarcity of wood BP2 : during rains, difficult to get dry wood; also, sometimes gas is also not available so family has to skip usual meal (this happens 2 3 times each rainy season) DG2: power cut timings during summer (affects sleep) B) What is appreciated most? No Responses BP1: not many power cuts DG2/3: availability; good voltage C) Most commonly occurring electricity & fuel issues? (responses to such issues)? NM1: scarcity of wood / dung; and recent rains NM3/9/10: power cuts (wait for electricity to return, go to roofto p, use candles, hand fan) NM4/5/6/7: cost is high (light candles) NM13: VERY HIGH COSTS (3025/mo doesn't matter what month) always too high (when power out, hand fan and uses candles NM14/17/18: power cuts and high bills NM15: electricity cost is an iss ue (so uses less); power cuts (uses candles, and goes to rooftop to sleep) NM16: power cuts (use candles and mobile torch) BP1: cost is high (6 Rs / unit) BP2: power cuts; expensive electricity and low availability of wood (especially during rainy season ) BP3: power cuts (candle, chargeable lamps, wait for electricity to return); costs BP4: cost of bill (3200 Rs for two months) BP5: power cuts (go to roof top; use candle) BP6/7/9: very high bills (use candles when power cuts and wait for power to return) BP10: bill exceptionally & erratically high BP11/13/14: high costs BP12: very high cost / erratic meter readings leading to higher cost (go to rooftop, use candles when power cuts) BP16: power cuts at midnight (go to rooftop, use candles) DG2: very high c ost bills (when power cuts, use inverter) DG3/13/15/16: power cuts (use back up generator; inverter; candles; wait outside) DG 4: high bills; meter readings high, unit cost high (emergency light and inverter used when power cuts) DG6/7/8/9/12: very high bi lls DG10/11: power cuts and high bills DG18: very high bills (more than 10,000 for 2 months) D) Responses to sudden loss of cooking fuels NM2: uses alternative fuel like coal NM3/4: looks for wood NM5: buys gas on black market with no government booking N M6: started eating out NM14: use alternative to gas chullha with wood NM15: carries small spare gas used when raining BP1: gas is always available BP2: skipped meals BP8: use a small cylinder whenever there is supply scarcity BP9: whenever there is shor tage, they borrow from neighbors BP10: "nothing, but price is erratically high" DG4: family starts eating out DG15: get food from outside (occasional shortage of 1 2 days so eat outside DG17: use another source (chullha with wood and upla) E) Important c hanges to address the current situation? NM1: more gas supply NM4: lower the cost; gas cylinder supposed to be for 400 Rs (HP gas); had to give 958 Rs. NM9/10/11/12: less price and power cuts NM13: electricity exceptionally high can't afford it; when didn' t pay, BSES came to cut electricity supply so they gave money; after 2 months not paying, BSES comes to house to cut supply NM15: wishes she had a gas connection NM15/16/17/18: costs and cuts should be less BP3/4/6: less price per unit BP5: for electricit y, no power cuts; for gas, shold be permitted 12 cylinders instead of 9 BP10: both meter reading and permit costs are erratic and need to do something about it; BP11: less rates / unit; meters should be checked for faulty readings BP12/34: bills should be less; regular meter checked up BP15: should be 12 cylinders BP16: less bill no power cuts; subsidized gas connection DG2: address power cut timings in summer (as it affects sleep); gas pipeline needed to lessen burden of logistics with gas cylinders [stor age] DG4; less bills; meters should be checked regularly; supposed to get 9 cylinders in timely manner, not happening doesn't get that much) DG8/9/10/12/16/17: should be provided with allotted cylinders of 12, not 9 per yr (not enough) DG18: very high ele ctric bills and gas is very costly on black market should be provided 12 cylinders / year instead of 9

PAGE 158

140 D. FOOD SUPPLY : Survey results for food supply conditions are presented below as it relates to self reported access to a healthy food supply (left) and whether fruit, meat, and dairy are included in their diet (right). In all study areas, roti (bread) and vegetables are part of the diet for the household (HH), and is therefore not included in the figure on right. New Mustafabad has the lowest number of re spondents stating their household has access to a healthy food supply and less then 20% of respondents are stating they have fruit in their diet. However, they do have highest proportion of households having meat in their diet, most likely due to the predo minantly Muslim population in this area. Brijpuri has slightly better conditions for healthy foods and fruit, yet still less than 50% have fruit in their diet. Figure 5. Initial % of Households Reporting Access to a Healthy Food Supply and Having a D iet Including Fruit, Dairy, and Meat Next, in Table 4, initial study findings are presented of responses regarding the current situation of how food supply conditions has affected health and livelihood, what the household appreciate s about their current s ituation, most common issues faced (how they respond), and the improvements to address current situation. Important responses revolve around food poisoning and less than optimal government subsidized food rations in NM and BJ, while households in DG are al so concerned about food quality and costs.

PAGE 159

141 Table 4. Qualitative Assessment of Food Supply Conditions: Sample Responses Questions New Mustafabad Brijpuri Dilshad Gardens A) How food supply has affected HH in terms of health and livelihood? NM1: episode of illness due to food poisoning BP2: son was sick and hospitalized for food poisoning once (1 month absent from school) B) Food supply factors appreciated most? BP10: no complaints about food BP16: no issues DG1: less oily DG2: balanced food DG3: good quality C) What are your major food supply issues? (responses to such issues)? NM2: ration not easily available thru ration cards NM3: costly food (look for alternative food items) NM4: very little of food items at govt subsidized rates NM5: no subsidiz ed food; shopkeepers do not give the amount entitled to be given NM6: high cost (cut meal sizes) NM8: poor food quality / food damage (cut meal sizes) NM9/10/12/14: poor quality / costs are very high (cut meal sizes / change shops / cut down budget) NM11: food damage NM13: poor quality / high food costs and food we can afford is not very good quality (cut down food budget and get quality food "we prefer quality over quantity"; when food scarcity, store cereals and other foods that don't spoil) NM15: very high costs (cut meal sizes ~4 days / month) NM18: no items on subsidized rate BP1/3/5: prices too high (cut meal sizes) BP2: quality is poor and price is high BP7: cost is very high (cut meal sizes) BP8: poor quality rice / cereals / lentils BP9: vegetab les not fresh / high food costs BP13: poor food quality BP14: less amount available in subsidized rate BP15: price of milk / related products too high BP17: poor food quality (so we go buy from Bhajanpura 2 kms away) BP18: costly DG1: poor food quality and fresh veg / fruits not available (colored, chemical used vegetables / milk products not fresh DG3: high cost DG4: quality not good (try to get food from different shops, spices bought in grains, powdered at home) DG8: high cost of items (cut down budge t at times) DG10: vegetables are not fresh / high price DG12/16/18: high cost (amount purchased is adjusted according to price) DG17: quality (gets rice, gehu, daal, oil etc from village home) D) Comments on food supply feature most interested in impro ving: NM1: "as income increases, food quality will improve especially for children liking milk" NM2/3/11/14/16/18: prices decreased NM4: "they should give amount we are entitled to get at subsidized (should get 35kg wheat; only getting 20kg); should get 10 kg rice, only get 5kg; should get 5kg sugar; for oil / kerosene (was getting 20 liters, now only 5)" NM5: government should make sure the items are available to us at subsidized rates, in the prescribed amount NM6: more amount in subsidized rate NM8: ratio n card not stamped; good quality food NM9: the govt. ration card they've applied for is not being stamped so not getting food at subsidized rate; food quality NM10/12/13: good quality food, cheaper price NM15: cheaper prices for healthy food BP2: quality improvements / prices decreased BP3/7: should be available on subsidized rates; cheaper BP4/16: subsidized priced food items should be provided BP5: "The subsidized rated food items should be in increased amount (15kg is too little) and according to no. of members in a household" BP8: quality; more amount of ration to be given according to no. of family size BP9: should be available at ration shop at subsidized rate BP10: "nothing, everything depends upon salary / income" BP11: "wish there was more supply; need more subsidized supply" BP13/17: quality of items should be improved BP15: price; and "ration card totally useless" BP18: should be able to make free for kids DG1: fresh and organic foods should be available, there should be vigilance on food quality DG3: main food ingredients should be cheaper DG4: good quality food should be supplied DG6/8/9/10/11/12/13/14/16/18: costs

PAGE 160

142 E.TRANSPORTATION: Survey results for transportation conditions by study area are presented below Findings are shown in Figure 6 with exhibits in the following order: A) Breakdown of primary modes of transportation for daily activities by neighborhood B) Breakdown of secondary modes used for daily activities by neighborhood ; C) Breakdown of preferred modes for both longer and sho rter trips: most walk in NM; D) Primary mode for work commute : for DG personal motor vehicles and for NM bus; E) Work commute self reported travel time / distance & % of h ouseholds with personal motor vehicles : highest time / distance for BJ and vehi cle ownership in DG > BJ > NM. Exhibit A Preferred Primary Modes Exhibit B Preferred Secondary Modes Exhibit C Preferred Modes for Short and Longer Trips

PAGE 161

143 Exhibit D Preferred Transportation Modes for Work Commute Figure 6. In itial comparative assessment of transportation conditions In Table 5 below, initial responses on how transportation conditions affected health and livelihood, factors appreciated most about primary and secondary travel modes and the transportation f eatu re s households are most interested in improving are described. Responses also note how price / supply of petroleum affects travel behavior and choices.

PAGE 162

144 Table 5. Qualitative Assessment of Transportation Conditions: Sample Responses Questions New Mustafaba d Brijpuri Dilshad Gardens A) Comments on what is appreciated most about your primary (and secondary) mode of transportation: [Note: cycle = bicycle; bike= motorcycle] NM1: since public transport is available, no need for private vehicle NM2: easily ava ilable buses NM3: walking saves money NM4: walking means you don't have to spend any money NM5: nothing to appreciate NM7: walking less then motorcycle (petrol) NM8/9/11/14/15: bus fare comparatively low NM9: scooter: when go out, it takes less time NM10: bus / cyclerickshaws both low fares NM11: cycle: saves time, money NM16: bus no. of buses and safe NM17: bus fare is cheap; autos are available BP1/2: cheap, always available (buses) BP4: motorcycle good for short distances BP5: bus (cheaper option) BP6: motorcycle fast way to all places, emergency, and easy to ride on busy roads BP10: bus less fare BP11: chooses auto b/c takes less time / always available BP15: bike takes less time; bus cheap BP17: bus less fare; auto: safer BP18: cycle no cos t DG1: Car is easy, fast, less troublesome DG2: Auto always available; metro: fast traveling DG3: Scooter saves time DG3: Car: status symbol; bike: fast, emergency, less expensive DG4: bike easy to travel, less costs; cyclerickshaw available DG6: bike petrol expenses less, easy to move around DG7: auto always fast, always available; metro fast DG8: metro saves time; having personal choice DG9: scooter easy to ride; car whole family can go DG10: car less time; bike less expensive DG11: Bu s cheap fare; Metro Air conditioning (AC), clean DG12: Bike saves time DG13: Bus cheaper; scooter easy DG14: Metro saves time; Bus cheap DG15: Rickshaw: available; Bike: easy to ride in traffic DG16: Bus cheap; Metro fast DG17: Bike easy, saves time DG18: Metro quick, AC, clean; Auto: available all the time B) What are your major transportation issues? (responses to such issues)? NM1: unsafe road / walking conditions ("have complained regarding roads; since, a wall made around nala -ch ildren fall less") NM2: bus stops are overcrowded NM3: dirty / unsafe roads; water logging NM8/12/14/15: infrequent / overcrowded PT; (less travel; step down to wait for bus that's less crowded which isn't always very soon) NM9/10: overcrowded PT NM11: tra ffic on roads (resorts to walking or travel by bus when cannot bicycle on roads) NM13: personal discomfort "very embarrassing&physically very difficult to travel as disabled person on buses" (travels less than she wants to) NM17: unsafe road / walking cond itions BP1/2/5/8/9/10/13/16/17/18: overcrowded PT BP2: As petrol price increase, bus fare increase BP3: no public transport to the main road BP4: fuel shortages / high costs BP5/8/13: infrequent unreliable public transport (PT) (goes by auto at times when infrequent); BP7: infrequent / unreliable / overcrowded PT; unsafe road / walking conditions; dirty roads BP9/18: very crowded PT; have to stand most of the time (avoids traveling outside) BP12: son's experience with crowded PT BP15: unsafe in PT due to p ickpockets BP16: overcrowded PT; compromised safety of ladies (waits for an empty bus) BP17: "overcrowded PT leading to harassment of us ladies by goons on roads and buses" DG1: infrequent / unreliable PT; too many traffic points / police DG2: overcrowded PT (wait for less crowded ones, choose to go by auto instead) DG3: traffic jams; fuel costs are very high (use motor vehicles less) DG4/9/10/12/15: traffic jams (DG12 looks for shortcuts) DG6: unsafe road / walking conditions DG7: unsafe roads during traff ic jams DG8: infrequent and unreliable PT DG13/16: overcrowded PT (DG16: do not go out much) DG18: overcrowded PT; high cost of petrol C) Comments on feature most interested in improving: NM2/8/9/10/17: increase in no. of buses NM3/18: cheaper prices fo r PT/ less bus fare NM12: better traffic control NM13: government aid in form of a scooter, 3 wheeler, handcycle, etc for disabled member of house to have comfort travelling NM14/15: more frequent bus service BP3/8/13: increase frequency of bus services B P4: lower petrol prices BP5/7/9/10/18: more buses BP9: more seats reserved for ladies / elderly BP10: more no. of public transport options BP14: smoother traffic BP15/17: more frequent bus with one security guard in each bus; more safety in buses BP16: les s price of traveling DG2/17: more frequent services: metro station should be nearer DG3/6/10/12/18: reduce petrol price DG4: better road traffic mgmt (more ways for vehicles in road remove illegal encroachment) DG7: traffic jam shold be controlled DG8/1 6: more no. of metro trains / buses DG9: traffic should be less DG10/13: more no. of buses, more reserved seats for ladies DG12: traffic control,15 to 20yr old vehicles banned from roads DG14: more convenient services for elderly

PAGE 163

145 F. SANITATION, ROAD & DR AINAGE CONDITIONS, WASTE, & PARKS: Findings on primary toilet facility arrangements (top left), handwashing facility (top middle), toilet sharing conditions (top right and middle left), road (middle) and drainage (middle right) conditions outside home, hou sehold waste facilities (bottom left) and parks access (bottom right) are presented below. Figure 7. Intial baseline comparative assessment of sanitation, roads, drainage, waste, parks and open space conditions Table 6 below summarizes the qualitative responses in terms of sanitation, roads, drainage, parks and waste. Similar to previous sections, this includes responses on how household health and livelihoods are affected, what's appreciated most about these infrastructure conditions the issues most often present, and the household suggestions on improvements to current conditions.

PAGE 164

146 Table 6. Qualitative Assessment of Sanitation, Roads, Drainage, Open Space and Waste Conditions: Initial Sample Responses Questions New Mustafabad Brijpu ri Dilshad Gardens A) How these conditions or has affected HH in terms of health and livelihood? NM1: daughter fell down in nala (large sewer drain) while playing, broker her hand; NM2: smelly nala just outside house NM5: no park areas for recreation; pe ople do not have place to walk NM14: lots of mosquitoes due to open nala BP1: lot of flies / mosquitoes breeding in open drains BP5: dirty clogged drains, fleas/flies/mosquitoes, no parks DG2: sewer pipe leakages causing children to be not allowed to play in park DG3: regular / frequent drain blockage causing flooding DG12: roads not in good shape, difficulty in riding bike B) Factors appreciated most? BP2: clean toilet, paved road in front of house DG2: regular cleaning by municipality worker DG4/6/9/ 13/15/16/16: no issues C) Issue that most often comes to mind: (responses to such issues)? NM2: garbage disposal / waste and nala outside house NM3/9/10: poor drainage and garbage disposal/waste NM5: no park areas for recreation NM6/8: drainage / nala / w aterlogging NM7/12/16: drainage / naala; no parks NM14: mosquitoes due to open naalas NM15/16/17/18: waste management BP1: garbage disposal / waste and waterlogging in nala BP2: drainage and garbage disposal/waste BP3/7/9/10/12/13/14: waste mgmt and drai nage ("very sad need sewer") BP4: flies in open drains BP5/15/16/17: dirty clogged drains BP5/10/15/16/17: no park areas DG1/3: regular / frequent blockage of drains and drains that are not covered DG2: sewer pipe leakage DG7/10/11/14: poor drainage and waterlogging DG8: more cleaning of road DG12: roads are not in good shape DG18: poor street and road conditions D) Important changes to address the current situation? NM2: address open nala near house NM3: paved roads NM4: frequent cleaning of drains NM 5/7/12/16: there should be a park; nalas (open sewage drains) should be covered NM8: improved toilet / drains Nm10: waste disposal govt cart should come; NM11: drainage system/ waste disposal improvements; NM13: drainage should be closed and wants cart to come collect the garbage (nobody to go and throw garbage so her elder mother has to do it for her as respondent is disabled) NM14: right positioning of garbage collecting car NM15: need a park; there's a park in Yamuna Vihar, o got here for walks it wo uld cost 20Rs, take 20 mins to get there (which is unreasonable to respondent) NM17: garbage dumping facility NM18: need to address garbage and drainage BP1/2: improve drainage system; cover nala BP3: need parks; waste mgmt (more dumping stations needed); road repair BP6: drains to be covered; sewer system improved BP7/8/9/10/11/12: sewer lines / park should be here BP13/17/18: park, waste mgmt, drainage (when drains are cleaned, the MCD workers do not pick the drain dirt, so the dirt again flows back into the drain/clogs it) BP14: waste mgmt and park BP15: park should be there; naala maintenance / covered DG1: address the narrow roads, open drains/often clogged, and open garbage dumpsters DG3: regular and frequent blockage of drains "all neighborhood hire cleaning worker, complain" for more regular maintenacnce of sewer drainage improvements DG7/10 drainage system should be better (regular cleaning and covering) DG8/12: roads should be cleaner DG11/14: drains should be maintained DG18: roads can be better Note: S imilar survey response data was also collected on experiences with environmental pollution and extreme weather (per Appendix E). Such data has yet to be summarized and could perhaps be developed into a separate future journal article from this th esis chapter acknowledging data collection efforts taking place during dissertation field work and as motivated by this chapter.

PAGE 165

147 Appendix G Household Survey Instrument Used SURVEY MODULE: ASSESSMENT OF INFRASTRUCTURE, ENVIRONMENTAL, AND EXTREME WEATHER C ONDITIONS Start of Module 2: Module 2 focuses on assessing infrastructure, environment, and extreme weather conditions (26 questions): researchers will document your views on 1) access to basic infrastructure services, such as adequate piped drinking wat er, sanitation, electricity, roads, and affordable / adequate housing; and 2) exposure to environmental conditions such as air and water pollution and severe weather events such as extreme heat and cold and how it affects you. Again, we will not keep on file your name or address, or ask any other personal information that can identify you. You do not have to answer any question you do not want to, and you can end the questionnaire at any time. Any information you give me will be confidential and if any qu estions, we will provide a phone number at the end of the survey for you to call to get more information. This survey module should take no more than 30 minutes to complete. Thank you for completing this survey. The information collected will be used to he lp identify better solutions for improved health in your community. This module asks questions in the following order: A. Infrastructure conditions (water supply, housing, energy supply, food, transportation, sanitation, drainage, waste, and street conditi ons); B. Environmental pollution conditions (outdoor/indoor air pollutionr); and C. Extreme weather conditions (extreme heat and drought, extreme cold, and urban flooding).

PAGE 166

148 Table 1. QUESTIONS ON VARIOUS INFRASTRUCTURE SECTOR CONDITIONS *infrastructure "Buniyadi Dhaancha"/ "Avsanrachna" Infrastructure Sectors ( Buniyadi dhancha kshetra ) What type of infrastructure conditions? (Buniyadi Dhancha ki avastha kis prakar ki hain?) Satisfaction with existing conditions: (Maujuda Haalat se santushti) : + What ab out this infrastructure you like or appreciate the most (nothing can be an option) (Is buniyadi dhaanche me kon si baat hain jo aapko pasand hain ya aap sarahte hain? "Kuch bhi nahin" bhi ek vikalp ho sakta hain ) What you would change or improve of the current situation and why? (Maujuda haalaat ko sudharne ke liye aap kaun sa badlao laana chahte hain?) WATER SUPPLY Q1a: Drinking Water Supply : How does drinking water come to your home? Is it piped into dwelling? 1 Yes 2 No; Ple ase explain: ___________________________________________ ( Peene ka Paani ki suvidha : Kya aap ke ghar me peena ka paani ki suvidha hain?) If it doesn't come to your home, skip to Q1f. ( agar Na', toh seedha Q2 jayen ) Q1b: If piped water does come to your h ome, how many taps do you have in your home? (Agar nal ka paani aapke ghar aata hain, toh aapke ghar kitne nal hain?) _____ Q1c: What is the frequency of daily water supply? (Paani din me kitne baar aata hain?) 1 Daily, once (Rojana ek baar) 2 Daily, twic e (Rojana do baar) 3 Daily, more than twice ojana do baar se adhik) 4 Always (Hamesha) 5 Alternate days (Ek din chodkar ek din) 6 Biweekly (Hafte me do baar) 7 Weekly (Hafte me ek baar) 8 Irregular supply (aniyamit sapply ) Q1d: What is the average hour s of water supply in a day? Duration in hours: (Prati din aushatan kitne ghanton tak paani ka supply rehti hain? Ghanton ke gisab se likhiye : ) Q1e: Please circle below the main and back up source of drinking water for your household: (Kripaya neeche diye gaye peene ka paani ki mukhya aur vikalp shrot jo aap ke ghar me hain, unme gol nishan banayen ) PIPED WATER: ( Nal ka paani ) 1 PIPED INTO DWELLING (Aavaash tak nal) 2 PIPED TO YARD/PLOT ( Parishar/chaukhat tak nal ) 3 PUBLIC TAP (sarvajanic nal) HAND PUMP: (Haath pump) 11 INTO DWELLING (Aavaash tak) 12 INTO YARD/PLOT (Parishar/chaukhat tak) 13 PUBLIC HAND PUMP ( Sarvajanic haath pump) 14 TUBEWELL OR BOREHOLE ( tube well ya bore hole) DUGWELL: (Kuwaan ) 21 PROTECTED WELL (Surakshit kuwaan) 22 UNPROTECTED WELL (Asurakshit 3a What is the most commonly occurring water supply issue 1 Don't have water when I need it 2 The water is dirty 3 High costs of water 4 Other. Please specify 3b What is one thing you have done in response to this problem? 1 Rainwater Harvesting 2 Conserving water 3 Community petition to municipal authority or water authority 4 Storage of water in containers 5 Other: Q3c How do you store your water ? (Jab nal ka Paani band ho jata hain toh aap kaise paani jama karke r akhte hain?) If using containers, are they covered? Yes, all are covered 1 | No, none are covered 2 | Some are covered 3 (Agar container/gamle ka istemaal kar rahe hain toh kya unhen dhak ke rakhte hain? 1 Haan, Sab dhake

PAGE 167

149 WATER QUALITY kuwaan) DISTRIBUTOR: ( Vitarak) 31 TANKER TRUCK (Tanker Truck) 32 CART WITH SMALL TANK (Chota tank wala Gaadi), 46 OTHER (SPECIFY) (Anya kripaya ullekh kijiye ) Q1f: If your source of drinking water is outside premises of home, how far is the drinking water source from your household? ____Meters; (Agar aap ki peene ka paani ki shrot aangan ki baahar hain, toh who aap ke ghar se kitna door hain?) -------Meters) Q1g: What is the time to go there, get water, and come back in one trip: ____Minutes (Wahan jane me, paani bharne me aur le kar wapas aane me kitna samay jaata hain : --------minutes ) Q2a: Water: Drinking Water Quality Type of Household Water Treat ment: Do you purify or treat the water every time you collect drinking water? Yes 1 | No 2 (kya aap peene ka paani ko filter ya kuch aur tarike se treat karte hain harbaar?) Haan 1 | Na 2 If Yes, what do you do to purify, filter or treat your water? (Not e: Record all mentioned of below): Agar haan, toh aap kaise paani ko treat karte hain? 1 Boiling (ubaalna) 2 Use Alum (alum istemaal karte hain) 3 Add Bleaching Tablets/ Chlorine Tablets (chlorine ki goli) 4 Strain through a cloth (kapde se filter) 5 Use Water filter: Please specify type: ceramic/ sand/ composite / reverse osmosis, other: (Paani ki filter mitti/baalu/RO ityadi) 6 Use Electric Purifier (Type: _________ ) (electric ) 7 Let it Stand and settle( ganda paani ko neeche jamne diya) 8 Others (Specify ) (anya) 9 Don't Know (pata nahin) Q2b: For your drinking water, please rank on a scale of 1 (excellent) to 5 (very poor ): peene ka paani ki bare me likhe taste kaisa hain cougue kaisa hain? Gandh kaisa hain?) How is the taste? ___ Color? ___ Smell? ____ huwe hain, | 2 Nahin, ek bhi Dha ka huwa nahin hain, | 3 kuch Dhake huwe hain ) Q4a: SMILEY EXERCISE: Overall, how satisfied are you with the water supply and drinking water to your household rank satisfaction with sticker: using a scale of 1 (very poor), 2 (poor), 3 (neutral), 4 (good), 5 (very good) (Aap ke ghar me paani ka supply ko le ke aap kitna santusht hain? Kripaya aap kitne santusht hain, neeche diye gaye vikalpon me se kisi ek se batayen 1(Bahut kharab), 2 (kharab), 3 (pata nahin), 4 (accha), 5 (Bahut ac cha) Q4b: What would you change about current situation and why ?(Maujuda haalaat me sudhar ke liye aap kya karna chahoge aur kyun?) HOUSING Q5a: Type of house (Hindi definition): Ghar ka Prakar : 1 kuchha (weak structure) (Kamzor Dhancha) 2 semi p ucca (fairly permanent but with mud/brick walls with plastic or thatch roof; marginally better than kuccha) ( AAdha pucca Kuch had tak sthyayi par mitti/eet ki deewaren aur plastic ya chappar ka chhatt, kuccha ghar se kaafi behtar) 3 pucca house (a perm anent structure) ( Pucca Ghar Sthyayi dhaancha) Q5b: Security of tenure: Are you protected against forced eviction from your house or the land you live on ? Yes / No; Please comment: ________ ?(Kya aap Zabardasti ghar ya zameen se bedakhal hone ke khila ph surakshit hain? Haan / Na, Kripaya ullekh kijiye .. Q5d: Housing Density: Total people in your home over the past 30 days: ( Aavashuya Ghanatva : Peechle 30 dinon me kul mila ke aapke parivaar me lohon ki sankhya): ___ Q5e : Sleeping Conditions: # of people / sleeping room in your home over past 30 days: ( Neend se judi haalaat peechle 30 dinon me aap ke ghar me logo ka/ sone ke liye kamre Q6a: SMILEY EXERCISE: Overall, how satisfied are you with your housing conditions: Please rank using scale of 1 (very poor), 2 (poor), 3 (neutral), 4 (good), 5 (very good). (Aavaasiya haalat se aap kitna santusht hain, Kripaya neeche diye gaye vikalpon me se chuniye 1(Bahut kharab), 2 (kharab), 3 (pata nahin), 4 (accha), 5 (Bahut accha) Q6b: What feature would you like to improve the most and why ?(Kon sa cheez aap behtar karna/banana chahte hain aur kyun?) Q6c: A ffordability of housing :

PAGE 168

150 kitne hain) ____________________ Q5f: Sewage Flooding (MEASURE / OBSERVE DON'T ASK): The plinth level (housing surface relative to street surface) of your home : ( Mal yukt paani ki bahaw : Aapke Ghar ka farsh bahar ki zameen ke star se kitna upar hain? ) ______ cm 5h Most commonly occuring housing issues for your household? Y / N? Circle if an issue: 1 poor construction quality 2 structural damage 3 water damage from flooding 4 poor indoor ventilation in summer 5 poor insulation in winter 6 poor lighting / dimmed low voltage lighting 7 heavy rains 8 rent 9 Other. Please specify: 5i: What have you done in response to problem/s? Q5g: (OBSERVE Don't ask) Is your house nearby ( Kya aapka ghar iske paas me hain? ) 1 Industrial Facilities Y | N ( Oidyogic kshetra Haan | Na ) 2 Market Areas Y | N (Bazaar Haan | Na) 3 Bus Stations Y | N (Bus Terminal Ha an | Na) 4 Trains or Railway Stations Y | N ( Rail ya Rail Station Haan | Na ) 5 Water Bodies Y | N ( Jalashay/Paani Bhara jagah Haan | Na) 6 Main roads Y | N ( main Rasta Haan | Na ) 7 Construction Y | N ( Nirman Sthal Haan | Na) 6 Loud Noise Pollution Y | N ( Sthaniya Shabd Pradushan Haan | Na ) (Aavashiya Samarthyata) Do you own a house or are you in rental accommodation? (Aapka apna ghar hain ya kiraye ka makaan me rehte hain?) __________________________ If renting, how much do you pay for your housing each month ?( Agar kiraye par rehte hain toh har mahina kitna kiraya dete hain?) _____ Rs. / mo ( Rupaye/mahina) Q6d: Can you afford this cost each month? (Kya har mahina yeh kharcha uthane me aap samarth hain?) 1 Yes, no problem (ji haan, koi dik kat nahin) 2 Most of the time (jyadatar samay ) 3 Sometimes ( Kabhi Kabhi) 4 Rarely (muskil se kabhi) 5 Very rarely ( Shayad hi kabhi) ENERGY: Q7a: Electricity Access : Do you have access to electricity? 1 Yes | 2 No ( Bijle e ki pahunch : kya aapke ghar me bijlee hain ? Haan | Na ) Q7b: Frequency of Electricity Supply : For about how many hours do you have electricity daily (i.e. available <4 hrs / day, 24 hrs / day, etc): __________( Bijlee supply ki raashi : Har din aapke g har kitne ghante bijlee rehti hain ? . Q7c: Cooking Fuels: What type of fuels does your household mainly use for cooking? ( Kana banana ka indhan : Aapke ghar me khana banana ke liye kon sa indhan ka istemaal hota hain?) 1 Wood ( lakdi) 2 Dung Cakes (Gobar /upla ) 3 Coal/Coke/Lignite/ Charcoal ( Koyla/coke/lignite/angaar) 4 Kerosene (kerosene) 5 Electricity ( Bijlee ) 6 Liquid Petroleum Gas (LPG/Gas) How satisfied are you with your energy supply specifically your electricity and fue l supply conditions .( Aap apne ghar urza/power supply ko leke kitna khush hain, khaskar bijlee aur indhan ka supply ) Q9a: SMILEY EXERCISE Satisfaction with Electricity Supply: Please rank using scale of 1 (very poor), 2 (poor), 3 (neut ral), 4 (good), 5 (very good )(Bijlee ki supply ko leke santushti : Kripaya neeche diye gaye vikalpon me se chune 1(Bahut kharab), 2 (kharab), 3 (pata nahin), 4 (accha), 5 (Bahut accha)

PAGE 169

151 Electricity & Fuel Supply 7 Bio gas ( Bio gas/ Jaivik Gas) 8 Other (Specify): (Anya Kripaya ullekh kijiye) Separate room for cooking? Y / N Kitchen Exhaust fan? Y / N Poor Ventilation / Good Ventilation: ______ (observe) Q8a: What is most commonly occurring issue with electricity supply: circle below. 1. power outages 2. costs for back up diesel generator 3. low voltage 4. Other: Q8b: Ho w have you coped with or responded to sudden disruptions in electricity supply? For example, when you had power outages/blackouts, did you: Use a store battery... 1 Back up diesel generator.. 2 Wait for electricity to return. 3 If other, please specify: ____________________________________ Q8c : what have you done in response to sudden loss of cooking fuels ? How have you coped with or responded to sudden disruption in cooking fuel supplies? Did you cut the size of fuel consumption because there wasn't en ough availability ? Y / N Used another source of cooking fuel ? Y / N Please specify which type: _______ Started eating out? Y / N Other steps you took? Please specify: Q9b: SMILEY EXERCISE Satisfaction with Cooking Fuels Supply: 1 (very poor), 2 (poor), 3 (neutral), 4 (good), 5 (very good ) (Khana pakane ki indhan ko le ke santushti : Kripaya neeche diye gaye vikalpon me se chune 1(Bahut kharab), 2 (kharab), 3 (pata nahin), 4 (accha), 5 (Bahut accha) Q10: What featu re would you like to improve the most about your electricity and gas supply? And why ?( Kaun si cheez aap behtar karna sabse jyada chahenge aur kyun?) Electricity: Gas:

PAGE 170

152 FOOD SUPPLY Q11a: Food: Is there access to a healthy food supply? Y/N Khana : ( s waccha khane ka supply hain? Haan /Na) Q11b: Diet: What does your diet include? Please circle : (Aahaar : Aap ka Aahar me kya kya hota hain? Kripaya gol nishan banayen) 1 Roti ( Roti) 2 Vegetables (Subji) 3 Fruit (phall) 4 Non veg (maansahwri) 5 Dairy ( Du dh / dudh utpaad) 6 Other: Please: specify ( Anya : Kripaya ullekh kijiye ) _____________________ 12a What is the most commonly occurring food issue. Please tick below. __ poor food quality __ food damage __ food scarcity __ unhealthy food __ other: __ ___ 12b What have you done in response to this problem? 12c When food scarcity, did you: c ut meal sizes because there wasn't enough availability of food (or money for food)? Y / N Did you store more food? Y / N Please specify: Q13a: SMILEY EXERCISE H ow satisfied are you with your food supply conditions: Please rank using scale of 1 (very poor), 2 (poor), 3 (neutral), 4 (good), 5 (very good) .( Aap ke khadya utpaad ki supply se aap kitna khush hain? Neeche diye gaye vikalpon me se chu niye 1(Bahut kharab), 2 (kharab), 3 (pata nahin), 4 (accha), 5 (Bahut accha) Q14b: What feature would you like to improve the most and why ?( kis cheez ko aap behtar karna sabse jyada chahenge?)

PAGE 171

153 TRANSPORT ATION Q15a : Transportation Options: Wha t is your primary and secondary mode of transportation for daily activities : (yatayat ke vikalp : aap ki yatayat ki pehla aur dusra sadhan kya hain?) 1 Walk (paidal) 2 Bicycle (cycle) 3 Metro ( metro) 4 Bus ( bus) 5 Cyclerickshaw (Rikshaw) 6 Autorickshaw ( auto) 7 Drive (motorcycle )( bike) 8 Drive (car) (gaadi ) 9 Other: Please specify (anya : kripaya illekh kijiye ) _____________________ Q15b : For trips less than <1km (to the grocery), what's your primary mode: ( 1 km se kum door jane ke liye aap ka sadha n kya hain?) Q15c : When you have to go far (i.e. for work) more than 3km, what is your primary mode: ( Jab aap ko 3 km se door jana ho toh aap kaise jate hain?) For member of HH that travels farthest to reach work, what is the Time: _____ mins; Distanc: ____ kms and primary travel mode t o reach work: _______ Where do the men and women in your household work? Are they leaving neighborhood to go to work? 1 Yes 2 No; For those leaving describe how: Q15d: Transportation Fuels: If owning or using a vehicle what type of vehicle is it: __________( Yatayat ke liye indhan : Agar aap ke paas koi saadhan hain ya fir aap kisi saadhan ka istemal karte hain, toh who kis prakar ka hain?) 16a What is most commonly occurring transportation problem. Please circle bel ow. 1 infrequent / unreliable public transport 2 Overcrowded public transport 3 Fuel shortages / high costs 4 unsafe road / walking conditions 5 Other. Please specify _______ 16b What is one thing you have done in response to this problem? Q17a: SMILEY EX ERCISE: How satisfied are you with your transportation options: (Aap apni yatayat ke vikalpon se kitna khush hain?) 1(Bahut kharab), 2 (kharab), 3 (pata nahin), 4 (accha), 5 (Bahut accha) Q17b: What about your primary and secondary mode of transport do you appreciate the most and why? (Aap ki yatayat ki pehli aur dusri saadhan ki kis cheez ko aap sabse jyada sarahte hain aur kyun?) 2 Q17c: What features would you like to improve the most and why ?( Kin cheezon ko behtar banana chahenge aap sabse jyada?) Q17d: How does price or supply of gasoline affect your travel behavior / choices? ( Patrol/gas ki keemat ya supply ka aap ki yatayat par kaisa prabhav padta hain?)

PAGE 172

154 SANITATION STREET & DRAINAGE CONDITIONS WASTE & OPEN SPACE Q18a: What is the arrangement for the toilet facility for your household? ( Aap ke ghar me sauchagaar ki kaisi suvidha hain? 1 Individual/ private (private/niji) 2 Shared ( aapsi) 3 Community (saamuhic) 4 No facility/ open defecation (koi suvidha nahin, ku le me ) Q18b : What toilet facility type does your household have: (Aapke ghar me kis prakar ki sauchalay ki suvidha hain) 1 Flush Toilet ( flush sauchalay) 2 Flush Toilet to Septic Tank ( Septic tank tak Flush sauchalay) 3 Flush Toilet to Pit Latrine (Pit latrine tak flush sauchalay) 4 Flush Toilet to open drain (kule naala tak flush sauchalay) 5 Flush Toilet to Somewhere else (kahin aur tak flush sauchalay) 6 Pit latrine (gaddha sauchalay) 7 Dry toilet (sukha suchalay) 8 No toilet : dÂŽfÂŽcation in open by a ll men, women, children ( koi sauchalay nahin/ khule jagah par purush, stree aur baccho ka mal tyag karna) Q18c: Is there a handwashing facility with soap right near your toilet? Yes | No (Aap ke suchalay ki paas sabun se haath dhone ka intezaam hain? Ha an | Na) Q18d : If toilet is shared, how many households use this toilet facility? No. of Households: ____ (Agar ek hi suchalay kayi parivaar aapas me istemaal karte hain, toh aise parvaaron ki sankhya kitna hain? ) Don't Know ( pata nahin) Q18e : Stree t / Drainage Conditions: What type of roads and drainage is by your house? Please circle all that apply: 1 Unpaved roads | 2 Both paved/unpaved roads | 3 Paved roads | 4 No Drains | 5 Open Drains/Sewage channels | 6 Drains are often clogged | 7 Majority o f area has underground drains ( Rasta/Jal nikashi vyavastha : Aapke ghar ke paas sadakon ki naalaon ki haalat kaisi hain? Kripaya neeche diye gaye vikalpon me se gol nishan banakar chune 1 kucchi, 2 kucchi aur pucci, 3 pucci sadak, 4 koi naala nahin, 5 khule naalen/sewage channels, 6 naala aksar block rehte hain, 7 jyadatar jagah zameen ke neeche naala hain.) Q18f: Open Space / Parks: Do you have access to public green space, park and playground areas by your neighborhood streets: Yes | No( Kya aap ke ghar ke paas khula hariyali se bhari jagah, park aur khelne ki jagah hain.? Haan | Na) Q18g: Garbage Waste: Where or how do you dispose of household garbage? (ghar ka kuda aap kahan aur kaise fenkte hain?) 1 Sweeper takes it from home ( Ghar se jamadaar l e jata hain) 2 Nagar Nigam Garbage dumpster (nagar nigam ki kuda lene wali gaadi aati hain) 3 Street throw (sadak par fenkte hain ) 4 Nala (nale me) 5 Open space ( khuli jagah pe) 6 Compost ( compost karte hain) 7 Trash burning (jala dete hain) 8 If other, p lease specify: ( Anya, kripaya ullekh kijiye ) ___________ 18h Which issue below comes to mind most often: 1 Poor toilet facilities 2 Poor street and road conditions 3 Poor drainage and waterlogging 4 No park areas for recreation / children playing 5 Garbage burning & waste Q18i: SMILEY EXERCISE How satisfied are you with your (Aap kitna khush hain ) sanitation / toilet conditions ( safai vyavastha se): street / road pavement conditions drainage systems (rastaa/galiyon ki ha alat aur drainage se) : green space / parks (hariyali/udyan se) waste management (kuda fenkne ki vyavastha se): Please rank each using scale of 1 (very poor), 2 (poor), 3 (neutral), 4 (good), 5 (very good ). (Neeche diye gaye vikalpon me se chuniye 1(Bahut kharab), 2 (kharab), 3 (pata nahin), 4 (accha), 5 (Bahut accha) Q18j: What feature/s would you like to improve the most and why ?( kaun si baat aap behtar karna chahenge aur kyun?)

PAGE 173

155 Thank you. Th is completes the first half of this module on infrastructure conditions and related health impacts. T he remaining questions are on ENVIRONMENT & WEATHER CONDITIONS e.g. pollution and extreme weather episodes as it relates to your health: Table 2. Quest ions on Environmental Conditions Paryavaran ki sthiti What type of environmental conditions? Paryavaran ki kis prakar ki sthiti ? OUTDOOR AIR POLLUTION Baahar ki vaayu pradushan Q19a: Do you sometimes experience breathing difficulty outside? Y | N (if N, skip to next Q) Q19b: SMILEY SCALE EXERCISE: Does it disrupt your life in terms of health and livelihood? (skip if only minor inconvenience) 1 minor incovenience; 2 headache (e.g. visual someone carrying a heavy load); 3 causing serious disruptions (si ck in bed) Q19c: Can you give us examples as to when, where, and why:: _______________ ( Baahar ka kon se kaam karte waqt aapko saans lene me padeshaani hoti hain?) Q19d How do you respond to this situation? Q19e: Do you have ideas of actions that othe r people can take that can help you? i.e. the community, the government authorities? INDOOR AIR POLLUTION GHAR KE ANDAR VAAYU PRADUSHAN Q20a: Do you sometimes experience breathing difficulty inside? Y | N (if N, skip to next Q) Q20b: SCALE EXERCISE: Does it disrupt your life in terms of health? (skip the rest of this question if only minor inconvenience) 1 minor incovenience; 2 headache (e.g. visual someone carrying a heavy load); 3 causing serious disruptions (sick in bed) Q20c: Can you give us example s as to when, where, and why:: _______________ ( Baahar ka kon se kaam karte waqt aapko saans lene me padeshaani hoti hain?) Q20d How do you respond to this situation? Q20e: Do you know how others have responded? Members of your community? Govt. authori ties? Table 2 (cont'd.) Questions on Extreme Weather Conditions: Inconveniences, Disruptions, and Responses EXTREME COLD Q22a: Do you recall episodes of extreme cold in past winter? 1 Yes 2 No (skip to next Q if no) In sardiyon me kya aapne bahut j yada thandi mehsoos ki? 1Haan | 2Na Q22b: SCALE EXERCISE: Did the cold events this December disrupt life for you and members of your household in terms of health and livelihood? 1 minor incovenience; 2 (xxx); 3 causing serious disruptions (sick in bed) ( **skip the rest if only minor inconvenience) Q22c: Can you give examples as to how your life was disrupted: Q22d How did you respond to this situation? Q22e: Do you know how others have responded? Members of your community? Govt. authorities?

PAGE 174

156 EXTREM E HEAT AND DROUGHT Q23a: Do you recall episodes of extreme heat or drought in this past summer? 1 Yes 2 No In garmiyion me kya aapne bahut jyada garmi mehsoos ki? 1Haan | 2Na Q23b: SCALE EXERCISE: Did heat and drought events this past summer disrupt life f or you and members of your household in terms of health and livelihood? 1 minor incovenience; 2 (xxx); 3 causing serious disruptions (sick in bed) (skip if only minor inconvenience) Q23c: Can you give examples as to how your life was disrupted: Q23d How did you respond to this situation? Q23e: Do you know how others have responded? Members of your community? Govt. authorities? FLOODING 24a: Do you recall episodes of flooding in past rainy season? 1 Yes 2 No Is baar bearish me kya baadh aaya tha? 1Haa n | 2Na Q24b: SCALE EXERCISE: Did rain and flooding events this past rainy season disrupt life for you and members of your household in terms of health and livelihood? 1 minor incovenience; 2 (xxx); 3 causing serious disruptions (sick in bed) *(skip the re st if only minor inconvenience) Q24c : Can you give examples as to how your life was disrupted: Q24d. How did you respond to this situation? Q25e: Do you know how others have responded? Members of your community? Govt. authorities? Q24e: (OBSERVE DON'T ASK) Is your home within 1km of any of the below. Please circle: Kya aap ka ghar neeche diye gaye cheezon se 1 km ke andar ki duri par hain? 1 water bodies (e.g. rivers/streams) Nadi/taalab | 2 sewage bodies Naala | 3 garbage piles kude ki dher 4 ro ads with poor drainage and prone to waterlogging khule nalaa wale rasten | 5 Other: anya ____________________ Q25. Final p rioritization: Now addressing some of these infrastructure, environmental, and extreme weather conditions: What are the top t hree most important changes if considering infrastructure, environmental conditions, AND extreme weather conditions toward improving your well being (i.e. your health, livelihood, and standard of living). Please rank the top three. Aap ki bhalai ke li ye abhibhut sanrachna, paryavaran aur mausam ki teebrata ko dhyan me rakh ke kon si cheez sabse jyada zaroori hain, neeche diye gaye vikalpon me se chuniye __ W ater __ Housing __E lectricity for lighting, fans, tvs and appliances in your house __ Access to fuels for cooking __ T ransportation __ Food __ Sanitation __ Street and road drainage conditions __ Waste mgmt __ Access to green parks / open spaces __ Managing extreme events that result in lacking access to infrastructure services __ Access to cl ean outdoor air __ Access to clean indoor air __ Access to clean water __ Access to clean drinking water __ Limiting exposure to hazardous chemicals and waste __ Limiting exposure to occupational hazards __ Improved management of extreme heat and drought events __ Improved management of flooding events paani makaan pankha, TV, light aur anya saadhano ki liye bijlee rasoi indhan ki purti yatayaat khana safai rasta aur naale ka vyavastha kuda hari bhari udyan/ khele ki jagah/ khula jagah saaf taaza hawa saaf paani saaf peene ka paani haanikarak rasayanik kaam karne ki jagah pe haani mausam ki teebrata se bachaw baadh se bachaw Thank you for completing this survey. You have now completed Module 2. If you have any questions you can call the research coordinator, _____ ______, at xxxxx.xxxx Aap ki yogdaan ke liye dhanyabaad.

PAGE 175

157 Appendix D. Summary of Response Coding *(Survey Response Data Available Upon Request) Summary of Response Coding Sample Household A (BJ) Sample Household B (NM) Sample Household C (DG) HH Access to Drinking Water: 1 Y 2 N 1 2 1 No. of Taps in Home 1 0 1 Frequency of Daily Water Supply: 1 Daily, once 2 Daily, twice, 3 Daily, more than twice 4 Always 5 Alternate Days 6 Biweekly 7 Weekly 8 Irregular Supply 2 (morning, evening) 2 2 HH Avg Hrs Water Supply Per Day 6 2 4 Main Source of Drinking Water (see code) 14 (tubewell / borehole at neighbor's house) 3 (Public Tap) 1 (Piped into Home) Backup Source of DrinkingWater (see code) Public tap (2nd preference) No back up source Distance to drinking water source from HH: __ (in meters) 50 to 100 m 10 NA Time taken to fetch water from source outside HH premises: ___ minutes to go there, get water, and come back in one trip NA 15 NA Total water consumption pe r household member each day: (liters) 18 Total Water Consumption for Drinking Per Member of Household (liters) 2.0 Total Water Consumption for Sanitation Services Per Member of Household (liters) 4.0 Total Water Consumption f or Bathing Per Member of Household (liters) 10.0 Total Water Consumption for Washing Per Member of Household (liters) Total Water Consumption for Cooking and Kitchen Per Member of Household (liters) 2.0

PAGE 176

158 How Water Supply an d Lacking Services Has Affected HH in terms of Health and Livelihoods 2 episodes: 15 days each had to fetch water from neighbour's borehole, had to carry 50 100m Water Supply Factors Appreciated the Most Tastes good / amount sufficient Most common ly occuring water supply issue: 1 Don't have water when I need it 2 The water is dirty 3 High costs of water 4 Other. Please specify 1 (not regular) | 2 (dirty) None What is one thing you have done in response to this problem: 1 Rainwater Harvesting 2 C onserving water 3 Community petition to municipal authority or water authority 4 Storage of water in containers 5 Other: 2 | 4 4 How do you store your water? If using containers, are they covered? 1 Yes, all are covered 2 No, none are covered 3 Some are covered 1 1 Water Supply Satisfaction: 5 Very good 4 Good 3 Neutral 2 Poor 1 Very Poor 4 (good) 2 4 Most important changes to current Water Supply Situation and Why: Landlord always forces them to use less water; costly (do not pay the bill 5 Rs per day) Regular water daily; should come twice daily both regularly and on time nothing WATER POLLUTION / DRINKING WATER QUALITY What types of water pollution sources are in your neighborhood: Sewage 1 Y 2 N 1 1 1 2 Hazardous chemicals and waste 1 Y 2 N 2 2 2 Garbage 1 Y 2 N 2 1 2

PAGE 177

159 Factories 1 Y 2 N 2 1 2 Types of Factories: 1 Plastic 2 Paints 3 Dyes 4 Printing 5 Other: 2 Water: Drinking Water Quality Do you purify or treat the water every time you collect drinking water : 1 Y 2 N 2 2 2 What do you do to purify, filter, or treat your water? 1 Boiling 2 Use Alum 3 Add Bleaching Tablets / Chlorine Tablets 4 Strain through a cloth 5 Use water filter (ceramic, sand, composite, RO) 6 Use Electric purifier 7 Let it stand and settle 8 others 9 dont know 10 nothing, direct drinking 10 10 10 Overall: How is the taste? Color? Smell? 1 very poor to 3 (sometimes issues) to 5 Very good Taste: Good; Smell: OK Taste: 3; Color: 3; Smell: 3 ("Sometimes dirty, smelly, and doesn't taste good -typhid an issue in this area") 4 Ways water pollution and poor drinking water quality affected health and livelihoods: occasional diarrhea Type of House: 1 kuccha 2 semi pucca 3 pucca 3 3 3 Security of Tenure: 1 Yes 2 No 1 1 (own house) 1 (own) Geographical Setting: HH is nearby Industrial Facilities: 1 Yes 2 No 1 1 2 Market Areas 1 Yes 2 No 1 1 1 Bus Station 1 Yes 2 No 2 2 2 Train or Railway Stations 1 Yes 2 No 2 2 1 Water Bodies 1 Yes 2 No 1 (20 m) 1 2 Main Roads 1 Y 2 N 2 2 1

PAGE 178

160 Construction 1 Y 2 N 1 1 1 Noise Pollution 1 Y 2 N 1 (noisy chair making factory) Total Ppl per home 5 6 11 Ppl per sleeping room 5 6 4 Plinth Level: ___ cm 2nd floor 30 5 How housing or lack of this service has affected HH in terms of health and livelihood since it i s on top floor, during summer last year, children fell ill due to extreme heat What are your major housing issues: 1 poor constrution quality 2 structural damage 3 water damage from flooding 4 poor indoor ventilation in summer 5 poor insulation in win ter 6 poor lighting / dimmed low voltage lighting 7 heavy rains 8 rent 9 other: please specify: 10 no problem leakage from rooftop 3 (when heavy rains, water also comes thru the roof) 10 What have you done in response to problem/s? they put plastic shee ts over roof Satisfaction with housing conditions: 5 Very Good 4 Good 3 Ok / Neutral 2 Poor 1 Very Poor 4 (poor) 3 3 Comments on what is appreciated most about house: open space in front of house; there is extra sleeping space within house extra spa ce to host guests sleeping Comments on feature most interested in improving: repair for leakage on rooftop more space, adress leaking roof entrance is smaller; needs to be widened

PAGE 179

161 Housing is 1 Rented 2 Owned Accomodation 1 2 2 Rent / Month 1300 Housing Affordability / Ability to pay for housing can you afford this cost each month? 1 Yes, no problem 2 Most of the time 3 Sometimes 4 Rarely 5 Very rarely 4 Electricity Access 1 Yes 2 No 1 2 1 Frequency of Electricity Supply Hours dai ly of electricty supply on average: 24 hrs (in winter); 15 20 hrs (in summer) 18 22 Cooking Fuel Type Mainly Used: 1 Wood 2 Dung Cakes 3 Coal / Coke / Lignite / Charcoal 4 Kerosene 5 Electricity 6 Liquid Petroleum Gas (LPG / Gas) 7 Bio gas 8 Other: 1 (co mmonly used); 6 (only at times) 6 6 How electricity and cooking fuels or lack of this service has affected HH in terms of health and livelihood: during rains, it is difficult to get drywood; also, sometimes gas is also not available so the family has to s kip usual meal (this happens 2 3 times each rainy season) Most commonly occuring issue with electricity supply: 1 power outages 2 costs for back up diesel generator 3 low voltage 4 other 1 (power cuts) | 4 expensive electricity and availability of woo d (especially during rainy season) 1 | 4 VERY HIGH COSTS (3025 Rs / mo doesn't matter what month), always too high 4 bill is very high Responses to sudden disruptions in electricity supply: 1 Use a storage battery 2 Back up diesel generator 3 Wait for e lectricity to return 4 Other 4 (hand fan, candles) 4 emergency light, candles Responses to sudden losses of cooking fuels: 1 Cut size of fuel consumptions when not enough avaialbility; 2 used another source of cooking fuel 3 started eating out 4 Other steps taken: skipped meals no sudden loss no scarcity Satisfaction with electricity supply conditions: 5 Very Good 4 Good 3 Ok / Neutral 2 Poor 1 Very Poor 3 (just ok) 1 3

PAGE 180

162 Satisfaction with cooking fuels supply conditions: 5 Very Good 4 Good 3 Ok / Neutr al 2 Poor 1 Very Poor 1 (very poor) 3 4 Comments on electricity and gas supply feature most interested in improving: Pro: not much power cut; Con: Make electricity cheaper; Pro: Cheap, tasty and healthy food (if cooked with woods); Con: Gas (LPG) should b e available at cheaper price Electricity: expectionally high can't afford to pay it; when didn't pay, they came to cut electricity supply so she gave a check; after 2 months not paying, BSES will come to house to cut supply E: cost should be less; G: nothi ng (except the no. of cylinders allowed since, it is a big family, 9 cylinders / yr is not enough) HH Access to a Healthy Food Supply: 1 Yes 2 No 2 2 1 Diet includes: Roti 1 Yes 2 No 1 1 1 Vegetables 1 Yes 2 No 1 1 1 Fruit 1 Y 2 N 2 2 1 Non Veg / Me at 1 Y 2 N 1 (one time / month) 1 1 Dairy 1 Y 2 N 2 (rarely | Paneer) 1 1 Other: Please specify How food or lack of food supply has affected HH in terms of health and livelihood: son was sick and hospitalized for food poisioning once (1 month abse nt from school) Most commonly occuring food issue: 1) poor food quality 2) food damage 3) food scarcity 4) unhealthy food 5) other, please specify 1 (quality is poor); 5 price is high so quality needs to be imrpoved, price needs to be decreased 1(poor quality) | 5 (high costs) Note: "high food costs and food we can afford is not very good quality" 5 high cost

PAGE 181

163 Responses to problem/s cut down food budget, and get quality food ("we prefer quality over quantity") amount is adjusted according to price Wh en food scarcity, did you 1 cut meal sizes b/c not enough availability of food (or money for food); 2 did you store more food 1: No 2: Yes (store cereals, etc, foods that don't spoil are stored) no scarcity Satisfaction with food supply conditions: 5 Ver y Good 4 Good 3 Ok / Neutral 2 Poor 1 Very Poor 1 (very poor) 2 3 Comments on what is appreciated most about food: nothing Comments on food supply feature most interested in improving: (quality is poor); price is high so quality needs to be imrpoved, price needs to be decreased better quality, less price should be cheaper Primary Mode of transportation for daily activities: 1 Walk 2 Bicycle 3 Metro 4 Bus 5 Cyclerickshaw 6 Autorickshaw 7 Drive (motorcycle) 8 Drive (car) 1 6 7 Secondary mode of transpo rtation for daily activities: 4 (for distant places) | 6 (rare times) 4 4 For trips less than <1km primary mode: Walk (brother has to carry her on his back for 1 2 km (she is disabled and cannot afford wheelchair)) motorbike If traveling far more tha n 3 km primary mode: Bus (more often) / Autrorickshaw (rarely) bus / auto motorbike Where do men and women in househould work Laborer in Bali Maran (Chadni Chowk) Ghaziabad Are they leaving the neighborhood to go to work? 1 Y 2 N 1 1

PAGE 182

164 Time to work: ___ mins 60 30 Distance to work: ___ kms 15 20 Transportation Mode: to reach work: ______ Bus Motorbike Use of Transportation Fuels: Owning a vehicle 1 Y 2 N (Type: _____ i.e. motorcycle) 2 2 1 (two motorbikes) How transportation or lack of this service has affected HH in terms of health and livelihood: Most commonly occuring transportation problem: 1 infrequent / unreliable public transport 2 overcrowded public transport 3 fuel shortage or high costs 4 unsafe road / walking conditions 5 o ther. Please specify with increasing petrol price, the bus fare also increases 5 Personal discomfort "very embarassing and physcially ver difficult to travel as disabled on buses" 5 Traffic jams One thing you have done in response to this problem: tra vels less than she wants to look for shortcuts Satisfaction with transportation options: 5 Very Good 4 Good 3 Ok / Neutral 2 Poor 1 Very Poor 1 (not at all) 1 2 Comments on what is appreciated most about primary and secondary mode of transport: nothing saves time by bike Comments on feature most interested in improving: buses should be frequent, roads should be better, bus ticket prices should be less Government aid in the form of a scooter, 3 wheeler, handcycle, etc would be very helpful to have more comfort travelling petrol price should be less, traffic controled, those cars which are 15 to 20 yrs old should be banned from

PAGE 183

165 roads How price or supply of gasoline affect your HH travel behavior / choices With increasing petrol prices, the bus fare incre ases none doesn't affect as he has to use bikes every day anyway Toilet facility arrangement: 1 Individual / Private 2 Shared 3 Community 4 No facility / open defecation 3 2 1 Toilet Facility Type: 1 Flush Toilet 2 Flush Toilet to Septic Tank 3 Flush Toi let to Pit Latrine 4 Flush Toilet to open drain 5 Flush Toilet to Somewhere else 6 Pit Latrine 7 Dry Toilet 8 No Toilet 2 2 1 (to sewer line) Handwashing facility with soap right near toilet: 1 Yes 2 No 1 2 1 If toilet is shared, No. of Households using toilet facility: 4 hourseholds (20 people) 5 NA Type of Roads: 1 Unpaved 2 Both paved / unpaved 3 Paved 3 3 3 Type of Drainage: 4 No Drains 5 Open Drains / Sewage Channels 6 Drains are Often Clogged 7 Majority of Area has Underground Drains 5 5 | 6 7 Ac cess to public green space, park, and playground areas by your neighborhood streets: 1 Yes 2 No 2 2 1 How household garbage is disposed of: 1 Sweeper takes it from home 2 Nagar Nigam Garbage dumpster 3 Street throw 4 Nala 5 Open space 6 Compost 7 Trash bu rning 8 Other 4 8 Partially enclosed Garbage collection area by nala 2 Comments on what is appreciated most about sanitation conditions: Clean toilet

PAGE 184

166 How lacking access to adequate sanitation, drainage, parks & open space, or waste mgmt services has affected HH in terms of health and livelihood "not much affected used to it" Issue that most often comes to mind of below: 1 poor sanitation 2 lack of facilities for defecation 3 waterlogging 4 garbage burning 5 no park areas for recreation / childr en playing 6 other: UPDATED VERSION ON MAR 1: 1 Poor toilet facilities 2 Poor street and road conditions 3 Poor drainage and waterlogging 4 No park areas for recreation / children playing 5 Garbage disposal and waste 3 | 5 2 (roads are not in good shape, difficulty in riding bike) Satisfaction with sanitation conditions: 3 (ok) clean toilet 4 4 Satisfaction with street / road pavement conditions: 3 (ok road in front of house is paved) 3 2 Satisfaction with drainage conditions: 1 (very poor not at a ll satisfied) 1 4 Satisfaction with green space / parks conditions: 1 (very poor not at all satisfied) 3 5 Satisfaction with waste management conditions: 1 (very poor not at all satisfied) 1 4 Features of above you'd most like to improve and why: dr ainage system should be improved Drainage should be closed and wants cart to come collect the garbage (nobody to go and throw garbage so her elder mother has to do it b/c she's disabled) roads should be better

PAGE 185

1 67 Prioritization: Top 3 most important changes to infrastructure for improving your well being (i.e. your health, livelihood, and standard of living) 1 Cooking Fuels (free govt supply should be there) 2 Water (should be more openly available and not time bound); 3 Street and road/ drainage conditions (should be improved) Infrastructure service provided you would most like available at a cheaper price (please rank 3:) 1 Affordability of cooking fuels (should be cheap); 2 Electricty (less price / unit); 3 Healthcare OUTDOOR AIR POLLUTION (OAP) Types of OAP sources near your home i.e <1km: Dust pollution along unpaved roads 1 Y 2 N 1 1 Main Roads 1 Y 2 N 1 Large Crowded Highways 1 Y 2 N 2 Power Plants 1 Y 2 N 2

PAGE 186

168 Trash Burning Smoke 1 Y 2 N 1 Brick Kilns 1 Y 2 N 2 Factorie s 1 Y 2 N 2 Incinerators / Garbage Dumps 1 Y 2 N 1 Bus / railway stations 1 Y 2 N 2 MOST SIGNIFICANT Pollution source of above: dust on roads Do you sometimes experience breathing difficulty outside? 1 Y 2 N 1 walking (the kids to school) 1 1 Does OAP disrupt your life in terms of health and livelihood? 1 minor inconvenience 2 significant inconvenience (i.e. headaches, or someone carrying a heavy load); 3 causing serious disrutpions (i.e. sick in bed) 3 (respondent and son have vomiting bouts a fter breathing in outside air occasionally) 3 3 Can you give us examples as to when, where, and why: walking kids to school vomiting, headache, and breathing difficulty (she also perceives that ther are illegal factory areas which cause area discomfort fr om small, smoke -they are a little far, but still bothersome dust allergy, eyes irritation, breathing difficulty

PAGE 187

169 How do you respond to this situation? cover mouth / nose with dupatta (scarf) covers face, hands Do you have ideas of actions that othe r people can take that can help you? i.e. the community, the government authorities? no Satisfaction with OAP: 5 Very Good 4 Good 3 Ok / Neutral 2 Poor 1 Very Poor 3 (ok) Comments on OAP feature most interested in improving: pollution / smoke sho uld be controlled INDOOR AIR POLLUTION What types of IAP sources are in your home: smoke from 'chulha' Dust particles 1 Y 2 N 2 Sewage smell 1 Y 2 N 2 Mold from water damage 1 Y 2 N 2 Smoke from 'chullah' 1 Y 2 N 1 Other Do you s ometimes have difficulty breathing inside? 1 Y 2 N 1 (while cooking) 1 1 Does IAP disrupt your life in terms of health? 1 minor inconvenience 2 significant inconvenience (i.e. headaches, someone carrying a heavy load) 3 causing serious disruptions i.e. ( sick in bed) 1 (not affected much) 3 3 Can you give us examples as to when, where, and why: while cooking, there's smoke from

PAGE 188

170 breathing difficulty kitchen, smell, is a source of allergy to respondent How do you respond to this situation? she just goe s away from area where there's smoke he comes out (kithcen separate and exhaust fan) Do you have ideas of actions that other people can take that can help you? i.e. the community, the government authorities? no Do you have difficulty breathing when y ou or others are cooking? 1 Y 2 N 1 1 1 Do you have a separate room, which is used for cooking? 1 Y 2 N 2 2 1 Do you have a kitchen exhaust fan 1 Y 2 N 2 2 1 Has IAP affected your Health and Livelihood: "not affected much" Satisfaction with IAP: 5 Excellent ("Very Clean") 2 Very good 3 Ok / Neutral 4 Fair 5 Poor ("Dirty") 3 HAZARDOUS WASTE POLLUTANTS & OCCUPATIONAL HAZARDS Types of Hazardous waste sites and pollution sources within 1 km of your home? 1 Factories 2 Mining 3 Lead acid battery waste 4 Paints and paint sludge 5 coal based thermal power plants 6 electrical appliances / electronic waste 7 hospital medical waste 8 incinerators / garbage dumps 9 Slaughter house, rotting dead animals Are you or other members of your household exp osed to daily occupational hazards (working conditions that could lead to your illness or death)? 1 Y 2 N 2

PAGE 189

171 Episodes of Extreme Cold this past winter: 1 yes 2 no 1 (4 or 5 times episodes in past 90 days during this past winter; frequency was some weeks but not every week) 1 (for 1 month) 1 (10 days) Did cold events this December disrupt life for you and members of your household in terms of health and livelihoods? 1 minor inconvenience 2 signficiant inconvenicence 3 causing serious disruptions (sick in bed) 2 3 3 Can you give us examples as to how your life was disrupted: burned fire wood (also needed for cooking) more expenses for bonfire / clothing she also had cold and felt extremely cold felt extreme cold during bike rides; less going out; work disrupted How did you respond to this situation? burned firewood Bonfires all the time (using coal and wood), lots of warm clothes -Always have to the buy the coal (wood they can get); and costs about 40 Rs / day to have bonfire for 4 5 hrs firewood at home, more warm clothes Do you know how others have responded? i.e. members of your community? Govt. authorities? nil no Episodes of Heat and Drought in past summer: 1 Y 2 N 1 (continuous extreme heat for 1 month this past 1 (doesn't recall # of days, but many really 1 (2 months)

PAGE 190

172 summer) hot days) Did heat and drought events this past summer disrupt life for you and members of your household in terms of health and livelihoods? 1 minor inconvenience 2 signficiant inconvenicence 3 causing serious disr uptions (sick in bed) 2 2 3 Can you give us examples as to how your life was disrupted: stomach infection, liver problems Felt the heat How did you respond to this situation? bathing, ceiling fan, sitting outside disabled so asks someone to hand fan he r coolers, drink more cold water Do you know how others have responded? i.e. members of your community? Govt. authorities? One or two inconveniences your family had with recent heat and drought events: Responses How did you deal with heat an d drought this past summer? bathing, ceiling fan, sitting outside How heat and drought affected health and well being in terms of health and livelihoods: typhoid / jaundice common in summer in this area Episodes of Rain and Flooding in Past Rainy Season 1 Y 2 N 1 1 Did rain and flooding events this past rainy season disrupt life for you and members of your household in terms of health and livelihoods? 1 minor inconvenience 2 signficiant inconvenicence 3 causing serious disruptions (sick in bed) 3 2 Can you give us examples as to how your life was disrupted: water came into their house during work gets disrupted due to

PAGE 191

173 monsoon; also leakage thru the roof of house affect in transportation How did you respond to this situation? lots of dirty and stinky naala water in their house, had to clean up for 2 3 days; they also use plastic sheets to cover roof nothing Do you know how others have responded? i.e. members of your community? Govt. authorities? raise the plinthe levels One or two inc onveniences your family had with recent rain and flooding events: 2 Home within 1 km of: 1 water bodies 2 sewage bodies 3 garbage piles 4 roads with poor drainage and prone to waterlogging 2 1 | 2 | 3 | 4 3 (Garbage dumpster just 20 meters away, but c overed) Final Prioritization: Addressing some of these infrastructure, environmental, and extreme weather conditions what are the top three most important changes if considering infr, env cond, & extr wthr toward improving your well being (i.e. heal th, livelihood, and standard of living. Pls rank the top 3. 1 Access to cooking fuels 2 Electricity 3 Street and road drainage conditions 1 Drainage; 2 Electricity; 3 Oudoor Air Pollution --AND AS A PERSONAL PROBLEM FOR RESPONDENT : TRANSPORTATION 1 Road ; 2 Extreme Cold; 3 Outdoor Air Pollution

PAGE 192

174 CHAPTER VI. VI. A FOCUS ON EXPERIENCES WITH AND BARRIERS TO ACCESS TO URGENT HEALTHCARE IN THREE NEIGHBORHOODS OF DELHI INDIA Abstract Many urban inhabitants within Indian and Asian cities lack adequate access to urgent healthcare despite the fact that urban areas are typically associated with closer proximity to these services. The objective of this study is to conduct an initial comparative assessment of three neighborhoods within Delhi, India to explore multiple dimenions of access to urgent healthcare while also addressing socio economic status (SES), household health and well being. Household surveys (n=116) and an initial sample (n=30) of household health diaries for the month of April, 2013 are used to begin assessing sociodemograhic, household health and well being, and experiences with and barriers to accessing urgent care facilities T wo low SES and one improved, high SES area are compared. Initial r esults indicate the two low SES areas, New Mustafabad (NM) and Brijpuri (BJ), had relatively similar inadequate ATH conditions. The time needed to reach an urgent care facility was 30 to 40% worse then the high SES neighborhood, Dilshad Gardens (DG). In addition, over 70% of households in NM and BJ currently trav el more than 5km to reach a health facility and while only 10% of BJ respondents having more than one hour in travel time to reach a facility, 50% of NM households have this experience. Not only did NM and BJ have worse ATH conditions, they also had 1.7 ti mes higher disease incidence for person sick days vs. total person days for April, 2013.

PAGE 193

175 Findings show the two low SES areas to have similar travel costs to reach care and different abilities to pay for medical care. The high SES area also had easier acces s to care yet with quality of care less acceptable relative to low SES areas that had issues with wait times and affordability suggesting future study should address such factors and its' effects on health outcomes. Introduction Many urban inhabitants in Indian and Asian cities are experiencing inadequate or delayed access to urgent healthcare facilities despite supposed proximity to such facilities. Inadequate or delayed access to urgent healthcare (ATH) specifically for serious health episodes or condi tions requiring hospitalization is likely to have immediate impacts on human health. This study conducts a comparative study of community health with households in three neighborhoods within the Northeast District of Delhi, India to address these gaps. H ousehold surveys (n=116) and an initial sample (n=30) of household health diaries for the month of April, 2013 are used to begin assessing sociodemograhic conditions (e.g. socioeconomic status, age, religion, education, place of origin (rural vs. urban), l ength of residence in neighborhood, etc), household health and well being, and experiences with and barriers to accessing urgent care facilities T wo disadvantaged or low socioeconomic status (SES) areas and one improved, high SES area are compared. Initia l r esults indicate the two disadvantaged neighborhoods, New Mustafabad (NM) and Brijpuri (BJ), had relatively similar inadequate ATH conditions. The time needed for households in these areas to reach an urgent care facility was 30 to 40% worse than the hig h SES neighborhood, Dilshad Gardens (DG). In addition, over 70% of households in NM and BJ currently have to

PAGE 194

176 travel more than 5km to reach a health facility. At the same time, only 10% of BJ respondents have the experience of more than an hour in travel ti me to reach a facility compared with 50% of NM households. Not only did NM and BJ have worse ATH conditions, they also had 1.7 times higher disease incidence (using the indicator of person sick days / total person days for the month of April). The key weak nesses of this preliminary study that need to be addressed through further community based research are that the associations between access to urgent healthcare and health outcomes are not yet explored due to the infrastructure conditions in each area bei ng different and the small n' sample size. Despite these weaknesses, the findings of this study provides a useful initial baseline assessment of the access to healthcare, health, and socio demographic conditions in these three neighborhoods, and an extend ed follow up, larger n' cohort study in NM and BJ of households sharing the same infrastructure conditions and different ATH conditions will help to assess the important associations between ATH and health while controlling for infrastructure conditions a s a confounder. Rationale The five points below motivate the initial exploratory inquiries in this comparative baseline study in Delhi, India as a representative rapidly growing Asian city, where: >45% of population lives in slums, resulting in a large portion of population exposed to increased health risks while lacking adequate access to healthcare, vaccinations, income, education, mobility and various other services (ToI, 2012); Over 50% of all recorded deaths in Delhi are not classified by cause due to the majority of deaths occuring at home rather than in health instutions where medically certified cause of death records are maintained (NCT of Delhi, 2008).

PAGE 195

177 9% of all deaths recorded by cause in Delhi India are related to pregnancy and birth (Sperl ing & Ramaswami, 2012) ; Temporary and recent migrants are often denied access to health services or are difficult to track for follow up health services (Aggarwal, 2004); and 40% of women in the Northeast District of the NCTof Delhi India don't make it t o a ny health facility for delivery of their child (DLHS, 2008) (see Figure 1). Figure VI. 1 Geographic Distribution of % Slum population to Total Population (Census, 2001) and Births as Institutional Deliveries by Delhi Di stricts (DLHS, 2008) Literature Review and Defining Access To the best of our knowledge, few studies in Asian cities have been conducted to characterize experiences with and barriers to accessing urgent care as differentiated by socio economic conditions and health outcomes. Furthermore, few infrastructure and environment related health effect studies account for this lack of access as an important factor for health effect estimates (Chapter 3). This study is one of few in Asian cities to

PAGE 196

178 characterize hous ehold experiences with and barriers to accessing urgent healthcare (ATH) in the case of serious health episodes that require immediate hospitalization across neighborhoods with different socio demographic and economic conditions. In defining access, many studies simply address distance, cost, and quality of care. For example, a study by Goli et al. 2011 is identified as relevant to this study as it begins assessing poor quality of care and the extent to which urban inhabitants (in both slum and non slum ar eas) lack access to healthcare. However, the study does not address the experiences with or barriers to access to urgent healthcare for whole populations (e.g. beyond instititutional deliveries and a few of these factors). Access in the public health liter ature has been referred to as a multi dimensional concept that describes people's ability to use health services when and where they are needed (Aday & Anderson, 1981). How ATH can impact health of whole populations and various subgroups beyond slum and n on slum conditions e.g. different SES and education levels, by length of residence within a neighborhood and in terms of characterizing multiple access factors (as defined by McLafferty et al. 2002 as the fiva A's) are unique contributions of this study Other factors related to social norms, values, beliefs and length of residence also exist; and in terms of willingness to use particular health services, these were identified in the literature to be shaped by gender, culture, ethnicity, sexual orientati on, multiple experiences based on length of residence, etc. Sense of comfort and satisfaction in receiving services are also dependent on if clients are well treated, providers and clients are able to communicate openly, and if providers are confident abou t the quality of care delivered. Below is a summary of just a few relevant studies on AT H and healthcare utilization. Based on this literature review of various ATH measures, the objective of this

PAGE 197

179 study included identifying measures most sensitive to diffe rences in urban communities and how differences in household characteristics shape awareness of health care options and experiences. Table VI. 1 Review of Studies Characterizing Access to Healthcare and Providing Quantitative F indings in Terms of Experiences with Access to Healthcare. Study Author, Year Literature Review on Barriers to Access to Healthcare McLafferty et al., 2002 Defines and characterizes access to healthcare in terms of five A'variables: 1) Availability : the su pply of services in relation to needs e.g. are the capacity and types of services adequate to meet the urgent care needs? 2) Accessibility : the geographic barriers, including distance, transportation, travel time, and cost. 3) Accommodation : the degree to whi ch services are organized to meet clients' needs, including hours of operation, application procedures, and waiting times (often referred to as the queue'); 4) Affordability : the price of the services with regard to people's ability to pay (Note: both income levels and insurance coverage can be critical aspects of affordability); 5) Acceptability : refers to the client's views of health services and how service providers interact with clients. Views on willingness to use particular health services can be shaped by gender, culture, ethnicity, sexual orientation, etc. Sense of comfort and satisfaction in receiving services can also depend on if clients are well treated, providers and clients are able to communicate openly, and if providers are confident about the q uality of care delivered. Agarwal, 2012 Awareness of Healthcare Providers in which (1) length of time in a neighborhood and (2) i nvolvment in local social networks both influence levels of awareness and selection of providers Goli et al., 2011 Uses the Demographic Health Survey to quantify p roxy attributes (e.g. mother 's delivery of their child) of access to healthcare within slum and non slum neigh borhoods of eight Indian cities; explores % of women who have received three or more ANC visits; % of wome n who had an institutional delivery in any health facility or a government health facility; % of children who have received full immunization; % of households not using a government health facility for birth delivery for both slum and total population; K ey study finding : more than 66% of households within eight Indian cities studied in the India 2005/6 Demographic Health Survey have stated they did not utilize government health facilities due to lack of accessibility and poor qualit y services Fan and Ha bibov, 2009 "Empirical results demonstrate that poverty, chronic illness and disability are the most important determinants of health care utilization andaffordability in Tajikistan. Other significant determinants include gender, the level of education of the household head, and the availability of medical personnel at a given population point." Scheil Adlung and Kuhl, 2011 34% of survey respondents in Croatia, Macedonia, and Turkey report difficulties in accessing medical care due to the cost of seeing a doctor.

PAGE 198

180 While not summarized in above table, analyses using WHO data is shown in the figure below depicting both the density of physicians per 100,000 population (for the year 1998 below) and life expectancy by country vs. physicians per capita. In this study, an initial exploration of these phenomena is conducted via household survey on availability of services' and socio demographic conditions within the three neighborhoods. Figure VI. 2 Physicians Per 100,000 populatio n by Country and Life Expectancy vs. Physicians Per Capita by Country (UC Atlas of Global Inequity and WHO) Methods First, three neighborhoods were selected for participation in: 1) a household socio economic, health and access to healthcare conditions sur vey administered between February and April, 2013 and 2) a health diary collected in April, 2013. The survey, designed mainly based on already existing standardized surveys (e.g. WHO / NIH National Health Interview Surveys, Behavioral Risk Factor Surveilla nce System Medical Outcomes Study I ndia Ministry of Health Survey, the Demographic and Health Survey), were pre tested and then administered to randomly selected households in each community (total of ~113 surveys; see Appendix for survey module question s). Then, using the survey and initial health diary data, baseline and comparative assessment was conducted across each neighborhood in terms of:

PAGE 199

181 1) Socio economic conditions; 2) Health and well being conditions; and 3) Experiences with and barriers to accessing u rgent healthcare facilities Study Area Selection and Assessment of Infrastructure Conditions : Three neighborhoods where access to healthcare was an issue were selected within the Northeast District of Delhi, India. As shown in the figure below, the NE dist rict had the least percentage of institutional deliveries (less than 60%), indicating the potential lack of access to healthcare facilities for such purposes (DLHS, 2008) Figure VI. 3 % of Births as Institutional Deliveries by Delhi Districts T wo slum areas were then selected to distinguish between households with some access to suitable healthcare that is usable by the residents, compared to households with much more limited access. Infrastructure conditions an identified confounding factor for exploring access to healthcare links with health outcomes, were also ma pped in all three neighborhoods (Chapter 5). B uilding type, plinth level (i.e., housing surface relative to street surface, to address flooding risk), provision o f piped water, availability of toilets

PAGE 200

182 type of cooking fuels, density of sleeping conditions (e.g. number of persons per sleeping room), as well as exposure to hazards such as heat and flooding were all assessed. Socio Economic Conditions Assessment: As income is typically a sensitive survey question, which many households prefer to not respond to, we use monthly expenditures to form a basis of socioeconomic status in each of the three neighborhoods. We also consider educational attainment, length of resi dence, place of origin, and other socio demographic factors for each study area (as shown in Appendix). Health and W ell Being Conditions Assessment: In the survey, r espondent s' were asked to self assess their health status ; whether they felt their well bei ng or quality of life was heading in a good direction or not; and whether there were i llness es of household members in the past 30 days. Pr eliminary health diary results of self reported health outcomes for each household for the month of April, 2013 are a lso presented. For the health diaries, households were requested to maintain a monthly health log / diary report of sicknesses or deaths associated with their homes, using a monthly Hindi Language calendar page provided to each household. A medically t rain ed research assistant visited the sampled households once every two weeks to encourage them to maintain the lo g and write the sickness, dates, and the age / gender of household member that was sick. For those household s that did not have a literate person the medically trained individual assisted in this process The diary instrument ( or Hindi language household level calendar to keep record of sickness, care sought, dates) was field tested on 10 households in March 2013 and adapted with input from communi ty participants. The medically trained research assistant helped households to maintain the calendar of events, to collate

PAGE 201

183 data bimonthly and log reported cause s of death (where a death has occurred). Small i ncentives were also provided for participation f rom useful kitchenware to lunch boxes Assessment of Experiences With and Barriers to Accessing Urgent Healthcare: Drawing on the literature review summarized in Table 1, specifically McLafferty et al. 2002, five key ATH drivers are identified and assessed in this comparative study: Affordability : costs of urgent care services and ability to afford them; Accessibility : travel distance, travel time, costs to reach an urgent care facility, and the transportation modes available to reach the preferred facili ty; Availability : services available are able to address the serious illnesses experienced; Accomodation : hours of operation waiting times, emergency care application / registration procedures can appropriately accommodate all patients ; Acceptability : sa tisfaction or acceptabilityof the quality of care delivered A few additional variables identified in the literature are also utilized in the household survey, including awareness of healthcare p roviders length of time residing in neighborhood, and place o f origin While our initial literature review is still very far from comprehensive, the factors of involvement in local social networks length of time in a neighborhood, and place of origin were determined to be other factors worth exploring due to their potential for influencing levels of awareness, selection of providers, and overall access to healthcare. Results Results are presented in the same order of how the survey was conducted: 1) Socio economic conditions in each neighborhood, 2) Health and well bein g conditions

PAGE 202

184 3) Experiences with and barriers to accessing urgent healthcare Socio economic conditions : a focus on multiple expenditures incl uding those related to health, which can be calibrated w ith other surveys ( e.g. consumer ex penditure surveys (Govt. o f NCT of Delhi, 2008)) are used, and the Figure below (as shown previously in Sperling Thesis Chapter V) summarizes differences. Even t hough expenditures are relatively similar, the previous chapter demontrates household infr astructure conditions are widel y different and that priorities are also widely different, showing that it would be invalid to assume from income, asset ownership, or expenditures what priorities may be. Figure 4 indicates that Dilshad Gardens had higher monthly expenditures than Brijpur i, and Bripjuri higher than New Mustafabad. In addition, NM residents spent relatively more then BJ on rent / housing and a relatively equal amount for cooking fuels. Across all three areas, levels of employment are also relatively the same, although incom e levels while not quantified due to sensitivity were unequal based on the anecdotal discussions with participating households. Analyses also indicate New Mustafabad (NM) has lowest female to male ratios, levels of education, and expenditures on food, d aily goods, education of children, healthcare related, water, electricity, local transportation, and for telecommunications (e.g. phone / cellphone).

PAGE 203

185 Figure VI. 4 Socio Economic Factors By Study Area: Avg. Monthly Expenditu res Health and Well Being Conditions: Preliminary assessment of health conditions took place through use of health diaries and through use several survey questions (these analyses are in Appendix A). The findings from analyzing the self reported health out comes provided by each household in a health calendar included a 1.7 times higher disease incidence for NM residents relative to DJ residents, measured as a ratio of person sick days to total person days for April, 2013. However, these results are very pre liminary as the health calendars were collected for only 30 households in April. Assessing Experiences with and Barriers to Access ing Urgent Healthcare (ATH) : In figure below, we present the survey findings in terms of ATH conditions including questions s uch as how far households typically have to go for care for a serious health episode whether the costs are affordable, and how long the waiting times can be. Key findings include high SES residents in DG (n=20) having higher accessibility to care yet the y find quality of care less acceptable. Low SES respondents (BJ_n=48; NM_n=42) seem to have issues with waiting times, care affordability, travel time, distance, and costs.

PAGE 204

186 Figure VI. 5 Comparing Self Reported Access to He althcare in Three Neighborhoods

PAGE 205

187 The analyses shown demonstrates that high SES residents experience higher accessibility to care yet they find quality of care less acceptable versus low SES residents whom have larger issues with waiting times, care affordab ility, and travel time / distance / costs. Findings suggest future study needs to address travel time, wait times and costs which are issues for both low SES areas. Results indicate that the higher SES neighborhood, Dilshad Gardens, had less respondents stating having good or excellent quality of care ', in others words had less satisfaction with the costs and quality of care despite having impro ved ATH conditions. In addition, more than 45% of household residents in all three communities found hours of o peration to not be "good or excellent". For NM and BJ residents, results indicate waiting times, care affordability, and travel time/distance/costs were larger barriers to accessing urgent healthcare than DG residents. Interestingly, more BJ residents are also traveling further distances (more than 5km) to reach urgent care than NM/DJ, while time for reaching urgent care appears similar for both BJ and NM residents ( with NM residents not traveling as far, perhaps due to costs) As shown, the time needed to reach an urgent care facility was over 60 minutes for just over 60% of residents in BJ and NM while no residents in the high SES neighborhood, DG, faced this issue. It was also more costly for NM and BJ residents to reach care using the common metric of r esidents taking autorickshaws (~ 30% and ~50% of NM and BJ residents paid over 100 rupees to reach health facility, respectively). Statistical Significance of Findings Using T tests: While many conditions are shown in the figures to be different or simila r for the two low SES neighborhoods, T tests to determine whether statistically significant is important for determining if the data and results indicate average access to healthcare conditions for the low SES study areas to be

PAGE 206

188 different (research hypothes is) or similar (null hypothesis). To test the travel costs to reach care, we use a t test to explore the research hypothesis that NM and BJ have different access to healthcare conditions for this factor. Per Table 2, we find the average number of household s who have to spend >100 rupees for BJ are not statistically higher than for NM (t=1.53, df = 87, p< 0.1), so the null hypothesis (NM and BJ have similar access to healthcare conditions) is accepted in this case. These statistical tests indicate NM and BJ are statistically different on average in terms of medical care costs (t=1.96, df=79, p < 0.1), below the threshold (t = 1.671) and so the research hypothesis holds here with results significant at p < 0.10. Future analyses can compare across all ATH facto rs. Table V. 2 Access to Healthcare: Costs to Reach Care and Affordability of Urgent Care T Test #1: Costs to Reach Care >100 Rs. by Auto (1= Yes; 2 = No) T=Test #2: Affordable Urgent Care: Able to Pay for Care (1= Yes; 2 = No ) Data Summary New Mustafabad (NM) Brijpuri (BJ) Data Summary New Mustafabad (NM) Brijpuri (BJ) Number of observations 44 45 Number of observations 41 40 Mean 1.68182 1.51111 Mean 1.63415 1.825 Standard Deviation 0.47116 0.50553 Standard Deviation 0.48 765 0.38481 Variance 0.2169 0.2499 Variance 0.2320047 0.144375 Test #1: t=1.53, df = 87, p< 0.1, where t > or = 1.664 indicates statistically similar T est #2: t=1.96, df = 79, p< 0.1 where t > or = 1.664 indicates statistically different Qualitative Res ponse Assessment : The results in Table 3 below are presented as qualitative data and focus primarily on the p referred h ealth facilities for urgent and general medical c are (with listed advantages and disadvantages of each); Local Knowledge of Urgent Care Facilities and Neighborhood Health Programs '; the comments on Seeking Medical Care, Affordability of Services, & Ability to Pay '; and the comments on Acceptability of Urgent Medical Care Services'. In addition, we identify

PAGE 207

189 the L ength of Residence in City for Head of Households (By Generation and Years) that provide these comments as a potentially relevant factor that may be influential in terms of awareness about urgent healthcare providers and accessing relevant healthcare facilities Table VI. 3 Comparing Qualitative Responses on Knowledge and Experiences With Health Facilities along with Length of Residence Study Author, Year Qualitative Responses Preferred Health Facilities for Urgent and General Medical Care and Acceptab ility of Urgent Medical Care Services (including advantages and disadvantages of some hospitals) For both NM and BJ, preferred facilities in order are listed as first Delhi State government hospitals, then Municipal Corporation of Delhi hospitals, then New Delhi Municipal Council hospitals, followed by private hospitals and private clinics. For DG, most respondents list private hospitals, clinics, and doctors as the highest preference. Government Hospitals / Health Facilities : + "Value for money", "free t reatment available", "not costly"; +/ "Treatment is good, but very crowded"; "treatment is great, all facilities are there, but staff is not well mannered and long queue "services can be improved", "not always cooperative", "long queue of 3 to 4 hours" "the worst manners, services, and long waiting times" "not at all happy with attitude of staff; doctor was very bad, not attentive, and negligent", "behavior and attitudes can be improved", Private Hospital s / Clinics: + "Very satisfied with the trea tment; all my relatives are also treated there"; Local Knowledge of Urgent Care Facilities and Neighborhood Health Programs Note: The below highlighted responses are followed by length of residence in neighborhood for head of household by generation/year s in (): A. How did you come to learn of facilities you use? DG:"Doctor came for drycleaning at my shop and we became friends" (1/25) DG: "Born in locality, so knowledge of options since childrhood" (2/36) DG: "practitioners are in our same neighborhood" ( 2/44); DG: "have doctor in family" (2/28) BJ: "Doctor used to be neighbor" (1/28); BJ: "Officer gave information to father" (1/7) BJ: "From work"; "neighbors"; "relatives/in laws"; "since childhood"; NM: "Pvt Doctor told her two govt hospitals that woul d be good" (2/20) NM: "Clinic next to home" (1/7); "neighbors" (1/2); "since childhood" (1/5)

PAGE 208

190 Table VI.3 (cont'd.) Local Knowledge of Urgent Care Facilities and Neighborhood Health Programs B. Are there local health programs or campaigns going on in your area? DG: "At GRC bank accounts on community basis; regular OPD"(1/25) DG: "yes, but do not bother with them" (1/5) DG: "No knowledge of programs" (1/6; 1/5; 1/0.5) DG: "Polio camps by metro station" (1/10) DG: "Eye camps by Pvt. Hospital" (1/3) NM: Health and nutrition camp; law camp" (1/10); "eye camp (1/20) NM: "Program in savings for health" (1/10); "health camp" (1/35) NM: "No knowledge of programs" (1/5; 1/ 2; 1/7) NM: "GRC social welfare programs; Anganwallis (public health workers) GRC" (1/30) ; "Came to know that GRC facilitating AADHAAR card" (1/21) BP: "Anganavadi, mid day meal, camp for eye checkups" (1/7) BP: "GRC has health camps, medicines" (1/8); "knows of medical camps being conducted in neighborhood by GRC" (1/8 previously in NM for 20 yrs) BP: "Medical store; vocational courses" Responses on Seeking Medical Care, Affordability of Services, & Ability to Pay' "We are poor people and cannot afford even traveling to hospital"; "Health is a priority, but have to spend so much money for it"; "Poor value for money at govt hospital, and for private, we had to borrow"; "Costs have become an issue"; "price is a concern"; "fees to high"; "Costs are a big issue for us"; "it is better to get treatment at local doctor's place then to travel t o a big govt. hospital";"Free treatment available"; "Since we go to a govt hospital, we don't need a lot of money"; "although govt hospital has no fees, the transportation and medicine cost is too high"; "She needed to get an ultrasound done at govt hospi tal for 250 Rs, but due to lack of money, she missed her date for it" ; "Transportation and medicine cost are major barriers"; "have to buy medicine from outside"; "Had to sell off house to continue treatment of mother"; "Healthcare is a priority, and gov t hospital has good value and is not costly"; "Cost is a major concern, but since health is a priority, we do it anyway"; "no matter what, our household members go when they need care" As is shown in the table for some cases, knowing neighbors who 1) may be a doctor, 2) have had an positive or negative past experience with a specific government or private hospital, or 3) who may be able to assist in helping those who are ill reach an urgent care facility e.g. having a car or motorcycle can all be influe ntial. However, there are also households who do not know their neighbors as they've recently moved or are living without security of tenure. Given these factors, we also explore n eighborhood c hoic e motivations and social networks upon migration to neighbo rhood.

PAGE 209

191 Discussion This baseline assessment provides an initial starting point for further study in the three neighborhoods. A weakness of this study is that we are not yet able to link lack of access to healthcare and health outcomes. A well designed long term cohort study with households sharing similar infrastructure conditions but different access to urgent healthcare conditions could prove useful to build on this preliminary assessment, and may also help to more explicitly explore the extent to which ac cess to healthcare might shape health effects and outcomes (including illness, hospitalizations, and mortality). Other confounding factors that may need to be controlled and/or accounted for would be seasonal temperature, precipitation, and environmental e xposure variation. Conclusions This study utilizes bottom up field methods including household surveys and a one month health diary for initial baseline assessment comparing socio demographic conditions, health conditions, and experiences with and barrier s to access to urgent healthcare in three neighborhoods. The findings help to inform and motivate further assessment of access to urgent healthcare and how it shapes health outcomes. As presented, slum areas can have higher health risks due to different in frastructure deficiencies (Chapter 5) that are excaberbated by their additional lack of access to urgent healthcare, among various other services. However, further research is still needed to improve the quantification of relationships between health and i nfrastructure conditions, socioeconomic conditions, and specifically to address the extent to which lacking adequate access to healthcare in Asian cities is linked with severity of illnesses and health outcomes such as mortality in the case of traffic acci dents and various other hazards.

PAGE 210

192 Appendix A. Supplementary Analyses Table 1 below compares self reported health outcomes by slum vs. non slum and Table 2 below summarizes initial April health diary results for all three neighborhoods. Table 1. Summary com paring slum vs. non slum neighborhoods Neighborhood Person Sick Days and Type Total Person Days Total Sick Days / Total Person Days Non Slum: DG 26 1380 0.0188 Slums: NM / BJ 36 1110 0.0324 % Reduction for Non Slum Neighborhood Vs. Slum Neighborhoods 41 .9% Relative Risk of Sick Days 1.72 Table 2. Preliminary Summary of Self reported Health Outcomes by Neighborhood

PAGE 211

193 Assessments of H ealth and Well Being Conditions: As shown below in Figure 1, New Mustafabad (NM) has ranked their perception of their he alth as the lowest and have the highest % of respondents perceiving their life (or well being) as heading in not such a good direction. An interesting result is that the average household in Dilshad Gardens may perceive their health as worse off then someo ne in Brijpuri. Figure 1. Ranking of General Health Status in Household by Study Area Figure 2. Ranking of Recent Perceptions of Whether Life Heading in a Good Direction

PAGE 212

194 Socio Demographic Conditions : (Prior Preliminary Analyses, n_total = 64; n_ NM=21; n_BJ=23; and n_DG=20) Figure 3. Comparing Socio Demographic Factors By Study Area: Gender Ratios, Employment Levels, and Education In Figure 4 below, the proportion of expenditures on food and education of children provides a useful indicator for assessing the differences between the three study areas (e.g. for NM, food is >40% of total expenditures). Across all three neighborhoods, food, housing, and transportation make up more than half of total expenditures.

PAGE 213

195 Figure 4. Percentage Break down of Typical Monthly Expenditures by Study Area Figure 5 below shows average monthly expenditures. The numbers, once averaged, appear under reported and may not be a useful indicator as this was also a sensitive question for some households. If accura te, NM residents currently live off of spending close to one dollar per day and DG residents on just over two dollars per day.

PAGE 214

196 Figure 5. Average Monthly Household Expenditures By Study Area Health Related Quality of Life for Respondents (Prior Prelim inary Analyses, n = ~64): Figure 6 shows preliminary and more general indicators of health and well being, presenting the % of household respondents feeling ill in the past 30 days and of those days ill, the average number of days respondents were not well during the past 30 days. >40% of respondents reported feeling ill in past 30 days for all study areas; with over 50% and 60% of respondents in Brijpuri and New Mustafabad reporting illness, respectively. NM residents who were ill also spent a third of the entire month feeling ill.

PAGE 215

197 Figure 6. Presence and Duration of Illness for Respondents in Past 30 Days The questions on illness in past thirty days were followed up with asking about the number of days lost productivity versus the number of days that individual still worked when ill / tired. Figure 7 presents these results and demonstrates that NM respondents would continue working more than others while ill, perhaps due to the greater need for income as a survival net for their families. Figure 7. Days of Lost Productivity from Illness and Working Despite Illness Figure 8 below presents the average hours of sleep each night, sufficiency of that sleep, and factors for not sleeping enough. For all three study areas, the top three reasons for not sle eping enough in order was (1) Sickness; (2) Stress; and then (3) Busy Schedule. In many cases, this question provided a nice way of also learning about existing mental health and stress factors in each area (e.g. safety, drug problems among teenagers, etc)

PAGE 216

198 Figure 7. Sleep Conditions and Sleep Deprivation Factors as Indicator of Well Being An observation from survey administration is that this softer question about sleep and many of previous questions made individuals more comfortable discussing har der question/s asked next about hospitalizations in past year. It was a lso helpful to have a medically trained doctor involved in asking these questions on type of illness experiences. Figure 8 below presents the breakdown of types of illnesses in past y ear requiring hospitalization. For both DG and BJ, heart ailment was a larger cause of hospitalization, and for NM, abdomen pain, typhoid, accidents, and osteoporosis had higher levels of incidence of causing hospitalization.

PAGE 217

199 (n_NM=21; n_BP=23; n_DG=2 0) Figure 8. Preliminary Analyses on Type of Illnesses Requiring Hospitalizations in Past Year by Neighborhood

PAGE 218

200 Appendix B. Consent Form and Household Survey Instrument Used Consent Form : Invitation to Participate in Research Study by University of Colora do Denver (USA) Namaste. "My name is and I am working with the University of Colorado. This interview is for a research study that is being conducted by University of Colorado in collaboration with Urban Health Resource Centre You are being invite d to participate in a household survey and research study conducted by University of Colorado in Delhi, India. The purpose of this study is to explore health impacts related to infrastructure and environment for people living in Delhi neighborhoods. We pla n to explore relationships between civil urban infrastructures (e.g. water supply and sanitation, roads), environmental conditions (e.g., air and water pollution) and health outcomes (deaths, illness, visits to healthcare facilities) in your local communit y T he goal is to improve health and well being by designing better solutions in your community and city such as improving civil infrastructures, reducing pollution or making healthcare more accessible. This study is funded by the United States National Science Foundation and is being conducted by Joshua Sperling and Dr. Anu Ramaswami from the University of Colorado (Denver, USA) in cooperation with Dr. Siddharth Agarwal of the Urban Health Resource Centre (Delhi, India) We seek your participation in this study for three key data collection efforts: (1) Household survey: there will be two modules, the first takes roughly 15 minutes to complete a nd the second takes roughly 25 minutes : a. Module 1: Information about the household, your assessment of your health conditions, and the current situation for access to healthcare (15 questions) b. Module 2: Assessing Infrastructur e, Environment, and Extreme Wea ther Conditions (20 questions): researchers will document your views on 1) access to basic infrastructure services, such as adequate piped drinking water, sanitation, electricity, roads, and affordable / adequate housing; and 2) exposure to environmental c onditions such as air and water pollution and severe weather events such as extreme heat and cold and how it affects you. (2) Monthly household health diary to assess community health outcomes ; and (3) Discussion of study results for the design of potentially fe asible infrastructure, environment and healthcare solutions for improving health and well being in your community. If you are willing to participate in this study, we would like to work with you to have you fill in the survey. This will take you about 40 minutes with a small break in between Module 1 and Module 2 Your privacy is important and your household is not linked in any way to your responses. Names and addresses of participants will not be included in analyses or reports, nor will information about your household be shared with anyone. In addition, all of your answers will be kept confidential and will be aggregated with the responses of other households in your community. Therefore, when we report survey findings there will be no way to identi fy your household from any of the others. There are no disadvantages, if you decide not to participate or not to answer certain questions. You are free to participate or not to participate in this survey. It is important that you answer each question hone stly. Anything you share will be kept private. You do not have to answer any question that makes you uncomfortable. Again, you do not have to participate in this research study. If you decide to be in the study, you can change your mind later. If you par ticipate in today's survey, we will share a summary of the results when they are available in a future community meeting. This st udy will help the study team to design better solutions for providing better infrastructure, environment, and health services f or your community We hope that you will participate in this survey since your views are important. If you have questions, you can ask at this time, call us at (+91) 82859 78260, and reach us by email at XXXXX@gmail.c om Sincerely, UHRC field coordinator / community liason / translator (on behalf of Joshua Sperling)

PAGE 219

201 Module 1 Questionnaire Q1(a): General Health Status: the first question asks about your health and how you feel. Would you say that in general your heal th is: ( Prashn 1 : Sadharan Sharirik Sthiti : Aapki Sharirik Sthiti kaisi hain? / aap kaisa mehsoos kar rahe hain? Please read and circle only one choice below 1 Poor 2 Fair 3 Good 4 Very good 5 Excellent ( Neeche likhe gaye uttaron ko padhiye aur jo uttar aapko sa hi lage uske saamne likhi gayi number par circle /vritt/gol chihn banayiye) 1 Bahut Accha 2 Accha 3 Pata nahin 4 kharab 5 Bahut kharab Do not read these responses: 6 Don't know / not sure 7 Refused In Pratikriyaon ko nahin padhen 6 nahin pata/ nischit nah in hoon 7 Jawab dena manzoor nahin Q1(b): When you think of your life, is it going in a good direction? 1 Y 2 N Please describe: Q2a: Recent Health Related Quality of Life / Number of Sickness Days: (Shwasthya se judi jivan ki gunvatta/ ashwasth dinon ki sankhya) Now thinking about your physical health (includes illness and injury that affects your body), for how many days were you ill or not feeling well during the past 30 days ?( Agar aap ki sariric sthiti ke bare me baat Karen, kya aap bata sakte hain k i peechle 30 dinon me kitne din aap bimar rahen ya swasth anubhav nahin kar rahe the?) ___ Number of days (__) None Note: If None is the response skip question 3. ___ rog mukt dinon ki sankhya (__) koi aswasth din nahin Note: A gar din rog mukt nahin hain, toh kripya seedha 3 number sawal ka jawab dijiye Q2b: Impact of Sickness on Activities : Following up on Q2: Of the days that you were ill, for about how many days were you able to do your regular activities such as going to wo rk, going to play, taking care of the household and so on? ( Bimari ka dainik kaam kaaj pe asar Following up on Q2: Bimari ke dauran aap kitne din apne dainik kaam kaaj jaise ki naukri pe jaana, khel kud, ghar ka kaam kaaj bagerah kar sakte the? ) ___ Number of days (__) None ___ Dinon ki sankhya (__) koi bhi din nahin Q2c: How many days in this period did you feel tired / ill and didn't want to go to work, but went anyway? Q3: Number of household (HH) members: _____ Q4a : What are the income generating activities for people in your household how are members making money? ( Aapke ghar ke sadasya rojgaar ke liye kya kaam karte hain / unki aamdani ki jariya kya hain?) _________________________________________________________ _________________

PAGE 220

202 Q4b: What is your estimated monthly HH expenditure in Rs.: __________ ( Aapki mahine ka kharcha kitna hota hain?) ________ Q4c: Is your income enough to cover basic needs? (kya aapki aamdani ghar ke zarooraton ko pura karne ke liye par yapt hain?) Yes ( haan) | No (nahin) Q4g: How many people contribute to your household income? __ Write in number of people: (Ghar ki aamdani me kitne logon ki yogdaan hota hain? __ logon ki sankhya likhiye) Q4h: Over the last 1 year, what was the a verage monthly household expenditure on the following? fiNys ,d lky esa] vkSlru ifjokj dk ekfld [kpkZ bu en ij fdruk Fkk i Food [kkuk ii Daily Goods jkstkuk dh oLrq, iii Rent (if applicable) fdjk;k iv Education of children cPpksdh f'k{kk ij v Health care related LokLF; laca/kh ns[kHkky ij vi Water fctyh vkSj ikuh ij vii. Electricity viii. Cooking fuels ix. Local transportation LFkkuh; ;krk;kr esa x. P hone / cell phone xi. Total average monthly HH expenditure vkSlru ifjokj dk dqy ekfld [kpkZ Rs. Rs. Rs. Rs. Rs. Rs. Rs. Rs. Rs. Rs. Rs. Q4i: What percentage of your income is going to the mont hly expenditure roughly? 1 100% 2 90% 3 80% 4 70% 5 60% 6 50% of income going to these expenditures Q4j: Which of these expenditures are most important for your life and improvements in your life and that of your household members? Q4k : During the last 1 year, what was the other major household expenditures other than regular household expenditure on the following (described below)? fiNys ,d lky esa] ifjokj ds fu;fer [kpsZ ds vykok] bu enksa ij fdruk [kpZ Fkk \ i. Marriage ceremonies 'kknh ds lekjksg ii. Religi ous festivals /kkfeZd R;kSgkj iii. Visits outside Delhi city fnYyh 'kgj ls ckgj vkxs tkus esa iv. Illness / accidents vkikr dkyhu chekjh@ ,DlhMsUV v. Birth in the family ifjokj esa tUe ij vi. Death in the family ifjo kj esa e`R;q ij vii. Total yearly HH expenditure dqy okf"kZd ikfjokfjd [kpkZ viii. Other i. Rs ii. Rs. iii. Rs. iv. Rs. v. Rs. vi. Rs. vii. Rs. viii. Rs. Q5: : Educational Attainment: What is the level of education you completed? shaikshik/ shiksha se judi yogyata: ( Aap kahan tak shikshit/padhe likhe hain? ) 1. Can't Read and Writ e ( Padhna Likhna nahin aata ) 2. Can Read and Write ( Padhna Likhna aata hain ) 3. Can read and write, but without formal schooling ( Padhna Likhna aata hain, par Koi Aanusthan/ Vidyalay nahin gaye) 4. Can Read and write with Formal Schooling, below Primary ( Padh na Likhna aata hain par Prathmik/ buniyadi vidyalay pura nahin kiya) 5. Primary (4 th grade completed) ( Prathmik / chauthi kaksha Pura kiya' ) 6. Middle (7 th grade completed) ( Madhyamik / saatwi kaksha pura kiya' ) 7. Secondary (10 th grade completed) ( Ucch Mad hyamik / daswi kaksha pura kiya' ) 8. Higher Secondary(12 th grade completed) ( Ucchtar Madhyamik / barahwi kaksha pura kiya' ) 9. Diploma / Certificate Course ( Diploma ya Certificate course kiya ) 10. Graduate & Above (BS/BA/BComm/BTech and above) ( Snatak ki ya fir uske aage tak padhai ki

PAGE 221

203 BA/BSc/BCom/BTech aur uske aage tak padhai ki ) ACCESS TO URGENT MEDICAL CARE ( AATYAVASHYAK SWASTHYA SEWAON KI UPLABDHTA): This is the second part of the survey. The first part was general questions about you and your hous ehold. This next part is about access to urgent medical care. For access to urgent medical care, we are interested in serious illness which is affecting your daily life and you feel requires you to go to a hospital. For example, this can include acute dia rrhea, heat stroke, high fever, accident or injury. BACKGROUND AND TRAINING FOR TRANSLATOR: DEFINING ACCESS TO URGENT MEDICAL CARE We define and characterize access to urgent medical care primarily in terms of variables identified in McLafferty et al (200 2): Affordability : the price of the services with regard to people's ability to pay (Note: both income levels and insurance coverage can be critical aspects of affordability); Accessibility : the geographic barriers, including distance, transpor tation, tr avel time, and cost. Availability : the supply of services in relation to needs e.g. are the capacity and types of services adequate to meet the urgent care needs? Accommodation : the degree to which services are organized to meet clients' needs, including hours of operation, application procedures, and waiting times (oft en referred to as the queue'); Acceptability : which refers to the client's views of health services and how service providers interact with clients. Views on willingness to use particular health services can be shaped by gender, culture, ethnicity, sexual orientation, etc. Sense of comfort and satisfaction in receiving services can also depend on if clients are well treated, providers and clients are able to communicate openly, and if provi ders are confident abou t the quality of care delivered; Awareness of Healthcare Providers : refers to knowledge of medical care options, which may depend on the length of time in a neighborhood and social networks that may influence access to urgent medical care. Q6a: Options for Accessing Urgent Medical Care: ( Atyavashyak Shwasthya Sevaon ke liye vikalp )

PAGE 222

204 When members of your household have a serious illness, where do they generally go for treatment? You can select more than one of the options below Plea se circle all appropriate numbers. If Other, please specify below: ( Jab aap ke parivaar wale gambhir roop se bimar hote hain, toh apne ilaaj ke liye who kahan jate hain? Aap dayen or diye gaye vikalpon me se ek se jyada vikalpon ko bhi chun sakte hain. K ripaya uchit sankhya par gol chihn banayiye. Agar koi anya vikalp hain, toh usse neeche diye gaye jagah pe ullekh kijiye ) PUBLIC MEDICAL SECTOR DIST HOSPITAL ................................ ................................ ..... 11 ( JILA ASPATAL ) GOVT. MUNICIPAL HOSPITAL ................................ .......... 12 ( SARKARI NAGAR NIGAM KSHETRA ASPATAL ) GOVT. DISPEN SARY ................................ ............................ 13 ( SARKARI DISPENSARY ) UHC/UHP/UFWC ................................ ................................ .... 14 ( SAHARI SWASTHYA KENDRA/ SAHARI SWASTHYA POST/ SAHARI PARIVAAR KALYAN KENDRA ) CHC/ SUB CENTRE ................................ ............................... 15 ( SAMUDAYIK SWASTHYA KENDRA/ UP KENDRA ) ANGANWADI/ICDS CENTRE ................................ .............. 16 ( ANANWADI/ ICDS KENDRA ) GOVT. MOBILE CLINIC ................................ ....................... 17 ( SA RKARI MOBILE CLINIC ) OTHER PUBLIC SECTOR HEALTH FACILITY ................. 18 ( ANYA SAMUDAYIK SWASTHYA SEVA ) NGO OR TRUST HOSPITAL/CLINIC .............................. 21 ( BESARKARI PRATISTHAN/ TRUST ASPATAL/ CLINIC ) PRIVATE MEDICAL SECTOR ( PRIVATE/ NIJI SWASTHYA PRTISTHAN ) PVT. HOSPITAL ................................ ................................ ..... 31 ( PRIVAT E/ NIJI ASPATAL ) PVT. DOCTOR/CLINIC ................................ ......................... 32 ( PRIVATE / NIJI DOCTOR CLINIC ) PVT. PARAMEDIC ................................ ................................ 33 ( PRIVATE / NIJI PARAMEDIC ) VAIDYA/HAKIM(UNANI)/ HOMEOPATH..34 ( VAIDYA,HAKIM/YUNANI/HOMEOPATHY ) PVT UN QUALIFIED DOCTOR ................................ ............ 35 ( NIJI?PRIVATE ASIKSHIT CHIKITSHAK ) T RADITIONAL HEALER ................................ ...................... 36 ( PARAMPARIC BYAM_KARTA ) PHARMACY/DRUGSTORE ................................ .................. 37 ( PHARMACY/DAWAKHANA ) DAI (TBA) ................................ ................................ .............. 38 ( DAI ) OTHER PRIVATE SECTOR ................................ .................. 39 ( ANYA NIJI/PRIVATE PRATISTHAN ) HEALTH FACILITY ................................ ............................... 40 ( SWASTHYA SUVIDHA ) OTHER ( ANYA ) SHOP ( DUKAN ) ................................ ................................ ..... 41 | HOME TREATMENT ( GHAR PAR UPCHAR ) 42 OTHER( ANYA ) ................................ ............................... 96 PLEASE SPECIFY ( KRIPAYA ULLEKH KIJIYE ): __________ ____________________ Q6b: Please give us the names and location of facilities which you mostly use: ( Kripaya aap jin swasthya suvidhawon ka ya Pratisthanon jyada istemaa l karte hai unka naam aur jagah bataye ) ______________________________________________________________________________________ Q6c : On a scale of 1 to 5, how well are these facilities able to meet your medical needs? ( Aap ki swasthya raksha ki zarooraton ko pura karne me yeh suvidhayen kis hadd tak saksham hain, 1 se 5 ke beech mulyankan Karen) 1 Never 2 Rarely 3 Sometimes ( Kabhi Kabhi ) 4 Most of the time ( Jyadatar ) 5 All the time ( Hamesha ) Q6d : KNOWLEDGE : How did you come to learn of these facilitie s that you use? ( JAANKAARI : Aap jin suvidhawon ka istemaal karte hain, unke bare me aap ko kahan se pata chala? ) Please describe: ( Kripaya Likhiye ) _______________________________________________________________________________ The next few questions ar e about the general accessibility of these health care facilities to meet your needs: Q7 : Were there serious health episodes in your family that you thought required urgent medical care over last 12 months that were dangerous or interrupted your life and kept you from your usual activities?( Kya aapko lagta hain ki peechle 12 mahino me aapke parivaar me koi gambhir swasthya se judi samasya hui

PAGE 223

205 hain jo janleva ho jiske karan dainik kaam kaaj me vadha aayi ho aur jiske liye aapko atyavashyak swasthya se wa ki zaroorat hui ho) 1 Yes( Haan ) | 2 No ( Nahin ) If yes, please describe the different serious episodes: ( Agar haan, toh kripaya un gambhir swasthya se judi samasyao ke bare me vistar se batayen) __________________________________________________ ____________________________________ Also, why did you consider these episodes serious (requiring hospitalization)? Please comment below: ( Yeh bhi batayen ki aapko woh samasyayen gambhir kyun lagi ? Kya aspatal me bharti ki aavashyakta hui? ) _________ _______________________________________________ Q10a: Medical Care Affordability and Ability to Pay : On a scale of 1 (worst) to 5 (best) how would you rate the affordability of medical care services provided to you. ( Swasthya sewa lene ke liye Samarth a ur paise aday karne ki kshamta : Aapko jo swasthya sevayen uplabdh hain, aapki unko upbhog karne ki kshamta ke aadhar par neeche diye gaye vikalpon ko 1 (sabse bura) se 5 (sabse accha) tak mulyankan kijiye) 1 Poor "My family needed care but I could not afford it" ( Bura Mere parivaar ko suvidha ki zaroorat thi lekin main lene me saksham nahin tha) 2 Fair "My family may use their services, but cost was a big barrier" ( Thik thak Mera parivaar unki sevao ko apna sakte hain, lekin keemat bahut jyada th a) 3 Neutral No opinion ( Koi rai nahin ) 4 Good "My family members go when they need care, cost was a small issue" ( Accha Mere parivaar ko jab bhi zaroorat hoti hain hum swasthya sevao ko apnate hain keemat itna bada mudda nahin hain ) 5 Excellent "Excellent value for money: we get a lot of care and quality care for the money" (Bahut accha keemat ke hisaab se bahut hi uttam matra ki syasthya sevayen uplabdh hain) Q10b: Please comment on your selection below: ( kripaya apne chune hue vikalp par tippani kijiye ) __________________________________________ ______________________ Q11a : Accessibility of Medical Care: What is the distance to your health facility (doctor's office, clinic, health center, or other place) that you usually go to if you hav e a serious illness and need to be hospitalized ? ( Swasthya sevao ki pahuch : Jab aap ko koi gambhir bimari ho aur aap ko Aspatal me bharti hone ki zaroorat aan pade, toh aap ilaj ke liye jis Doctor ya clinic ya aspatal me jaate hain, woh aap ki ghar se kitni duri pad hain? ) 1 Within a half kilometer (km) ( aadha km ke andar ) 2 Within 1km ( 1 km ke andar ) | 3 Within 3km ( 3 km ke andar ) | 4 Within 5km ( 5 km ke andar ) 5 Within 10km ( within 10 km ) | 6 More than 10km ( 10 km se jyada ) | 7 Doctors come to my h ome ( mere ghar Doctors aate hain ) Q11 b: How do you usually access medical care in case of a serious illness? ( Gambhir bimari hone pe aap aam taur par kis tareeke se ilaaj ke liye jate hain ?) 1 Autorickshaw ( autorickshaw ) | 2 Cyclerickshaw ( Cyclericks haw) | 3 Bus ( bus ) | 4 Medical transportation service ( Chikitsha paribahan sewa ) | 5 Friend/relative brings you ( dost/rishtedaar le ke aate hain ) 6 Drive yourself ( khud gaadi chalake aate hain ) | 7 Walk ( Paidal jaana ) | 8 Doctor comes to home ( Doctor khud ghar pe aate hain ) | 9 Other ( anya ) Q11 c: Of the below list, w hat transport mode /s are available for you to use to get to your health facility in case of a serious illness ? Please circle all available. ( Gambhir bimari hone par swasthya Kendra/Aspatal pa hunchne ke liye aap ke paas neeche diye gaye list me se kon kon se sadhan uplabdh hote hain ? ) 1 Autorickshaw ( autorickshaw) | 2 Cyclerickshaw ( cyclerickshaw ) | 3 Bus ( Bus ) | 4 Medical transportation service ( swasthya parivahan seva ) | 5 Friend/relative brings you ( dost/rishtedar le aate hain ) | 6 Drive yourself ( khud gaadi chalake aate hain ) | 7 Walk ( paidal ) | 8 Other ( anya)

PAGE 224

206 For those modes available, please estimate the costs of each trip to access care in case of serious illness, and the time it take s. Transport Mode 1: ______; __ rupees on average; ____ time it takes Transport Mode 2: ______; __ rupees on average; ____ time it takes Transport Mode 3: ______; __ rupees on average; ____ time it takes ( Agar aap ke paas kuch saadhan uplabdh hain, toh un sadhano ke jadiye gambhir bimari me ilaj ke liye jane me aap ko kitna kharcha hota hain aur kitna samay lagta hain, kripaya neeche likhiye Yatayat ki saadhan 1..rupaye aushatansamay lagta hain Yatayat ki saadhan 2..rupaye aushatansa may lagta hain Yatayat ki saadhan 3..rupaye aushatansamay lagta hain ) Q12a : Accommodation Ability to Meet your Medical Care Needs : On a scale of 1 (all the time) to 5 (never), how often are the medical services available to meet your needs for a serious illness? ( Swasthya suvidhao ki uplabdhta aapki zarooraton ki hisab se suvidhaon ki maujudgi : Aapki gambhir bimari ki dauran swathya suvidhayen kis hisab se maujad hain neeche diye gaye vikalpon me se chuniye 1 Never | 2 Rarely | 3 Somet imes | 4 Often | 5 All the time 1Kabhi nahin | 2 Bahut kum | 3 Kabhi Kabhi | 4 Aksar | 5 Hamesha Please explain giving examples of when they met or did not meet your needs: ( Kripaya Udaharan ke sath batayen kab Sevayen aap ki zaroorat ke hisab se maujud t he aur kab nahin ) ______________________________________________________________________________________ ________________________________________________________________________________ Q12 b: On a scale of 1 to 5, how well do the hours of operation cover your healthcare needs? Is the facility open when you need it? ( Aap ke swasthya ki zarooraton ke liye seva ya suvidhayen prapt karne ka samay kya aapke hisab se kaafi hain? Aapko jab zaroorat ho toh kya yeh jagahen aapko khuli milti hain? Neeche diye gay e vikalpon me se chuniye. 1 Poor, (e.g. rarely available when I need it) ( Kharab, e.g. jab zaroorat hoti hain toh nahin milti) 2 Fair ( Thik thak ) 3 Good ( Accha) 4 Very Good ( Bahut Accha ) 5 Excellent covers my needs very well (open all the time) ( Bahut jyada accha) Please comment on your selection: ( Kripaya apne chune huwe v ikalp pe tippani kijiye ) _____________________________________________________ Do the facilities adhere to the prescribed timings of their hours of operation? Yes | No (Kya Suvidhayen apni vidhigat samay ke mutabik kaam karten hain ? Haan/na) Q12 c: On a scale of 1 to 5, how easy was the application / registration procedures ? ( registration / panjikaran ki prakriya kitni aasaan thi, neeche diye gaye vikalpon me se (1 se 5 tak) chuniye.) 1 Poor ("very complicated") ( Kharab)/ bahut muskil 2 Fair, ("a l ittle difficult") ( Thik Thak ) / thoda muskil 3 Good ("just OK") (Accha ) / thik 4 Very good ("easy") (Bahut Accha/) /aasaan 5 Excellent ("very e asy no problems at all") ( bahut jyada accha ) /bahut aasaan Please comment on your selection: ( Kripaya apne chune huwe vikalp pe tippani kijiye) Q12 d: On a scale of 1 to 5: how long was the waiting time / the queue' ? (this can include at the registrat ion desk, in the doctor's chamber, in waiting for tests / diagnoses / pharmacy, etc) (Aapko Swasthya seva ki jagah par aapki baari aane tak lagbhag kitna der pratiksha karna padta hain ya line lagani padti hain ? neeche diye gaye 5 vikalpo me se chune 1 Poor ("very long") (kharab / bahut der ) | 2 Fair, ( thik thak / thodi der) | 3 Good ( accha / Jaldi ) 4 Very good ( bahut accha / bahut jaldi) 5 Excellent ("very quick") ( Bahut Jyada Accha / bahut jyada jaldi )

PAGE 225

207 Q12e: Please estimate the total waiting ti me and wait at various stages of being at the hospital: _______________ (Stage 1: e.g. Initial Admission, Registration or if other, describe: _________________) _______________ (Stage 2: ________________________________) (i.e. in front of Doctor's chamber) _______________ (Stage 3: ______________________________________) _______________ (Stage 4: ______________________________________) TOTAL WAITING TIME: _____________________ Q12f: Did you ever feel the delay in reaching hospital or waiting in hospital negatively affected your health? Y / N Please describe: _______________ Q12g: Were you ever turned away? Y / N Please describe: Q12h : Was wait ever too long when life or death situation ? Y / N Q13a: Reliability / repeat questions below (we use differen t way of asking same questions as above to ensure reliability in responses) Q13b : During the past 12 months, have you or anyone in your family delayed seeking medical care due to costs ? 1 Yes | 2 No; Please specify which reasons in particular: (Peechle barah mahine me kya aap ya aapke parivaar me kisine upar diye gaye kisi karan ke wajah se swasthya sevayen lene me der lagaye? 1 Haan | 2 Naahin Kripaya un karanon ka ullekh kijiye ) ___________________________________________ Q13c : During the past 1 2 months, have you or anyone in your family delayed seeking medical care due to health facility too far or there being no transportation ? 1 Yes | 2 No; Please comment on which: (Peechle barah mahine me kya aap ya aapke parivaar me kisine upar diye gaye k isi karan ke wajah se swasthya sevayen lene me der lagaye? 1 Haan | 2 Naahin Kripaya un karanon ka ullekh kijiye ) Q14a : Length of Residence in City : How long has the head of the household been staying in this city ? ( Parivaar ka Mukhiya kitne samay se is Shahar me reh rahe hain? ) By Generation : Peedhiyon ki hisab se By Number of Years Barson ke hisab se If < 1Yr. write Months Agar < 1saal, toh Mahine me likhiye Were you born here? Yes / No (Kya aap ka janam yehi par huwa tha ? Haan/Nahin) Q14b Len gth of Residence in Neighborhood: How long has the head of the household been staying in this neighborhood ? ( Parivaar ka mukhiya is mohalle me kitne samay se reh rahe hain? ) By Generation : Peedhiyon ki hisab se By Number of Years Barson ke hisab se If < 1Yr. write Months Agar <1saal, toh Mahine me likhiye Were you born here? Yes / No (Kya aap ka janam yehi par huwa tha?)

PAGE 226

208 Q14e: Location Choice Decision Making Factors: Reason for settling in this specific neighborhood ? ( Is mohalle me basne ka koi khas karan ?) ____________ _________________________ What are the main reasons for living here ? (yahan rehne ka mukhya karan kya kya hain ? ) ______________________________________ Q14g: If you recently moved to this neighborhood, did you know people here? Yes / No (Agar aap haal h i me is muhalle me aaye hain, toh kya aap yahan ke logo ko jaante the?) _____________________ Did you come to this neighborhood because you knew someone here Yes or No? (Kya aap is mohalle me isliye aaye hain kyunki aap yahan kisiko jaante the Haan ya naa ? ) _______________ Q15 Local Knowledge : Could you tell a newcomer to this neighborhood the best place to get medical care if they had just moved to this neighborhood and had a serious illness? Yes | No; If yes, please describe the available options and their quality: ( Sthaniya jaan kari : Kya aap koi aise byakti ko jo naya naya aapke muhalle me aaya hain aur gambhir roop se bimar ho gaya hain, yeh bol sakte hain ki swasthya seva paane ke liye yehi muhalla sabse acchi jagah hain ? Haan/Naa, Agar Ha an', toh maujud vikalpon ke bare me aur unki gunvatta ke bare me vistar se batayen) ______________________________________________________________________________________ If newcomer to neighborhood asked about urgent medical care options available in you r neighborhood, would you respond : (Agar aapke muhalle me koi naya naya aaya ho aur aapko muhalle me maujud swasthya sevaoon ke bare me puchta ho, toh aap kya jawab dena chahenge ?) I have: 1 No knowledge on this subject (I'm not aware of any decent opt ions') (is baare me kuch nahin pata, kisi acchi jagah ke baare me nahin pata) 2 Minimal knowledge on this subject (I think there might be at least one option') (Is Vishay me bahut kum jaankaari hain, /mujhe lagta hain, ko i ek vikalp toh hogi hi) 3 Some knowledge on this subject (I'm aware of at least one option') (Is Vishay me thodi si jaankari hain/ek vikalp ki jaankari hain mujhe) 4 Good knowledge on this subject (I'm aware of some decent options') ( Is bare me achi jaa nkari hain mujhe/ kuch achi vikalpon ke bare me pata hain mujhe ) 5 Excellent (I'm aware of many decent options') ( Bahut accha / mujhe kaafi sare sahi vikalpon ke bare me jaankari hain) Q16 : Are you aware of any health programs or campaigns going on in your community ? ( Kya aap ke muhalle me chal rahi koi swasthya karyakram ya abhiyan ke bare me aapko jaankaari hain? ) 1 Yes 1 Haan 2 No 2 Naa If yes, please describe: ( Agar Haan', toh vistar se batayen ) ___________________________________ ___________________________________________________ Thank you for completing this survey. You have now completed Module 1. (Yeh survey pura karne ke liye dhanyabaad, aapne abhi Module 1 katam kiya ) If you have any questions you can call the research c oordinator, _____ ______, at xxxxx.xxxxx

PAGE 227

209 CHAPTER VII. VII. WHAT IS KNOWN AND WHAT IS NEEDED TO ESTIMATE HEALTH BENEFITS OF INFRASTRUCTURE INTERVENTIONS CASE STUDY OF DELHI, INDIA Abstract T his chapter explores what is known and what is needed to estimate h ealth benefits of infrastructure interventions for the case of Delhi, India. This effort utilitizes both bottom up and top down analyses along with pri or published works from the IGERT Ramaswami research group, literature review, and collected mortality da ta. This chapter seeks to integrate and develop an understanding of knowledge relevant to informing opportunities for measuring health benefits of infrastr ucture interventions. First, this chapter explores what is known and what is needed to estimate healt h benefits of infrastructure interventions for the case of Delhi, India. Then, this chapter focuses on preliminary applications of what we do know that can be used for computing first order benefits of multiple infrastructure interventions for health (e.g. avoided mortality), while also exploring who may benefit more or less by age, gender, socio economic status. Challenges are identified in linking health to GHG risk reduction benefits, specifically due to uncertainty of how health scenarios shape air poll ution reduction geographically, as well as opportunities in some infrastructure sectors such as transportation perhaps easier to assess then power generation (due to uncertainty in which power plants reductions come from). Such discussion offers initial in sights for help ing guide decision making in cities and future modeling of infrastr ucture environment health linkaages

PAGE 228

210 Introduction Urban infrastructure related hazards globally such as substandard housing, crowded living conditions, food and water safe ty, inadequate sanitation and solid waste disposal services, air pollution, and road accidents (WHO, 2010) all have important associated health risks, especially in rapidly growing Asian cities such as Delhi, India where up to 19% of all recorded deaths ma y be infrastructure related (Sperling & Ramaswami, 2013). While building and upgrading civil infrastructures (e.g. water supply, drainage, energy, transportation) can make cities healthier to what extent? A case study of Delhi is conducted in this thesis to further explore how civil infrastructures (e.g. energy, water, sanitation, transportation, housing) and environment conditions (air pollution) shape health outcomes in cities. While it is well known that income is a strong predictor for mortality for example, the National Family Health Survey 2006 shows strong correlation between lower income and under 5 mortality rates for children the extent to which income may be a surrogate for other factors like literacy, access to health care, access to infrastr uctures, or reduced pollution exposure is still not well understood. The research question explored in this chapter is 1) what do we know and what is needed to estimate health benefits of interventions related to infrastructure and infrastructure related environmental factors that shape current urban health outcomes?' Such discussion can help improve the evidence base for decision making on priority infrastructure interventions that can improve health in rapidly growing cities like Delhi, India in ways tha t may also have co benefits for greenhouse gas mitigation.

PAGE 229

211 Literature Review T he three tables below help to summarize current knowledge for developing co benefits of potential infrastructure interventions for the case of Delhi, India. Table 1 presents h ow individual infrastructure affects health. Table 2 explores infrastructure related health outcomes and the strength of the evidence specific to mortality health effect estimates (HEEs) and HEEs available by different social factors. Finally, Table 3 is o n specific HEEs useful to estimating first order health benefits. Findings show an emerging knowledge base on energy and air pollution related health effects, and that knowledge gaps remain for other infrastructure sectors including traffic accidents, wate r sanitation, waste, parks and food. Appendix A also addressing initial methodological questions and insights.

PAGE 230

212 Table VII. 1 What We Know and What is Needed to Explore Health Benefits of Infrastructure for the Case of Delhi, Ind ia Sector What is Known (specific to Delhi, India, or globally if not in India) What is Needed (for Delhi, India) Energy and Air Pollution Short term dose response studies for Delhi, India by Cropper et al. 1997 and Rajarathnam et al., 2010 by age, gender Many additional studies exist in Asian, US, and other Global cities focused on multiple pollutants with findings for different subpopulations and for multiple health outcomes. Identified studies do not consider health effect estimates in terms of sensiti vity related to inadequate access to healthcare Accidents Risk factors for road fatalities divided into factors influencing exposure to risk, crash involvement, crash severity, and severity of post crash injuries. Limited findings in Delhi primarily ex posure fatality findings from other Asian city, national, global studies. Delhi / India transport modal split and traffic fatality data by mode available to compute mode shift interventions. Need number of incidents per unit population by each transport mo de. Estimates for avoided or increased exposure to risk by shifting from road use to rail, for example, and data on access to quality hospital emergency room care post crash involvement Water Supply and Sanitation Morbidity and mortality relative risks (R R) associations with water sanitation infrastructures exist globally, but does not address confounders so relative risk is small (i.e. RR = 1.2 (Cairncross, 2010, Gunther and Fink, 2010) relative to what we're finding (RR = 7.0 for India; RR=4.4 for Delhi) Inadequate control of confounders a problem in all but a few studies. Need whole population mortality data by infrastructure, SES & literacy conditions (Sperling Thesis, Chapter IV) Waste General associations (often not quantitative) on u rban malaria an d other vectorborne mortality effects; case specific dose responses may be known on exposure s to waste burning / toxic sites and cancer / other health risks. Waste incineration 10 key pollutants having greatest health impacts based on environmental persi stence, bioaccumulation, amount emitted and toxicity: cadmium, mercury, arsenic, chromium, nickel, dioxins, PCBs, PAHs, PM10, SO2 (Rushton, 2003) Waste burning health studies exist, but not specific to Delhi, India; Many studies do not know individual exp osure data and data on confounders such as SES and literacy often not addressed. Urban Parks Cooling effects can reduce urban heat related excess mortality and recreational areas can be of benefit toward regular physical activity reducing risk of prematur e death; reducing risk of coronary heart disease, hypertension, colon cancer, non insulin dependent diabetes, etc Too complex, as multiple infrastructures can impact cardiovascular disease. Food General associations between health and food poisoning, unh ealthy food relative risks for malnutrition / obesity, child morbidity, and BMI relative to nutritional status based on socioeconmic status and rural and urban residence (specific to Indian women by Bharati et al., 2008 ) Food poisoning deaths by location f or Delhi, India; effectiveness of acute malnutrition interventions on mortality

PAGE 231

213 Table VII.2 (cont'd.) Housing Health effect estimates for overcrowding and TB mortality known for Sao Paolo, Brazil (Antunes & Waldman, 2001) and on TB prevalence in India f or homes with overcrowded sleeping conditions and without a separate cooking space (Aggarwal, 2010). Literature suggests association of overcrowding and TB in both children and adults, with overcrowding in childhood affecting aspects of adult health. Assoc iations between overcrowding / physical health / prevalence of morbidity have been well documented, yet limited evidence exists on mortality health effects, specific to Asian cities. Can trace prevalence of TB cases in Delhi back to home s and housing condi tions to address need for relating Delhi overcrowded sleeping in homes to health data based on visits to hospitals or TB and other forms of airborne infrectious disease mortality so to develop relative health effect estimates for TB and other airborne infe ctious diseases in Delhi / Indian cities. Table VII. 2 Literature Review Integrating Multiple Infrastructure Related Health Risk Categories and Health Effect Estimates

PAGE 232

214 The Table 2 summary table is based on literature review conducted in Sperling Thesis, Chapter III, with the water related HEEs based on analyses in Sperling Thesis, Chapter IV that addresses under age five (child) mortality, but not adult mortality. If to scale up DHS child mortality data to adult populations, o ne approach would be to use the available Delhi whole population mortality by age dataset, and apply an adjustment factor based on the ratio of child mortality to adult mortality rates per 1000 births or 1000 population. This is just one way to scale up child mortality data from DHS child mortality data to adult populations. However, challenges remain in scaling up for specific cause related mortality. In terms of adult diarrheal mortality, for example, the below literature and available DHS / Delhi morta lity data highlights the needs and opportunities for scaling up: A synthesis by Walker and Black, 2010 highlights the need for improved diarrhea specific mortality and morbidity data for different age groups including older children and adults (as deaths f or this age group can also be quite high). Lamberti et al. 2012 also identifies the need for improved understanding of diarrhea duration and severity across age groups due to significant global diarrhea morbidity, and improved treatments leading to decrea ses in diarrhea mortality. A study by Boschi Pinto (2008) indicates deaths from diarrhea of children aged less than 5 represent approxim ately 19% of total all cause child deaths. The Medical Certication of Cause of Deaths report for Delhi, India roughly m atches this in indicating that of the total institutional deaths of 1945 (1146 males and 799 females) children, the infectious and parasitic diseases caused 13% of total all cause child deaths.

PAGE 233

215 However for the case of Delhi, India, analyses of 2008 death records indicate the percentage of diarrheal specific mortality relative to total deaths reported by cause for under 5 populations was 0.6% (likely indicating underreporting) and for adult populations as 0.1% (NCT of Delhi Annual Births / Deaths Report, 20 08). Such findings listed above are useful for exploring future mortality scenarios at city scale if the only data available is the < 5 mortality with cause n ot specified (e.g. DHS data in Delhi), yet it is important to note that scaling up from children to adults can be challenging due to not knowing DHS mortality by cause. At the the same scale of child mortality, Delhi DHS under five mortality and whole Delhi under five mortality data can be used to estimate reductions in excess deaths (as shown in result s). If to use same ratios as for children, analyses could assume similar potential reduction for adults, but this approach is not calculated, as such a method would likely be considered questionable. By reviewing and developing a synthesis of available, c ontext appropriate literature field data and models and utilizing health effect estimates health risk mitigation benefits of infrastructure intervention scenarios can be explored. However, challenges still remain. For example, PM10 source apportionment would be needed to explore potential interventions related to transportation, road dust, brick kilns, waste burning, and so on. Furthermore, exposure factors are still missing and multiple infrastructure interventions may contribute in different ways to th ese exposures making health linkages quite challenging. Therefore, exposure data is needed for linking multiple infrastructures and social factors (age, gender) with health risk reduction estimates.

PAGE 234

216 Table VII. 3 Summary of mor e health effect estimates useful for intervention scenarios Intervention Scenarios Health Effect Estimates / Mortality Risk Factors Source PM 10 NAAQS CVD: 0.43% reduction per 10ug/m^3 reduction ; All Cause: 0.07% 0.24%, 0.26%, 0.43%,0.11% 0.20%,and 0.1% 0 .8% reduction per 10ug/m^3 reduction for 0 4 yrs age group, 5 14 yrs, 15 44, 45 65, and >65, respectively. Rajarathnam estimates 0.15% reduction for 5 44 age group; 0.17% (females); 0.09% (males) Cropper, 1997 ; Rajarathnam, 2010 PM 10 NAAQS Respiratory: 0 .31% reduction per 10ug/m^3 PM10 Cropper, 1997 NO2 NAAQS All Cause: 0 .84% reduction per 10ug/m^3 Rajarathnam, 2010 Water Quality Diarrhea: 17% risk reduction ( Cairncross ) for improved water quality; 26% risk reduction (Fewtrell review of 46 studies on po int of use water treatment); Note: diarrhea deaths can also be due to poor food and / or other factors Cairncross, 2010 ; Fewtrell, 2005 Sanitation Diarrhea: 36% risk reduction ; Child Mortality: 8% risk reduction for Low SES Delhi and 19% risk reduction fo r Low SES All India (DHS re analyses) Cairncross, 2010 ; DHS, 2006 Water Supply Diarrhea Mortality : 42 48% risk reduction (by Cairncross); Child Mortality: 14% risk reduction (70 countries by Gunther and Fink); and 5% 23% risk reduction for Low and Mid SES population in Delhi (DHS re analyses by author) Cairncross, 2010 ; Gunther and Fink (2010); DHS, 2006 Solid Fuels CVD: 0.287 for children <5 (see text box below) Smith, 2000 Water & Sanitation Infrastructure 5 20% risk reduction for child mortality (70 c ountries by Gunther and Fink; 84.5% risk reduction for All India child mortality (Sperling DHS analyses, Ch. IV) Gunther & Fink, 2010; DHS, 2006 Solid Fuels ARI: 0.381 % reduction in <5 respiratory deaths Smith, 2000 Solid Fuels TB: 0.366 % reduction in T B deaths for <5 children Smith, 2000 Solid Fuels Lung Cancer: 0.1% reduction in deaths for Females Smith, 2000 Solid Fuels Asthma: 0.315 % of total population asthma deaths are premature mortality ( mid range estimate ) Smith, 2000 Solid Fuels vs. Clean Fuels Child Mortality: 67% risk reduction for Low SES population in Delhi (DHS re analyses by author) DHS, 2006 Rail Mode Shift From Auto: 9.7 deaths / 1% mode shift (transportation fatalities re analyses by author) NCRB, 2008 Rail Mode Shift From 2 Whee elers: 20.6 deaths / 1% shift (transportation fatalities re analyses by author) NCRB, 2008 Ped Safety 5 deaths / 1% reduction in ped fatality risk (transportation fatalities re analyses by author) NCRB, 2008 Physical Activity Diabetes: 25% reduction from walking/cycling to work (not specific to Delhi and so not applicable) De Hartog, 2010 Table 3 summarized key health effect estimates developed through review of literature, models, and analyses of field data that may be useful for intervention scenarios

PAGE 235

217 Where differential risk (e.g. by age, gender, and socioeconomic status) is measured, such results can be used in computations where appropriate. Where possible, the most up to date health effect estimates should be used. Findings from a range of studies can be considered to harmonize health effect estimates and used as a basis for first order computations (i.e. by using the more conservative or full range of appropriate health effect estimates) as shown in Thesis Chapter IV and the Appendix B of this chap ter. With mortality as a health risk reduction indicator, a 2008 baseline developed in Thesis Chapter II ( Sperling & Ramaswami, 201 3) can be used for estimating e xcess mortalit y (and differential risk to subpopulations where possible) for the case of Delhi India. Further data collection and analyses on mortality by age, gender, and other social factors would also be of use. A question that can be explored through a first order set of analyses is : w hat are the health effects of reducing air pollution, water sanitation clean fuels provisiosn, and transportation mode shif ts in Delhi. Preliminary efforts at such a question are explored in the next section. Results of First Order Computation Quantifying Risk Reduction Via First Order Estimation of Health Benefi ts : Equations, assumptions, and initial risk mitigation computations in this section offer three categories of sustainable infrastructure pathways that can be computed related to governance & policy, engineering, and urban design. A summary of preliminary findings as well as challenges for such first order assessment of health risk reduction benefits through three intervention scenarios are described next for the actions of reducing air pollution, under five mortality via basic infrastructure provisions (us ing DHS data analyses and findings from Thesis Chapter IV), and reduced traffic fatalities.

PAGE 236

218 Governance and Poli cy Toward Improved Air Quality: In the boxes below, health effect estimates are applied for excess CVD and Respiratory mortality in overall Delh i population using the Cropper et al. 1997 health effect estimate in first calculation box below (similar to the analys es shown in Thesis Chapter II). Next, computations are shown for all cause mortality for NO2 using HEEs from Rajarathnam et al. 2011. Usi ng this same study, first order computations for PM10 as modified by age and gender are also explored The limitation to this PM10 analysis by age and gender is that we currently assume same exposure for these different populations, and same exposure reduc tions for these sub populations, which is unlikely to be the case. Figure VII. 1 Example Analysis using Cropper et al. HEE from Thesis Chapter II As shown above and in Thesis Chapter II, reduction in premature mortality as re lated to infrastructure related environmental pollution can be computed using data on pollution change (i.e. annual change in PM10 in micrograms per meter cubed ( u g/m3 )),

PAGE 237

219 health effect estimates for specific population groups (i.e. the % change in mortalit y from a certain change in level of ug/m3 of PM10), mortality incidence (# of deaths per study population), and total people in population/s of interest.

PAGE 238

220 Figure VI. 2 Rajarathnam et al. HEE for NO2 and PM10, with PM10 by A ge and Gender To more fully understand the variation in estimates for reduced mortality in terms of various subpopulations, exposure data and data on changes in exposure would be needed. The sample calculation in the box demonstrated the variation in redu cing mortality for various subpopulations through the action of meeting PM10 pollution standards, assuming exposures are all the same. The analyses were conducted using the Rajarathnam et al. (2010) study on premature mortality and PM10 in Delhi, India tha t reports a 0.07 0.24% increase in all cause mortality risk for 0 4 yrs age group, 0.15% for 5 44 age group, 0.11 0.20% for 45 65 age group, and 0.1 0.8% for >65 age group; 0.17% for females and 0.09% for males per 10ug/m3 increase above prescribed standar ds. Note that where a range is provided, middle of range was used to compute mortality reduction: e.g. for the <5 age group where 0.07 0.24% is reported, we use .15%; t he same for the 5 44 age group. An important no te on the limitation of these initial a nalyses is that doing such fine grained analyses by social factors requires detailed analyses by exposures for these populations with such data not so readily available. So to summarize, we know HEE and have some idea of populations in subgroups (women, o ld, young, etc), but do not know differences of exposure. Engineering: Toward Clean Energy Water Sanitation for All : The below computation utilizes a modest health effect estimate from DH S re analyses conducted in Thesis Chapter 4, under five mortality dat a from the DHS and Delhi 2008 baseline mortality data, then considers homes without multiple improved conditions: water and sanitation off premises and using solid fuels using DHS survey results. In order words, DHS is used to estimate the homes of this s pecific condition and what improvements to

PAGE 239

221 under five mortality might be achieved if you improve these homes currently without provision of water supply / toilet on premises, and clean fuels for cooking. As mentioned in the literature review section, scali ng up from children to adults can be challenging due to not knowing DHS mortality by cause. Here analyses use child mortality from Delhi DHS under five mortality and whole Delhi under five mortality data to compute potential reductions in excess deaths. If to use same ratios as for children, analyses could assume similar potential reduction for adults, but this approach is not calculated, as such a method would likely be considered questionable. To summarize, what is not known is adult mortality and how to scale up to adults if having DHS data and Delhi whole population mortality data as summarized in this chapter's literature review. In addition, further work is needed on how to scale to other cities for <5 mortality and for adults with another limitation being appropriate city mortality data.

PAGE 240

222 Figure VI. 3 Thesis Ch. IV HEE Initial Application: Multiple Basic Provisions to Low SES Delhi Households Lacking Basic Provisions (Using DHS Survey and Delhi Data) Urban Design: Active Walkable & Transit Oriented Communities: The preliminary calculations below use Delhi mode share data and national fatality data by mode to make a first order estimate of reductions in traffic fatalities for a shifts in total trips to rail from from cur rent auto / two wheeler modes, respectively. While assumptions are needed for this computation, the major limitation remains exposure data and changes in individual exposure. For example, such a mode shift would likely result in larger pedestrian populatio ns heading to and from rail stations, so exposure data for such pedestrians would be needed to more fully account for reductions in traffic fatality when shifting a % of trips to rail from these other modes of travel. To summarize, current risk is calculat able (with modal shift assumptions needing improvement), future risk is not. Supplementary All India and Delhi modal split data, fatality by mode data, and preliminary HEEs useful for computations in Figure 4 are provided in the Appendix.

PAGE 241

223 Figure VI. 4 First Order Computation of Reduced Traffic Fatalities Discussion : Synergies in Sustainable Infrastructure Development and Urban Health Rapid urban population growth has significant local impacts, as seen in Delhi, and can place high stress on infrastructures and services, often leading to inadequate infrastructure and services (e.g., lack of clean drinking water, affordable housing, hazardous waste management, health services, etc.), which can lead to hazards of poor sanitation, disease, social and psychological stress, and increased demand for scarce resources (UN Cities in Africa Report, 2007 ). The possibility of employment based on some economic growth in urban areas is one of the major causes of rural to urban migration fo r Delhi and other cities. Meanwhile, employment and economic growth usually can't keep up with the population growth, leading to increased unemployment, poverty, and stresses on the local

PAGE 242

224 infrastructure and environment. Such drivers increase the importance of strategic decision making on urban infrastructure investments that can equitably improve health (e.g. reducing unnecessary mortality), environment (e.g. reduce greenhouse gas and black carbon emissions), economy (e.g. c reate jobs and income sources), a nd access to services. While economic benefits infrastructure finance, and policy actors have yet to be considered for detailed analyses in this study, they are identified as import ant drivers for decision making. R egarding equity, analyse s using health e ffect estimates presenting differential risk by age, gender, socioeconomic status, etc can also prove useful to decision making and determining whether various subpopulations may receive equal benefit from the initial high level interventions identified in this chapter (Appendix A) Conclusion Building on this synthesis of literature, models, and data, a first order assessment of health risk mitigation is shown that also demonstrates what is known and what is needed to develop scenarios and compute first o rder health benefits for the case of Delhi, India. Initial analyses and synthesis of health effect estimates are applied to estimate the potential benefits for three hypothetical infrastructure intervention scenarios including reducing air pollution, impro ving basic infrastructure provisions to low socioeconomic status households, and reducing traffic fatalities all of which could be aimed at mitigating ~4% of total 2008 mortality (up to 4365 excess deaths). These first order analyses remain preliminary a nd additional epidemiological study and analyses w ould be useful However, the data, literature and analyse s shown serve as a useful starting point for an analytical scenario tool that can help quantify infrastructure intervention benefits for public healt h

PAGE 243

225 Further exploration into the line of research of coupling health risk reduction and GHG mitigation may also be important for understanding relevant questions, such as: what future health and GHG outcomes can motivate development of infrastructures for l ow carbon and healthy cities in Asia today; when does infrastructure related health strategies lead to large GHG reduction benefits and when do they not; and when can GHG mitigation strategies lead to large health risk reduction benefits. Future analyses a ddressing such questions would also need to consider the spatial allocation of pollutant concentration reductions and changes in individual exposure due to GHG strategies, which would be a quite complex task. Further adding to this complexity, but relevant to moving forward is exploring equity in more depth regarding potential distribution of benefits to subpopulations. Based on this first order assessment, existing literature and community surveys, we conclude that for rapidly growing Asian cities where la ck of adequate infrastructures and high levels of environmental pollution are everyday experiences for most urban inhabitants, any primarily health based strategies for infrastructure development are likely to be more successful than GHG mitigation based s trategies for infrastructure development due to the fact that water supply, air pollution, and other local public health concerns will often outweigh global climate concerns. Further data and analyses on pollution exposure related health effects for waste burning, for overcrowded sleeping conditions, and on access to healthcare may also prove useful to explore additional scenarios and potential high priority interventions relevant to different subpopulations within Indian cities.

PAGE 244

226 Appendix A. First Order Com putations Supplementary Data: Mode Split & % of Mode Specific to Total Transport Fatalities

PAGE 245

227 CHAPTER VIII. VIII. SUMMARY & CONCLUSIONS While infrastructure and infrastructure related environmental factors shape health outcomes in various ways, assessing these r isks together can help to improve infrastructure decision making in ways that may reduce health risks (e.g. mortality). In this dissertation, we have examined the potential infrastructure and environment related impacts on health in Asian cities, includin g impacts from air pollution as well as access to water supply, sanitation, energy, and healthcare. The main outputs include a) both quantitative and qualitative studies exploring the extent to which socioeconomic conditions, multiple civil infrastructures (e.g. water, sanitation, energy infrastructures) and infrastructure related environmental factors (e.g. air quality) can shape health outcomes and b) addressing potential confounding factors for infrastructure environment health interactions in Indian cit ies including wealth, literacy, and access to healthcare. The community based participatory (CBPR) research methods in this thesis were also instrumental for improving the research, relevance and understanding achieved. Such community studies allowed for understanding local sustainability priorities, complexities of human settlement, and the potential impacts of infrastructure and environment on local health outcomes in areas with different infrastructure and socioeconomic conditions. In this thesis, fiv e phases of work are developed. In the first phase, Chapter Two, we explore the extent to which civil infrastructure (i.e., water, sanitation, energy, transport and building infrastructures) and environmental factors (e.g. air and water quality) associated with these infrastructures shape current urban health outcomes. We find that up to 19% of all recorded deaths in Delhi, India may be infrastructure related.

PAGE 246

228 While preliminary, the findings suggest health outcomes may be a large factor in motivating low ca rbon development in Asian cities. In Phase two, we develop a first comparative assessment of multiple civil infrastructure conditions associated with health, specifically the policy relevant outcome of reducing under age five mortality. We find that when controlling for wealth and literacy, significant reductions in mortality rates are achievable in Urban India by improving multiple infrastructures together including water, toilets, and cooking fuels. In Phase 3, knowledge is acquired through linking en gineering and epidemiological literature and models in order to more effectively quantify current health outcomes associated with infrastructure and environment related factors. Such literature can contribute to new measurement tools and evidence for local and inclusive decision making on infrastructure interventions that can have significant impacts on improved health and equitable access to adequate basic infrastructures. Phase 4 demonstrate bottom up methods for characterizing infrastructure environment extreme weather conditions and experiences, local priorities in different neighborhoods, in addition to access to healthcare and socioeconomic conditions. Finally, Phase 5 presents what is known and what is needed for computing health benefits by develop ing and utilizing health effect estimates and integrating past research to compute first order health risk mitigation potential scenarios. Such efforts in addition to bottom up studies of residents experiences with infrastructure deficiences and high lev els of environmental pollution on a daily basis suggest that GHG mitigation strategies serving primarily as health improvement strategies can likely be more

PAGE 247

229 successful than climate action strategies alone as water supply, air pollution, and addressing ot her local health concerns can often outweigh global climate concerns. A Preliminary Conceptual Diagram for Future Related Studies: To conclude, we return to the preliminary conceptual diagram to build on for future study related to urban infrastructures, environment, and health. As shown, the preliminary elements of this conceptual diagram (that has still yet to be applied fully ) includes biophysical system characteristics (e.g. infrastructure and environmental conditions) and human social system character istics (e.g. socio economic and biological). Future use could involve assessment of risk, differential vulnerability, and risk mitigation in cities. Figure VIII. 1 Schematic Representation of a Preliminary Analytical Framewor k

PAGE 248

230 Exposures, health effect estimates, and health outcomes must also be considered in order to improve decision making on both infrastructure and public health interventions in cities. The assessment of evidence on health effect estimates, preliminary top d own and bottom up methods for exploring infrastructure environment health conditions by SES, as well as understanding of local action priorities and first order benefits computations can all have implications for infrastructure in Delhi, India as a represe ntative rapidly growing Asian city. The preliminary conceptual diagram shown helps tie some of the key elements of this thesis together and can remain flexible to future refinement based on changing needs and semantics that may come about through continue d integrative interdisciplinary discourse. For example, health infrastructure' which the US Department of Health and Human Services currently defines as including three key components: (1) a capable and qualified workforce; (2) up to date data and infor mation systems; and (3) public health agencies capable of assessing and re sponding to public health needs may be a larger focus for using this diagram and the WHO framework in the future. To date, this thesis has only addressed the multiple civil infrast ructures, infrastructure related environmental factors, and to some extent the socioeconomic conditions (e.g. income, education, access to healthcare) that can shape public health. Further research is still needed on access to healthcare as it shapes healt h impacts and on differential risk to diverse subpopulations due to exposures. For such purposes, alternative frameworks should also be considered with perhaps integration of infrastructure services where not explicitly defined, such as highlighted in red on the WHO 2010 framework shown below.

PAGE 249

231 To conclude, the work conducted in this thesis can help open new line s of inquiry for improving health in cities via multiple infrastructure improvements, and can perhaps also lead to more scalable and replicable t ools, policies and practical solutions for application in real world c ities, with relevance to Delhi and other cities in India and Asia.

PAGE 250

232 REFERENCES Chapter 1 Amekudzi A, Ramaswami, A., Chan, E., Lam, K., and W. Meng. Contextualizing carbon reduction ini tatives: how should carbon mitigation be addressed by various cities worldwide? Carbon Manag. 2(4), 363 365 (2011). Amekudzi A. Placing carbon reduction in the context of sustainable development priorities: a global perspective. Carbon Manag. 2(4), 413 42 3 (2011). Aggarwal, R. 2013. Strategic Bundling of Development Policies with Adaptation: An Examination of Delhi's Climate Change Action Plan. International Journal of Urban and Regional Research. Vol. 37; September 2013; 1902 15. Campbell Lendrum, D. an d C. Corvalan. 2007. Climate Change and Developing Country Cities: Implications for Environmental Health and Equity. Journal of Urban Health: Bulletin of the New York Academy of Medicine, Vol. 84, No. 1. Chavez, A. and A. Ramaswami, 2013. "Articulating An Infrastructure Supply Chain Greenhouse Gas (GHG) Emissions Footprint for Cities: Mathematical Relationships and Policy Relevance", Energy Policy. Vol. 54: 376 384. March 2013. http://dx.doi.org/1 0.1016/j.enpol.2012.10.037 De, U.S., Singh, G.P, and D.M. Rase. 2013. Urban flooding in recent decades in four mega cities of India. J. Ind. Geophyc. Union. Vol. 17, No. 2, pp. 154 165. Accessed July 2013 at http://www.igu.in/17 2/4usde.pdf Murray, C., a nd A. Lopez. 1997. Mortality by cause for eight regions of the world: Global Burden of Disease Study. Lancet. Accessed on June 1, 2013: www.ncbi.nlm.nih.gov/pubmed/9142060 Pruss Ustun, A. and C. C orvalan. 2007. How Much Disease Burden can be Prevented by Environmental Interventions. Epidemiology and Society: A Forum on Epidemiology and Global Health. Ramaswami, Anu. 2013. "Understanding Urban Infrastructure Related Greenhouse Gas Emissions and Ke y Mitigation Strategies". Infrastructure and Land Policies, eds. Ingram, Gregory K. and Karin L. Brandt. Cambridge, MA: Lincoln Institute of Land Policy Pruss Ustun, A., Bonjour, S., and C. Corvalan. 2008 The impact of the environment on health by country : a meta synthesis. Environmental Health; 2008 (7:7).

PAGE 251

233 World Health Organization (WHO), 2010. Environment and health risks: a review of the influence and effects of social inequalities. WHO Europe. Chapter 2 Amekudzi A, Ramaswami, A., Chan, E., Lam, K. and W. Meng. Contextualizing carbon reduction initatives: how should carbon mitigation be addressed by various cities worldwide? Carbon Manag. 2(4), 363 365 (2011). Amekudzi A. Placing carbon reduction in the context of sustainable development prioritie s: a global perspective. Carbon Manag. 2(4), 413 423 (2011). Babinard, J. and P. Roberts. 2006. Maternal and child mortality development goals: what can the transport sector do? The World Bank. http://www.eldis.org/index.cfm?objectId=12FF2014 F8B6 EF21 2F C73C39FBAA7E93. Accessed May 2011. Boschi Pinto, C., Velebit, L. and K. Shibuya. 2008. Estimating child mortality due to diarrhoea in developing countries. Bulletin of the World Health Organization | September 2008, 86 (9) Butala, N., VanRooyen, M. and R Patal. 2010. Improved health outcomes in urban slums through infrastructure upgrading. Social Science & Medicine 71 (2010) 935e940 C. Venkataraman, A.D. Sagar, G. Habib, N. Lam, and K.R. Smith. 2010. The Indian National Initiative for Advanced Biomass C ookstoves: The Benefits of Clean Combustion. Energy for Sustainable Development 14 (2010) 63 72 CDC, 2011. "Chronic Disease Obesity At A Glance." Centers for Disease Control and Prevention. Accessed September, 2011: www.cdc.gov/chronicdisease/resources/publications/aag/obesity.htm Chavez, A., Ramaswami, A., Dwarakanath, N., Ranjan, R., & Kumar, E. (2011). Implementing Expanded Geographic Based Greenhouse Gas Accoun ting for Delhi, India. Data Availability and Methods. Journal of Industrial Ecology. In Press. Diabetes Foundation India DFI. Accessed August, 2011. http://www.diabetesfoundationindia.org/abo ut.htm Eisenburg, J. N. S.; Bartram, J.; Hunter, P. R. A Public Health Perspective for Establishing Water Related Guidelines and Standards. In Water Quality Guidelines, Standards, and Health: Assessment of Risk and Risk Management for Water Related Infe ctious Disease; WHO: Geneva, 2001; pp 229 256.

PAGE 252

234 U.S. Environmental Protection Agency (EPA). 2006. "BenMAPP: Environmental Benefits Mapping and Analysis Program | Case Study: Mumbai." Prepared by ABT Associates. Accessed September 2011: http://www.epa.gov/air/benmap/ Ezzati, M., Vander Hoorn, S., Lopez, A.D., Danaei, G., Rodgers, A., Mathers, C.D., & CJL, M. (2006). Global Burden of Disease and Risk Factors, Annex 4a. Oxford University Press: New York. Fang L Q de Vlas SJ, Liang S, Looman CWN, Gong P, et al. 2008 Environmental Factors Contributing to the Spread of H5N1 Avian Influenza in Mainland China. PLoS ONE 3(5): e2268. doi:10.1371/journal.pone.0002268 Friel, S. et al. 2009. Public health benefits of str ategies to reduce greenhouse gas emissions: food and agriculture. The Lancet. Published Online November 25, 2009 DOI:10.1016/S0140 6736(09)61713 X. www.thelancet.com Gadgil A.J. and Derby E.A., 2003. Providing Safe Drinking Water to 1.2 Billion Unserved People. Gleick, P. H. 2000. Dirty Water: Estimated Deaths from Water Related Diseases 2000 2020; Pacific Institute: Oakland, CA. Gordon Larsen, P., Nelson, M. C., Page, P., Popkin, B. M., 2006. Inequality in the Bu ilt Environment Underlies Key Health Disparities in Physical Activity and Obesity. Pediatrics 117, 417 424. Government of India Central Bureau of Health Intelligence. 2010. National Health Profile. Accessed July, 2011: www.cbhidghs.nic.in Government of National Capital Territory (NCT) of Delhi. 2008a. Annual Report on Registration of Births and Deaths in Delhi. Government of National Capital Territory (NCT) of Delhi. 2008b. Report Medical Certification of Causes of Deaths in Delhi. Griffith J, Duncan RC, Riggan WB, Pellom AC. 1989. Cancer mortality in U.S. counties with hazardous waste sites and ground water pollution. Arch Environ Health 44: 69 74 Guerriero, C. and J. Cairns. 2009. The potential monetary ben efits of reclaiming hazardous waste sites in the Campania region: an economic evaluation. Environmental Health 2009, 8:28 doi:10.1186/1476 069X 8 28. http://www.ehjournal.net/content/8/1/28 Gunther, I. and G. Fink. 2011. Water and Sanitation to Reduce Child Mortality: the Impact and Cost of Water and Sanitation Infrastructure. Research Working Paper 5618. The World Bank: Development Economics Prospects Group.

PAGE 253

235 Gupta, N., Jain, S., Chawla, U., Hossain, S., and S. Venkatesh. 2007. Evaluation of Diarrheal Diseases and Acute Respiratroy Infections Control Programmes in a Delhi Slum. National Institute of Communicable Diseases and Division of Reproductive Health and Nutrition, Indian Council of Medical Resea rch, New Delhi, India Hajat, S., Armstrong, B.G., Gouveia, N., & Wilkinson, P. (2005). Mortality displacement of heat related deaths: a comparison of Delhi, Sao Paolo, and London. Epidemiology, 16(5), 613 620. Hess JJ, Bednarz D, Bae, JY, Pierce J. Petro leum and Health Care: Evaluating and Managing Health Care's Vulnerability to Petroleum Supply Shifts. American Journal of Public Health. 2011. doi: 10.2105/AJPH.2011.300233: 2011. Hillman, T., & Ramaswami, A. 2010. Greenhouse Gas Emission Footprints and E nergy Use Benchmarks for Eight U.S. Cities. Environ. Sci. Technol., 44(6), 1902 1910. ICLEI SA. (2009). Energy and Carbon Emissions Profiles of 54 South Asian Cities. ICLEI South Asia. International Energy Agency (IEA). 2010/2011. CO2 Emissions From Fuel Combustion. IEA. 2011. "Implication of Energy Poverty on Health & Environment." Accessed May 2011: . Jacobson, M. 2007. "On the Causal Link Between Carbon dioxide and Air Pollution Mortality". Geophysical Research Lette rs, Vol. 35, doi:10.1029/2007GL031101. Jemal, A., Bray, F., Center, M. M., Ferlay, J., Ward, E. and Forman, D. (2011), Global cancer statistics. CA: A Cancer Journal for Clinicians, 61: 69 90. doi: 10.3322/caac.20107 Kammen, D et al. 2011. Household Cook stoves, Environment, Health, and Climate Change: A New Look at an Old Problem. The Environment Department (Climate Change), The World Bank. Killoran, A., N. Doyle, S. Waller, C. Wohlgemuth, and H. Crombie. "Transport Interventions Promoting Safe Cycling a nd Walking." National Institute for Health and Clinical Excellence (2006). www.nice.org.uk Accessed March, 2011. Listorti, J. and F. Doumani. 2004. Energy and Environmental Health: A Literature Review and Recommendations. Joint UNDP/World Bank Energy Sect or Management Assistance Programme (ESMAP) Lo, A. (2010). China's Response to Climate Change. Environ. Sci. Technol., 44(15), 5689 5690.

PAGE 254

236 Markandya, A. et al. 2009. Public health benefits of strategies to reduce greenhouse gas emissions: low carbon electr icity generation. The Lancet. Published Online November 25, 2009 DOI: 10.1016/S0140 6736(09)61715 3. www.thelancet.com Matin, N., Mukib, H.B. & Khanam, D. 2002. Women's Empowerment and Physical Mobility. In: Fernand o, P. & Porter, G.(eds). Balancing the Load: Women, Gender and Transport. London: IFRTD/Zed Books. McMichael, A, Woodward, A. and R. Leeuwen. 1994. The Impact of Energy Use in Industrialised Countries upon Global Population Health. Medicine & Global Surv ival 1994; Vol. 1, No. 1 Miller, L., Ramaswami, A., and P. Kumar. 2011. Life Cycle Assessment of Energy Use and Greenhouse Gas Emissions for a Wastewater Treatment Plant: A Case Study in Hyderabad, India. (In Review: Water Environment Research Journal). Ministry of Urban Development, Government of India. 2011. Improving urban services through Service Level Benchmarking. Accessed Oct, 2011: http://urbanindia.nic.in/programme/uwss/slb/slb.htm TERI, 2010. "State of Environment Report for Delhi 2010". Access ed September 2010: www.environment.delhigovt.nic.in Mobley, L., Root, E., Finkelstein, E., Khavjou, O., Farris, R., and J. Will. Environment, Obesity, and Cardiovascular Disease Risk in Low Income Women. American Journal of Preventive Medicine, Volume 30, Issue 4, April 2006, Pages 327 332.e1, ISSN 0749 3797, 10.1016/j.amepre.2005.12.001. ( http://www.sciencedirect.com/science/article/pii/S0749379705005040 ) Mohan, D. 2004."Th e Road Ahead: Traffic Injuries and Fatalities in India", New Delhi, India. http://tripp.iitd.ernet.in/publications/paper/safety/road_ahead.pdf Accessed May 2011. Molesworth, K. 2006. Mobility and Health: Impact of Transport Provision on Direct and Indirec t Determinants of Access to Health Services. Montgomery, M. and M. Elimelech. 2007. Water And Sanitation in Developing Countries: Including Health in the Equation Environmental Science & Technology 2007 41 (1), 17 24 Murray, C.J.L. 1994. Quantifying th e burden of disease the technical basis for disability adjusted life years. Bulletin of the World Health Organization 72(3): 429 445. Nanda, G., Switlick, K. and E. Lule. 2005. Accelerating Progress towards Achieving the MDG to Improve Maternal Health: A Collection of Promising Approaches. Health, Nutrition and Population Discussion Paper. The World Bank Group.

PAGE 255

237 National Research Council (NRC). 2008 Committee on Estimating Mortality Risk Reduction Benefits from Decreasing Tropospheric Ozone Exposure. Natio nal Academies. Accessed September 2011: http://www.nap.edu/openbook.php?record_id=12198&page=R1 Nawrot, T., Perez, L., KŸnzli, N. Munters, E., and B. Nemery. 2011. Public health impor tance of triggers of myocardial infarction: a comparative risk assessment. The Lancet Vol. 377, Issue 9767, Pages 732 740 DOI: 10.1016/S0140 6736(10)62296 9 Pasala, S, Rao, A. and G. Sridhar. 2010. Built Environment and Diabetes. Int J Diabetes Dev Ctrie s Apr: 30(2): 63 8 Pasternak, A. 2000. Global Energy Futures and Human Development: A Framework for Analysis. US Department of Energy: Lawrence Livermore National Laboratory. Patz JA, Campbell Lendrum D, Holloway T, Foley JA. Impact of regional climate c hange on human health. Nature 2005; 438: 285. PRC. Public's Priorities for 2010: Economy, Jobs, Terrorism. The Pew Research Center. Washington, D.C. January 25, 2010 Ramaswami, A. and S. Dhakal. 2011. Low carbon policies in the USA and China: why cities p lay a critical role. Carbon Management (2011) 2(4), 359 362 Rushton et al. Occupation and cancer in Britain. British Journal of Cancer (2010) vol. 102 (9) pp. 1428 1437. Rushton L. 2003. Health hazards and waste management. British Medical Bulletin Vol ume 68 Issue 1: 183 197. doi: 10.1093/bmb/ldg034 Russell, A. (2008) EPA Supersites Program Related Emissions Based Particulate Matter Modeling: Initial Applications and Advances. Journal of the Air & Waste Management Association 58: 289 302. Rutherford et al., 2009. Access to health care and mortality of children under 5 years of age in the Gambia: a case control study. Bulletin of the World Health Organization 87:216 224. doi: 10.2471/BLT.08.052175 Saatkamp, B., Masera, O., and D. Kammen. 2000. Energy and health transitions in development: fuel use, stove technology, and morbidity in Jar‡cuaro, MŽxico. Energy for Sustainable Development, Volume IV No. 2, August 2000. Saelens, Brian E., Sallis, James F., Black, Jennifer B., Chen, Diana. 2003. Neighborho od Based Differences in Physical Activity: An Environment Scale Evaluation. Am J Public Health. 93: 1552 1558

PAGE 256

238 Samet JM, Zeger SL, Dominici F, et al. The National Morbidity, Mortality, and Air Pollution Study. Part II: Morbidity and mortality from air poll ution in the United States. Res Rep Health Eff Inst 2000;94:5 70 Schwartz B, Parker C, Hess JJ, Frumkin H. 2011. Public Health and Medicine in an Age of Energy Scarcity: The Case of Petroleum. American Journal of Public Health. doi 10.2105/AJPH.2010.20518 7. Sen, A., 1998. "Mortality as an Indicator of Economic Success and Failure," Economic Journal, Royal Economic Society, vol. 108(446), pages 1 25. Shaw M. Housing and Public Health. Annual Review of Public Health 2004;25:397 418. Smith, A., Lingas, E., and M. Rahman. 2000. Contamination of drinking water by arsenic in Bangladesh: a public health emergency. Bulletin of the World Health Organization, 2000, 78 (9) Smith, K. et al. 2009. Public health benefits of strategies to reduce greenhouse gas emissio ns: health implications of short lived greenhouse pollutants. The Lancet. Published Online November 25, 2009 DOI: 10.1016/S0140 6736(09)61716 5. Steinberger, J. and J. Roberts. 2009. Across a Moving Threshold: energy, carbon and the efficiency of meetin g global human development needs. Social Ecology Working Paper 114, Vienna. UN Water Decade Programme on Advocacy and Communication (UNW DPAC). Water and Cities: Facts and Figures. United Nations Framework Convention on Climate Change (UNFCCC). 2010. Gr eenhouse Gas Inventory Data. http://unfccc.int/ghg_data/items/3800.php. Accessed August, 2011. US Mayors (2011). U.S. Mayors Climate Protection Agreement. Retrieved May 20, 2011, from http:/ /www.usmayors.org/climateprotection/list.asp Water Quality and Health Council, 1995. "Water Wastewater Chlorination: An Enduring Public Health Practice." Water Quality and Health Council. Accessed August, 2011. . WHO. 2004. "Global Burden of Disease for the Year 2001, for Use in Disease Control Priorities in Developing Countries." 2nd edition. Accessed August, 2011: www.fic.nih.gov/ dcpp/gbd.html and http://7pointplan.org/global burden childhood diarrhoea.html

PAGE 257

239 WHO, 2004. World report on road traffic injury prevention. Accessed March 2011: www.who.int/violence_injury_prevention/publications/road_traffic/world_report/en/inde x.html World Health Organization. 2008. Safer Water, Better Health: Costs, benefits, and sustainability of interventions to protect and pro mote health. Accessed July, 2011: http://www.who.int/quantifying_ehimpacts/publications/saferwater/en/index.html WHO. 2011. "Cancer | Diabetes | Hepatitis A." Accessed August, 2011. < www.who.int/cancer/en/ > & < www.who.int/diabetes/en/>. Aguilera, N, and G. Marrufo. Can better infrastructure and quality reduce hospital infant mortality rates in Mexico? Health Policy 80 (2007) 239 252. Wilkinson, P. et al. 2009. Public health benefits of strategies to reduce greenhouse gas emissions: household energy. The Lancet. Wilkinson, P., Smith, K., Joffe, M., and A. Haines. 2007. A Global Perspective on Energy: Health Effects and Injustices. The Lancet. Woodcock, J. et al. 2009. Public health benefits of strategies to reduce greenhouse gas emissions: urban land transport. The Lancet. Ch apter 3 Agarwal, S. and A. Srivastava. 2009. Social Determinants of Children's Health in Urban Areas. Journal of Health Care for the Poor and Underserved. http://www.measuredhs.com/publications/journal details.cfm?article_id=1168&C_id=26&T_ID=0&P_ID=0 American Diabetes Association (ADA), 2010. Diabetes Statistics. http://www.diabetes.org/diabetes basics/diabetes statistics/ Antunes and Waldman, 1997 Tuberculosis in the twentieth century: time series mortality in S‹o Paulo, Brazil, 1900 97. Accesse d Jan, 2012: http://www.scielosp.org/scielo.php?pid=S0102 311X1999000300003&script=sci_arttext Antunes and Waldman, 2001. The impact of AIDS, immigration and housing overcrowding on tuberculosis deaths in S‹o Paulo, Brazil, 1994 1998. Soc Sci Med. 2001 Apr;52(7):1071 80. Accessed November, 2011: http://www.ncbi.nlm.nih.gov/pubmed/11266050 Arnett, J. 2002. Development sources of crash risk in young drivers. Inj Prev 2002; 8 (Suppl II): ii17 ii23.

PAGE 258

240 Baker, M., Das, D., Venugopal, K., and P. Howden Chapman. 2008. Tuberculosis associated with household crowding in a developed country. J Epidemiol Community Health 2008;62:715 721 doi:10.1136/jech.2007.063610. http://jech.bmj.com/content/62/8/715.full Baumgartner, J., Zhang, Y., Ezzati, M., Bautista, L., and J. Scha uer. 2011. Black carbon is more strongly associated with cardiovascular risk than PM2.5 in women cooking with biomass fuels. Bell, M., Zaobetti, A., and F. Dominici. 2013. Evidenec of Vulnerability and Susceptibility to Health Risks Associated with Short Term Exposure to Particulate Matter: A Systematic Review and Meta Analysis. American Journal of Epidemiology. Berrington, A., Sweetland, S. and E. Spencer. A meta analysis of obesity and the risk of pancreatic cancer. Br J Cancer 89(3):519 23. http://www.ncbi.nlm.nih.gov/pubmed/12888824 Bond, T., Sun, H. (2005), Can reducing black carbon emissions counteract global warming? Env. Sci. & Tech ., 39, 16. Bramley, S., and E. Breslin, 2011. Sanitatio n as a Business : A new spin on the challenge of sanitation Operation and Maintenance http://support.waterforpeople.org/site/DocServer/S SP article bramley and breslin.pdf/864370681?docID=1661 Bell, M., McDermott, A., Zeger, S., Samet, J., Dominici, F. 2004. Ozone and Short Term Mortality in 95 US Urban Communiteis, 1987 2000. JAMA 292 (19):2372 2378 Boschi Pinto, C., Velebit, L. and K. Shibuya. 2008. Estimating child mortality due to diarrhea in developing countries. Bulletin of the World Health Organization 2008; 86:710 717. Brunekreef, B and S. Holgate. 2002. Air Pol lution and Health. THE LANCET, Vol 360, October 19, 2002, Accessed at : www.thelancet.com Cairncross, S., Hunt, C., Boisson, S., Bostoen, K., Curtis, V., Fung, I., and W. Schmidt. 2010. Water, sanitation and hygiene for the prevention of diarrhea. International Journal of Epidemiology; 39: i93 i205. Oxford University Press. DOI: 10.1093/ije/dyq035 Cakmak, S., Martin, S., Vidal, C, Gultekin, T., Brion, V., and M. Rubio. 2010. Do socio demographic characteristics modify the association between air pollution and mortality & morbidity? Ad vanced Topics in Environmental Health and Air Pollution Case Studies Cancer Research UK. 2012. Body weight and cancer: the evidence, 2012. Accessed Feb, 2012: http:/ /info.cancerresearchuk.org/healthyliving/obesityandweight/howdoweknow/

PAGE 259

241 Carter, P., Gray, L., Troughton, J., Khunti, K., and M. Davies. 2010. Fruit and vegetable intake and incidence of type 2 diabetes mellitus: systematic review and meta analysis. BMJ 2 010; 341:c4229. http://www.bmj.com/content/341/bmj.c4229.full Central Pollution Control Board (CPCB). 2008. Study on Ambient Air Quality, Respiratory Symptoms & Lung Function of Children in Delh i. Ministry of Environmnet and Forests. Environmental Health Series: EHS/2/2008. www.cpcb.nic.in Chavez, A., Ramaswami, A., Dwarakanath, N., Ranjan, R., & Kumar, E. (2011). Implementing Expanded Geographic Based Green house Gas Accounting for Delhi, India. Data Availability and Methods. Journal of Industrial Ecology. In Press. Cohen AJ, Anderson HR, Ostro B, Pandey KD, Krzyzanowski M, KŸnzli N, et al 2004. Mortality impacts of urban air pollution. In: Comparative Quant ification of Health Risks: Global and Regional Burden of Disease Due to Selected Major Risk Factors, Vol 2 (Ezzati M, Lopez AD, Rodgers A, Murray CJL, eds). Geneva:World Health Organization. Clark M, Riben P, Nowgesic E. 2002. The association of housing d ensity, isolation and tuberculosis in Canadian First Nations Communities. International Journal of Epidemiology 31(5):940 945. Cropper, M. L., Simon, N. B., Alberini, A., & Sharma, P. K. (1997). The health effects of air pollution in Delhi, India. PRD wo rking paper 1860 (unpublished). New Ideas in Pollution Regulation. Washington, D.C.: World Bank. Accessed November, 2011. www.econ.umd.edu/wcropper/publications/jc36.pdf Dai, Z, Xu, YC, and L. Niu. 2007. Obesity and colorectal cancer risk: a meta analysis of cohort studies. World J Gastroenterol. 2007 Aug 21; 13(31): 4199 206. http://www.ncbi.nlm.nih.gov/pubmed/17696248 De Harto g, J., Boogard, H., Nijland, H., and G. Hoek. 2010. Do the health benefits of cycling outweigh the risks? Environmental Health Perspectives. Volume 188 / Number 8 / August 2010. Delhi Ministry of Environment, Climate Action Plan 2009 2012. http://delhi.gov.in/wps/wcm/connect/environment/Environment/Home/Climate+Change/ Dockery, D., Pope, C., Xu, X, Spengler, J., Ware, J., Fay, M., Ferris, B. and F. Speizer. 199 3. Association between Air Pollution and Mortality in Six U.S. Cities. http://www.nejm.org/doi/full/10.1056/NEJM199312093292401 Elvik R. A meta analysis of studies concerning the safe ty effects of daytime running lights on cars. Accident Analysis and Prevention 1996; 28(6):685 694.

PAGE 260

242 Englert, N. 2004. Fine particles and human health a review of epidemiological studies. Toxicology Letters 149: 235 242. Esrey, S and J. Habicht. 1985. T he impact of improved water supplies and excreta disposal facilities on diarrheal morbidity, growth and mortality among children. Cornell International Nutrition Monograph Series No. 15. Ithaca NY: Cornell University Esrey, S. and J. Habicht. 1986. Epidem iologic evidence for health benefits from improved water and sanitation in developing countries. Epidemiologic Reviews 8:117 127 Esrey, S.A., Potash, J.B., Roberts, L., and C. Shiff. 1991. Effects of improved water supply and sanitation on ascariasis, dia rrhea, dracunculiasis, hookworm infection, schistosomiasis, and trachoma. Bulletin of the World Health Organization, 69 (5): 609 621 Esrey SA (1996). Water, waste, and well being: a multicountry study. American Journal of Epidemiology, 143(6):608 623. Fe wtrell, L., & Kaufmann, R. (2005). Water Sanitation and Hygiene Intervetnions to Reduce Diarrhoea in Developed Countries: A Systematic Review and Meta Analysis. Lancet Infectious Diseases 5. Fink, G., Gunther, I., and K. Hill. 2011. The effect of water and sanitation on child health: evidence from the demographic and health surveys 1986 2007. Int. J. Epidemiol. 2011;1 9. Oxford University Press. Fischbeck, P., Gerard, D., Quesenberry, J. and B. Gengler. 2012. Death Risk Rankings. Carnegie Mellon Univers ity Center for the Study and Improvement of Regulation. www.deathriskrankings.com Franklin, M., Zeka, A., and J. Schwartz. 2007. Association between PM2.5 and all cause and specific cause mortality in 27 US communities. Journal of Exposure Science and Env ironmental Epidemiology (2007) 17, 279 287 Fox, A.J., Goldblatt, P. (1982) Longitudinal study: socio demographic mortality differentials 1971 1975, London, OPCS. Gentleman, Amelia. 2007. New Delhi's vertical ambition aims to ease overcrowding NY Times, Accessed X, 2012: http://www.nytimes.com/2007/03/12/world/asia/12iht delhi.4881178.html?pagewanted=all Gupta, R., Misra, A., Pais, P., Rastogi, P., and V.P. Gupta. 2005. Correlation of regional cardiovascular disease mortality in India with lifestyle and nutritional factors. International Journal of Cardiology. Volume 108, Issue 3, 14 April 2006, Pages 291 300. http://www.sciencedirect.com/science/article/pii/S0167527305007527

PAGE 261

243 Garg, A. 2011. Pro equity Effects of Ancillary Benefits of Climate Change Policies: A Case Study of Human Health Impacts of Outdoor Air Pollution in New D elhi, World Development, Volume 39, Issue 6, June 2011, Pages 1002 1025, ISSN 0305 750X, 10.1016/j.worlddev.2010.01.003. (http://www.sciencedirect.com/science/article/pii/S0305750X1100060X) Greenprint Denver, 2011. www. greenprintdenver .org Government of NCT of Delhi. 2006. Delhi Human Development Report: Partnerships for Progress. Oxford University Press Glennon, C, Kakah, M., Singh, M., Dhokaria, M, Shafiq, N, Vist, G, Underland, V., and J. Odegaard Jansen. An overview of systematic reviews for interventions aimed at reducing child mortality in low middle income countries (draft for discussion). WHO. www.who.int/rpc/meetings/Child_Mo rtality_Overview.pdf Gan, W. Traffic related Air Pollution, Community Noise, and Coronary Heart Disease. Doctoral Thesis. University of British Columbia, Vancouver. Global Burden of Disease, 2011 Transport Accidents in top ten causes of death globally Greenbaum, D. 2011. Vehicle Emissions and Health: A Global Perspective on Effects, Placed in an Indian Context. Health Effects Institute Presentation in Delhi, India on Sept. 1, 2011. Gunther, I. and G. Fink. 2010. Water, Sanitation and Children's healt h: Evidence from 172 DHS Surveys. Policy Research Working Paper; WPS 5275. Gunther, I. and G. Fink. 2011. World Bank: Water and sanitation to reduce child mortality: the impact and cost of water and sanitation infrastructure. Policy Research Working Paper ; WPS 5618. http://go.worldbank.org/LUKE643VQ0 Gupta, R., Guptha, S., Joshi, R., and D. Xavier. 2011. Translating evidence into policy for cardiovascular disease control in India. Health Research Policy a nd Systems 2011, 9 :8. doi:10.1186/1478 4505 9 8. Accessed Jan, 2012: http://www.health policy systems.com/content/9/1/8 Gupta D, Boffetta P,Gaborieau V, Jindal SK. 2001. Risk factors of lung cancer in Chandigarh, India. Indian J Med Res 2001;113:142 50. Hansen, J., M. Sato, P. Kharecha, G. Russell, D.W. Lea and M. Siddall. 2007. Climate change and trace gases Philosophical Transactions of the Royal Society of London A 365:1925 1954.

PAGE 262

244 Health Effects Institute Scientific Oversight Committee. 2004. Health Effects of Outdoor Air Pollution in Developing Countries of Asia: A Literature Review. Special Report 15. Health Effects Institute, Boston MA. Hedley Environmental Index, 2012. http://hedleyindex.sph.hku.hk/home.php Hendriks, C., Worrell, E., de Jager, D., Blok, K., and P. Riemer. 2004. Emissions reduction of greenhouse gases from the cement industry. Greenhouse Gas Control Technologies Co nference. Hooper, L., Summerbell, C., Higgins, J., Thompson, R., Capps, N., Smith, G., Riemersma, R., and S. Ebrahim. 2001. Dietary fat intake and prevention of cardiovascular disease: systematic review. BMJ Volume 322: 757 63. 21 March 2001. Hosgood, H. Dean, III. 2011. Household coal use and lung cancer: systematic review and meta analysis of case control studies, with an emphasis on geographic variation INTERNATIONAL JOURNAL OF EPIDEMIOLOGY, JUN 2011. Ichikawa M, Chadbunchachai W, Marui E. Effect of the helmet act for motorcyclists in Thailand. Accident Analysis & Prevention 2003; 35(2):183 189. ICLEI SA: http://www.iclei.org/index.php?id=970 International Council on Clean Transportation. 2010. A policy relevant summary of black carbon climate science and appropriate emission control strategies. January 2010. 10 pp. Jacobson, M.Z. 2005. Correction to Control of fossil fuel particulate black carbon and organic matter, possibly the most effective method of slowing global warming. Journal of Geophysical Research 110:D14105. Jacobson, M.Z. 2009. Short term effects of controlling fossil fuel soot, biofuel soot and gases, and methane on climate, Arctic ice, and air pollution health. Manuscript under r eview. 58 pp. http://www.stanford.edu/group/efmh/ jacobson/ Last viewed Feb. 20, 2010. Jacobson, M.Z. 2007. Testimony for the Hearing on Black Carbon and Global Warming House Committee on Oversight and Government Reform United States House of Representati ves The Honorable Henry A. Waxman, Chair. October 18, 2007 Jacobson, 2008. On the Causal Link Between Carbon Dioxide and Air Pollution Mortality. Geophysical Research Letters, Vol. 35, LO3809, doi: 10.1029/2007GL031101 James JW. Longitudinal study of the morbidity of diarrheal and respiratory infections in malnourished children. Am J Clin Nutr 25, 690 694, 1972. Jannsen, N., Hoek, G., Simic Lawson, M., Fischer, P., Bree, L., et al. 2011. Black Carbon as an Additional Indicator of the Adverse Health Effec ts of Airborne Particles Compared

PAGE 263

245 with PM10 and PM2.5. Environ Health Perspect 119:1691 1699 (2011). http://dx.doi.org/10.1289/ehp.1003369 [Online 2 August 2011] Jerrett M, Burnett RT, Pope CA 3rd, et al. Long term ozone exposure and mortality. N Engl J M ed 2009; 360: 1085 95. Jeuland, M., Fuente, D., Ozdemir, S., Allaire, M., and D. Whittington. 2013. The long term dynamics of mortality benefits from improved water and sanitation in less developed countries. PloS ONE 8(10): e74804. Ji, M., Cohan, D., a nd M. Bell. 2011. Meta analysis of the association between short term exposure to ambient ozone and respiratory hospital admissions. Environ. Res. Let.. 6 024006 (11pp) Available Online (Feb, 2012) at stacks.iop.org/ERL/6/024006 Johns Hopkins Bloomberg Sc hool of Public Health. LiST: The Lives Saved Tool: An evidence based tool for estimating intervention impact. Accessed November, 2011: http://www.jhsph.edu/dept/ih/IIP/list/ Katsouyanni K, Touloumi G, Spix C, Schwartz J, Balducci F, Medina S, Rossi G, W ojtyniak B, Sunyer J, Bacharova L, Schouten JP, Ponka A, Anderson HR. 1997. Short term effects of ambient sulphur dioxide and particulate matter on mortality in 12 European cities: Results from time series data from the APHEA project. BMJ 314:1658 1663. Ac cessed Feb 2012: http://www.bmj.com/content/314/7095/1658 Key, T.J. 2011. Fruit and vegetables and cancer risk. British Journal of Cancer (2011) 104 6 11. doi:10.1038/sj.bjc.6606032 Published online 30 November 2010 http://www.nature.com/bjc/journal/v104/n1/full/6606032a.html Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi Y, Turner MC, Pope CA III, Thurston G, Calle E E, Thun MJ. 2009. Extended Follow Up and Spatial Analysis of the American Cancer Society Study Linking Particulate Air Pollution and Mortality. HEI Research Report 140. Health Effects Institute, Boston, MA. Lamberti, L., Walker, C. and R. Black. 2012. Sys tematic review of diarrhea duration and severity in children and adults in low and middle income countries. BMC Public Health 2012, 12:276. Laden, F., Schwartz, J., Spelzer, F., and D. Dockery. 2006. Reduction in Fina Particulate Air Pollution and Mortal ity: Extended Follow up of the Harvard Six Cities Study. Am J Respir Crit Care Med Vol 173. pp 667 672, 2006 Leon, D.A., Davey Smith, G. (2000) Infant mortality, stomach cancer, stroke and coronary heart disease: ecological analysis', British Medical Jou rnal, 320, 1705 1706.

PAGE 264

246 Lienhardt C (2001). From exposure to disease: the role of environmental factors in susceptibility to and development of tuberculosis. Epidemiological Reviews, 23(2):288 300. Liu, S., Manson, J.E., Lee, I.M, Cole, SR, Hennekens, CH, Willett WV, and JE Buring. 2000. Fruit and vegetable intake and risk of cardiovascular disease: the Women's Health Study. Accessed 2012: http://www.ncbi.nlm.nih.gov/pubmed/11010932 Liu, CC, Chen, C.C., Wu, T.N., and CY Yang. Association of brain cancer with residential exposure to petrochemical air pollution in Taiwan. 2008. J Toxicol Environ Health A. (71)5:310 4. http://www.ncbi.nlm.nih.gov /pubmed/18214804 Lock, K., Pomerleau, J., Causer, L., Altmann, D., and M. McKee. 2005. The global burden of disease attributable to low consumption of fruit and vegetables: implications for the global strategy on diet. Bulletin of the World Health Organ ization / February 2005, 83 (2). Accessed Dec., 2011: http://www.who.int/bulletin/volumes/83/2/100.pdf Lubin JH, Boice JD Jr (1997). Lung cancer risk from residential radon: meta analysis of eight epidemiologic studies. Journal ofthe National Cancer Institute, 89(1):49 57. Malig,B., and B. Ostro. 2009. Coarse particles and mortality: evidence from a multi city study in California. Occup Environ Med; 66: 832 839. http://oem.bmj.com/content/66/12/832.abstract MacCarty, N., D. Ogle, D. Still, T. Bond and C. Roden. 2008. A laboratory comparison of the global warming impact of five major types of biomass cooking stoves. Energy for Susta inable Development 12(2):56 65. Marinaccio, A, Scarselli, A, Binazzi, A, Mastrantonio, M, Ferrante, P, and S Iavicoli. 2008. Magnitude of asbestos related lung cancer mortality in Italy. Br J Cancer http://dx.doi.org/10.1038/sj.bjc.6604450 http://www.nature.com/bjc/journal/v99/n1/full/6604450a.html McMichael AJ, Campbell Lendrum D, Wilkinson P, Wilson T, Nicholls R, Hales S, Tanser F, Le Sueur D, Schlesinger M, AndronovaN (2004). G lobal climate change. In: Ezzati M, Lopez AD, Rodgers A, Murray CJL, eds. Comparative quantification of health risks. Geneva, World Health Organization. Miettinen OS (1974). Proportion of disease caused or prevented by a given exposure, trait or intervent ion. American Journal of Epidemiology, 99:325 332. Mohan D (2004). The road ahead traffic injuries and fatalities in India. Delhi, Indian Institute of Technology.

PAGE 265

247 National Cancer Institute. 2012. Asbestos Exposure and Cancer Risk. http://www.cancer.gov/cancertopics/factsheet/Risk/asbestos National Capital Territory of Delhi, 2008. Socio Economic Profile of Delhi: 2007 2008. Planning Department. www.delhiplanning.nic.in National Crime Records Bureau (NCRB). 2008. Accidental Deaths and Suicides 2008. Ministry of Home Affairs (Table 1.8: Traffic Accident Fatalities Summary for Delhi, 2008): http://ncrb.nic.in/adsi2008/home.htm National Research Council. 1999. Health Effects of Exposure to Radon. Committee on Health Risks of Exposure to Radon www.nap.edu/catalog.php?record_id=5499 Novoa, A, Perez, K and C. Borrell. 2009. Evidence based effectiveness of road safety interventions: a literature review. www.ncbi.lm.nih.gov/pubmed/19896245 Ostro B, Chestnut L, Vichit Vadakan N, Laixuthai A. 1999. The impact of particulate matter on d aily mortality in Bangkok, Thailand. J Air Waste Manage Assoc 49:PM100 PM107. http://www.ncbi.nlm.nih.gov/pubmed/11002832 Ostro, B., Broadwin, R., Green, S., Feng, WY, and M. Lipsett. 2006. Fine pa rticulate air pollution and mortality in nine California counties: results from CALFINE. Environ Health Perspect.; 114 (1):29 33. Owen, Christopher G., Whincup, Peter H., Orfei, Lia, Chou, Qurratul Ain, et al.2009. Is body mass index before middle age rel ated to coronary heart disease risk in later life? Evidence from observational studies. International Journal of Obesity, August 2009 33(8); pp866 877. Parent, M., Desy, M., and J. Siemiatycki. 2009. Does Exposure to Agricultural Chemicals Increase the Ri sk of Prostate Cancer among Farmers? Mcgill J. Med. 2009 January; 12(1): 70 77. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2687920/ Pavia M, Bianco A, Pileggi C, Angelillo IF (2003). Met a analysis of residential exposure to radon gas and lung cancer. Bulletin of the World Health Organization, 81(10):732 738. Princeton Carbon Mitigation Initiative. 2011. Stabilization Wedges Introduction Accessed May, 2013 at: http://cmi.princeton.edu/wedges/intro.php Pope, CA, Thun, M., Namboodiri, M., Docerky, D., Evans, J., Speizer, F., and C. Heath. 1995. Particulate air pollution as a predictor of mortality in a prospective study of U.S. adult s. Am J Respir Crit Care Med. 1995 Mar; 151 (3 Pt 1):669 74.

PAGE 266

248 Pope, C. Burnett, R., Thun, M., Calle, E., Krewski, D., Ito, K. and G. Thurston. 2002. Lung Cancer, Cardiopulmonary Mortality, and Long term Exposure to Fine Particulate Air Pollution. JAMA. 20 02;287(9):1132 1141. doi: 10.1001/jama.287.9.1132 Pope, CA, Burnett, R., Thurston, G., Thun, M., Calle, E., Krewski D., and J. Godleski. 2004. Cardiovascular Mortality and Long Term Exposure to Particulate Air Pollution: Epidemiological Evidence of Genera l Pathophysiological Pathways of Disease. Circulation, 2004 Pope CA III, Burnett RT, Turner MC, Cohen A, Krewski D, Jerrett M, et al. 2011. Lung Cancer and Cardiovascular Disease Mortality Associated with Ambient Air Pollution and Cigarette Smoke: Shape o f the Exposure Response Relationships. Environ Health Perspect 119:1616 1621. http://dx.doi.org/10.1289/ehp.1103639 Poppel, F., and C. Heijden. 1997. The effects of water supply on infant and child morta lity: a review of historical evidence. Health Transition Review 7, 1997, 113 148. Radin .U.R.S., MacKay GM, Hills BL. Modelling of conspicuity related motorcycle accidents in Seremban and Shah Alam, Malaysia. Accident Analysis & Prevention 2003; 28(3):32 5 332. Rajarathnam U, Seghal M, Nairy S, Patnayak RC, Chhabra SK, Kilnani, Santhosh Ragavan KV. 2011. Part 2. Time series study on air pollution and mortality in Delhi. In: Public Health and Air Pollution in Asia (PAPA): Coordinated Studies of Short Term Exposure to Air Pollution and Daily Mortality in Two Indian Cities. Research Report 157. Health Effects Institute, Boston, MA. Rao C, Lopez AD, Yang G, Begg S, Ma J. Evaluating national cause of death statistics: principles and application to the case of China. Bull World Health Organ 2005;83:618 25. PMID:16184281 Rastogi, T., Vaz, M., Spiegelman, D., Reddy, K., Bharathi, A., Stampfer, M., Willett, W., and A. Ascherio. 2004. Physical activity and risk of coronary heart disease in India. International Jour nal of Epidemiology 33: 759 767. Doi: 10.1093/ije/dyh042 Samet JM, Dominici F, Zeger SL, Schwartz J. Dockery DW. 2000. The National Morbidity, Mortality, and Air Pollution Study. Part I: methods and methodologic issues. Res Rep Health Eff Inst 2000; 75: 5 14. Samet JM, Dominici F, Curriero FC, Coursac I, Zeger SL. 2000. Fine particulate air pollution and mortality in 20 U.S. cities, 1987 1994. N Engl J Med. 2000;343(24):1742 1749. Samet JM, Dominici F, Curriero FC, Coursac I, Zeger SL. Fine particulate air pollution and mortality in 20 US cities, 1987 1994. N Engl J Med 2000; 343: 1742 49.

PAGE 267

249 Schartz, J. 1994. Air Pollution and Daily Mortality: A Review and Meta Analysis. Environmental Research 64, 36 52 Schwartz J. 1993. Air Pollution and Daily Mortalit y in Birmingham, Alabama. Am J Epidemiol 137:1136 1147. Accessed Feb 2012: http://aje.oxfordjournals.org/content/137/10/1136 Smith, K., Jerrett, M., Anderson, H. R., Burnett, R. T., Stone, V ., Derwent, R., et al. (2009). Public health benefits of strategies to reduce greenhouse gas emissions: health implications of short lived greenhouse pollutants. The Lancet. doi:10.1016/S0140 6736(09) 61716e5, Published Online November 25, 2009. Smith, A., Lingas, E., & Rahman, M. (2000). Contamination of drinking water by arsenic in Bangladesh: a public health emergency. Bulletin of the World Health Organization, (9), 78. Smith, K., 2000. National burden of disease in India from indoor air pollution. P NAS vol. 97 no. 24, pp 13286 13293 Smith KR, Corvalˆn FC, Kjellstršm T (1999). How much ill health is attributable to environmental factors? Epidemiology, 10(5):573 584. Also available online at http://www.who.int/quantifying_ehimpacts/methods/en/smith.pd f Smith KR, Mehta S, Maeusezahl Feuz M (2004). Indoor air pollution from solid household fuels. In: Ezzati M, Lopez AD, Rodgers A, Murray CJL, eds. Comparative quantification of health risks. Geneva, World Health Organization. Smith, R., Xu, B., and P. S witzer.2009. Reassessing the relationship between ozone and short term mortality in U.S. urban communities. Inhal Toxico. Suppl 2: 37 61. Smith, K. National Burden of Disease in India from Indoor Air Pollution. 2000. PNAS. Vol. 97, no. 24; 13286 13293 Published November 21, 2000 Snilstveit B. and Waddington H. (2009), "Effectiveness and sustainability of water, sanitation, and hygiene interventions in combating diarrhoea", Journal of Development Effectiveness, Vol 1 Issue 3: 295 335. Sperling, J. B., & Ramaswami, A. 2013. Exploring health outcomes as a motivator for low carbon city development: Implications for infrastructure interventions in Asian cities, Habitat International, doi:10.1016/j.habitatint.2011.12.013 Sridhar, K. 2010. India Infrastructu re Report: Carbon Emissions, Climate Change, and Impacts in India's Cities. http://www.idfc.com/pdf/report/Chapter 21.pdf Suglia, S. F., et al. (2008a), Association between traffic related black carbon exposure and lung function among urban women,

PAGE 268

250 Environmental Health Perspectives 116 (10), 1333 1337. Suglia, S. F., et al. (2008b), Association of Black Carbon with Cognition among Children in a Prospective Birth Cohort Study., American Journal of Epidemiology 167 (3), 280 286. Tarlov, A.R., Public Policy Frameworks for Improving Population Health. Annals of the New York Academy of Sciences, 1999. 896 Socioecononomic Status and Health in Industrial Nations: Social, Psychological, and Biological P athways: p. 281 293 UK Deputy Prime Minister Office. 2004. The Impact of Overcrowding on Health & Education: A Review of Evidence and Literature. Walker, C and R. Black. 2010. Diarrhoea morbidity and mortality in older children, adolescents, and adults Epidemiol Infect. 2010 Sep; 138 (9):1215 26. UN Habitat, The Challenge of an Urban World. http://www.intlhc.org/docs/The_Challenge_of_an_Urban_World.pdf UN Habitat. 2008. Th e State of African Cities: A Framework for addressing urban challenges in Africa. http://www.unhabitat.org/pmss/listItemDetails.aspx?publicationID=2574 USAID, 2010. Black Carbon Emissions in Asia: Sources, Impacts, and Abatement Opportunities. http://cleanairinitiative.org/portal/sites/default/files/ USAID_RDMA_Black_Carbon_Emi ssion_in_Asia_Apr2010.pdf Waddington, H., Snilstveit, B., White, H., & Fewtrell, L. (2009). Water, sanitation and hygiene interventions to combat childhood diarrhoea in developing countries. International Initiative for Impact Evaluation. New Delhi, India: 3ie. Wagner, Gilson, Berry and Timbreell. 1971. Epidemiology of Asbestos Cancers. http://members.optusnet.com.au/geoffberry1/GB%20Web_files/Pub%2019.pdf Welborn, TA, Dhaliwal, SS., and SA Bennett. 2003. Waist hip ratio is th e dominant risk factor predicting cardiovascular death in Australia. Med J Aust. 2003 Dec 1 15;179(11 12):580 5. Access December, 2011: http://www.ncbi.nlm.nih.gov/pubmed/14636121 World Health Org anization (WHO). 2009. Global Health Risks: Mortality and burden of disease attributable to selected major risks. WHO, 2000. Obesity: preventing & managing the global epidemic: report of a WHO consultation. WHO Technical Report Series: 894 WHO, 2006. Gui delines for Treatment of Malaria: Second Edition

PAGE 269

251 http://www.who.int/malaria/publications/atoz/9789241547925/en/index.html WHO, 2011. Air Quality and Health: Fact Sheet 313. Accessed November, 2011: http://www.who.int/mediacentre/factsheets/fs313/en/ WHO, 201x. Risk Factors for Chronic Respiratory Diseases. WHO, 2009. Global Health Risks: A response to th e need for comprehensive, consistent and comparable information on health risks at global and regional level. http://www.who.int/healthinfo/global_burden_disease/global_health_risks/en/index.html WHO. 2004. World report on road traffic injury prevention. Accessed March 2011. www.who.int/violence_injury_prevention/publications/road_traffic/world_report/en/inde x.html WHO. 2000. Part II: Establishing the true costs of the problem of overweight and obesity. WHO. 2011 Diabetes: Fact Sheet N312. http://www.who.int/mediacentre/factsheets/fs312/en/ WHO/FAO, Joint WHO/FAO Expert Consultation on Diet, Nutrition and the Prevention of Chronic Diseases, in WHO Technical Report Series. 2003, WHO: G eneva. p. 95 104. Yang, Y., Runkui, L., Wenjing L., Wang, M., Cao, Y., Wu, Z., and Q. Xu. 2013. The Association between Ambient Air Pollution and Daily Mortality in Beijing after the 2008 Olympics: A Time Series Study. PLoS ONE 8(10): e76759. Caulfied, L., de Onis, M., Blossner, M. and R Black. 2004. Undernutrition as an underlying cause of child deaths associated with diarrhea, pneumonia, malaria, and measles. Am J Cli Nutri. 2004 July; 80(1):193 8. http://www.ncbi.nlm.nih.gov/pubmed/15213048 Zachariah, R., M. Spielmann, A. Harries, and F. Salaniponi. Moderate to severe malnutrition in patients with tuberculosis is a risk factor associated with early death. Trans R Soc Trop Med Hyg. 2002 May Ju n; 96(3):291 4. http://www.ncbi.nlm.nih.gov/pubmed/12174782 Conn, V. and T. Ruppar. 2010. Interventions to Promote Physical Activity in Chronically Ill Adults. American Journal of Nursing, July 201 0. http://www.nursingcenter.com/prodev/ce_article.asp?tid=1034217 Wong C M, Lai H K, Ou C Q, Ho S Y, Chan K P, et al. (2008) Is Exercise Protective Against Influenza Associated M ortality? PLoS ONE 3(5): e2108. doi:10.1371/journal.pone.0002108. http://www.plosone.org/article/info:doi/10.1371/journal.pone.0002108

PAGE 270

25 2 Mollah, K. and T. Aramaki. 2010. Soc ial epidemiological study for evaluation of water supply and sanitation systems of low income urban community in Dhaka, Bangladesh. Journal of Water and Health. http://www.iwaponline.com/jwh/0 08/0184/0080184.pdf Kanarek, M and T. Young.1982. Drinking Water Treatment and Risk of Cancer Death in Wisconsin. Environmental Health Perspectives, Vol. 46, pp. 179 186. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1569038/pdf/envhper00463 0177.pdf Implications for diarrhea disease control and environmental health indicators.2004. Chapter 3: International Review of Diarrheal Disease Control Studies. Pp 18 25 http://www.bvsde.paho.org/bvsacd/eco/031900/031900 04.pdf Blumenthal, U. and A. Peasey. 2002. Critical review of epidemiological evidence of the health effects of wastewater and excreta u se in agriculture. London School of Hygiene and Tropical Medicine. http://www.who.int/water_sanitation_health/wastewater/whocriticalrev.pdf NAE, 2009. Vulnerable Infras tructure and Waterborne Disease Risk. Global Issues in Water, Sanitation, and Health: Workshop Summary. Institute of Medicine (US) Forum on Microbial Threats. http://www.ncbi.nlm.nih.gov/books/NBK284 63/ Fobil et al., 2011. Neighborhood Urban Environmental Quality Conditions Likely to Drive Malaria & Diarrhea Mortality in Accra, Ghana. Journal of Environmental and Public Health Volume 2011, Article ID 484010 doi: 10.1155/2011/4841010 http://www.hindawi.com/journals/jeph/2011/484010/

PAGE 271

253 Chapter 4 Sen, A. (1998). Mortality as an indicator of economic success and failure. Economic Journal, Royal Economic Society, 108(446), 1e25. UN, 2013. G oal 4 Reduce Child Mortality: Target 4.A Reduce by two thirds, between 1990 and 2015, the under five mortality rate. Accessed July 2013: http://www.un.org/millenniumgoals/childhealth.shtml UNICEF, 2008. Millennium Development Goals Child Mortality. Ac cessed June 2013: http://www.unicef.org/mdg/index_childmortality.htm The World Bank: Under Five Mortality Rates. 2013. Accessed June 2013: http://data.worldbank.org/indicator/SH.DYN.MORT Claeson, M. 2000. Reducing child mortality in India in the new millennium. World Health Organization 78 (10) 1192. www.who.int/b ulletin/archives/78(10)1192.pdf Agarwal, S and S. Taneja. 2005. All Slums Are Not Equal: Child Health Conditions Among the Urban Poor. USAID EHP Urban Health Program. Indian Pediatr., 2005 Mar; 42(3):233 44. Chandramouli, C. 2013. Housing Stock, Ameni ties and Assets in Slums Census 2011. Available Dec 30 2013: www.censusindia.gov.in/2011 Documents/On_Slums 2011Final.ppt Gupta, K., Arnold, F. and H. Lhungdim. 2009. Hea lth and living conditions in Eight Indian Cities. Indian Ministry of Health. Accessed Feb 2013: http://www.measuredhs.com/pubs/pdf/OD58/OD58.pdf National Family Health Survey 3 (NFHS 3) | Demographic and Health Survey (DHS): India. 2006. Available Jan 2013 at: http://www.rchiips.org/NFHS/nfhs3.shtml and www.measuredhs.org UN Habitat, 2010/11. State of the World's Cities: Bridging the U rban Divide. http://www.unhabitat.org/pmss/listItemDetails.aspx?publicationID=2917 UN Habitat, 2006/7. State of the World's Cities. Slums: Past, Present, and Future. Available Jun 2013: http://www.unhabitat.org/documents/media_centre/sowcr2006/SOWCR%204.pdf Mohanty, S. 2009. Alternative wealth indices and health estimates in India. Genus, Vol65, No2, pp113 137

PAGE 272

254 Bharati, P, Bharati, S. et al. 2008. Chronic Energy Deficie ncy Among Indian Women by Residential Status. Ecology of Food and Nutrition. http://www.measuredhs.com/publications/journal details.cfm?article_id =613&C_id=26&T_ID=0&P_ID=0 Mollah, K. and T. Aramaki. 2010. Social epidemiological study for evaluation of water supply and sanitation systems of low income urban community in Dhaka, Bangladesh. Journal of Water and Health. http://www.iwaponline.com/jwh/008/0184/0080184.pdf Goli, S., Arokiasamy, P. et al. 2011. Living and health conditions of selected cities in India: Setting priorities for the National Urban Health Mission. Cities Volume 28, Issue 5, October 2011, Pages 461 469. Government of India, Ministry of Health and Family Welfare. 2010. Annual Report to the People on Health. http://mohfw.nic.in/W riteReadData/l892s/9457038092AnnualReporthealth.pdf Das, K.C. and V. Gautam. 2013. Household Environmental Factors and Its Effects on Child Morbidity in Madhya Pradesh. Journal of History and Social Sciences. Vol IV, Issue 1, Jan Jun 2013. Ed. P. Tiwari Agarwal, S. and A. Srivastava. 2009. Social Determinants of Children's Health in Urban Areas. Journal of Health Care for the Poor and Underserved. http://www.measuredhs.com/publications/journal details.cfm?article_id=1168&C_id=26&T_ID=0&P_ID=0 Gunther, I. and G. Fink. 2010. Water, Sanitation and Children's health: Evidence from 172 DHS Surveys. Policy Research Working Paper; WPS 5275. Fink, G ., Gunther, I., and K. Hill. 2011. The effect of water and sanitation on child health: evidence from the demographic and health surveys 1986 2007. Int. J. Epidemiol. 2011;1 9. Oxford University Press. Van Poppel, F. and C. Heijden. 1997. The effects of wa ter supply on infant and childhood mortality: a review of historical evidence. Health Transition Review 7, 1997, 113 148. Kaiser, H. 1974. An Index of Factorial Simplicity. Psychometrika, March 1974 Vol 39 Issue 1 pp 31 36 Neill, J. 2008. Exploratory Fac tor Analysis. University of Canberra. Lecture on Mar 25, 2008. Available Nov 2013: http://www.slideshare.net/jtneill/exploratory factor analysis

PAGE 273

255 Clark, M. 2009. Correlation. Uni versity of North Texas. Lecture available Dec 2013 at: http://www.unt.edu/rss/class/mike/index.html Moss, W., Darmstadt, G., Marsh, D., Black, R., and M. Santosham. 2002. Research Priorities for the Reduction of Perinatal and Neonatal Morbidity and Mortality in Developing Country Communities. Journal of Perinatology, Sept 2002, Vol 22, Num 6, pp 484 495. Stoll, B. 1997. The Global Impact of Neonatal Infection. Clin Perinatol; 24:1 21. http://www.ncbi.nlm.nih.gov/pubmed/9099499?dopt=Abstract&holding=npg Bang, A., Bang, R., Baitule, S. Reddy, M. and M. Deshmukh. 1999. Effect of home based neonatal care and manageme nt of sepsis on neonatal mortality: field trial in rural India. Lancet; 354: 1955 61. http://www.ncbi.nlm.nih.gov/pubmed/10622298?dopt=Abstract&holding=npg Lawn, J., Wilczynska Ketende, K. and S. Cousens. 2006. Estimating the causes of 4 million neonatal deaths in the year 2000. Int. J. Epidemiol. (June 2006) 35 (3): 706 718. http://ije.oxfordjournals.org/content/35/3/706.long {Additional Relevant References Below Not Included in Final Chapter} Agarwal, S. and A. Srivastava. 2009. Social Determinants of Children's Health in Urban Areas. Journal of Health Care for the Poor and Underserved. http://www.measuredhs.com/publications/journal details.cfm? article_id=1168&C_id=26&T_ID=0&P_ID=0 Antai, D. and T. Moradi. 2010. Urban Area Disadvantage and Under 5 Mortality in Nigeria: The Effect of Rapid Urbanization. Environ Health Perspect. 2010 June; 118(6) 887 883 Bhan N, Srivastava S, Agrawal S, Subrama nyam M, Millett C, Selvaraj S, Subramanian SV. 2012. Are socioeconomic disparities in tobacco consumption increasing in India? A repeated cross sectional multilevel analysis. BMJ Open 2(5). doi: 10.1136/bmjopen 2012 001348. Bhargava, A., 2011. Health care utilization, socioeconomic factors and child health in India.J. Biosoc Sci. Claeson, M. 2000. Reducing child mortality in India in the new millennium. World Health Organiza tion 78 (10) 1192. www.who.int/bulletin/archives/78(10)1192.pdf Cleland, J. and J. Ginneken. 1988. Maternal education and child survival in developing countries: The search for pathways of influence Social Science & Medicine, Volume 27, Issue 12, 1988, Pages 1357 1368

PAGE 274

256 Das, K. and V. Gautam. 2012. Household environmental factors and its effects on child morbidity in Madhya Pradesh. Journal of History and Social Sciences. Volume III, Issu e 1. Fotso, J., and B. Kuate Defo. 2005. Measuring Socioeconomic Status in Health Research in Developing Countries: Should We Be Focusing on Households, Communities or Both? Social Indicators Research, v72 n2 p189 237 Jun 2005 Gupta, K. 2009. Health and l iving conditions in Eight Indian Cities. Indian Ministry of Health. http://www.measuredhs.com/pubs/pdf/OD58/OD58.pdf Goli, S., Arokiasamy, P. et al. 2011. Living and health conditions of sele cted cities in India: Setting priorities for the National Urban Health Mission. Cities Volume 28, Issue 5, October 2011, Pages 461 469. Ghosh, S. 2003. Demographic and Socioeconomic correlates of Neonatal, Post neonatal and Childhood Mortality in Uttar Pr adesh, India: A Study based on NFHS 2 Data. Journal of Health and Population in Developing Countries. www.jhpdc.unc.edu Gunther, I. and K. Harttgen. 2012. Deadly Cities? Spatial Inequalities in Mortality in sub Saharan Africa. Population and Development R eview, Vol 38 Issue 3. http://www.measuredhs.com/publications/journal details.cfm?article_id=1240&C_id=0&T_ID=56&P_ID=0 Hazarika, I. 2010. Wom en's reproductive health in slum populations in India: evidence from NFHS 3. J Urban Health. 2010 Mar; 87 (2): 264 77. India Demographic and Health Survey. 2005 6. http://www.measuredhs.com/Publications/Publication Search.cfm?ctry_id=57&country=India Jankowska, M. M. Benza, and J. Weeks. 2013. Estimating spatial inequalities in urban child mortality. Demographic Research, Volume 28 Article 2, Pages 33 62. Kumar C, Singh PK, Rai RK (2012) Under Five Mortality in High Focus States in India: A District Level Geospatial Analysis. PLoS ONE 7(5): e37515. doi:10.1371/journal.pone.0037515 Lauridsen, J. and J. Pradhan. 2011. Socio economic inequality of immunizat ion coverage in India. Health Economics Review, 1: 11 doi: 10.1186/2191 1991 1 11. http://www.healtheconomicsreview.com/content/1/1/11 Mishra V, Retherford RD, Smith KR (2005) Cooking smok e and tobacco smoke as risk factors for stillbirth. International Journal of Environmental Health Research 15(6):397 410.

PAGE 275

257 Mohanty, S. 2009. Alternative wealth indices and health estimates in India. Genus, Vol 65, No 2, pp. 113 137 National Family Health Survey, India. 2006. http://www.rchiips.org/NFHS/nfhs3.shtml Pathak,P. and A. Singh. 2011. Trends in malnutrition among children in India: Growing inequalities across different economic groups, Social Science & Medicine, Volume 73, Issue 4, August 2011, Pages 576 585, ISSN 0277 9536, 10.1016/j.socscimed.2011.06.024. Subramanian SV, Smith GD, Subramanyam M (2006) Indigenous Health and Socioeconomic Status in India. PLoS Med 3(10): e421. doi:10.1371/jou rnal.pmed.0030421 Chapter 5 Agarwal, S and S. Taneja. 2005. All Slums Are Not Equal: Child Health Conditions Among the Urban Poor. USAID EHP Urban Health Program. Indian Pediatr., 2005 Mar; 42(3):233 44. Allianz 2013 World's Fastest Growing Megacitie s. http://knowledge.allianz.com/demography/population/?2079/The worlds fastest growing megacities Bartlett, J. Kotrlik, J, and C Higgins. 2001. Organizational Research: Determining Appropriate Sample Size in Survey Research. Information Technology, Learning, and Performance Journal, Vol. 19, No. 1, Spring 2001 Brebbia, C., Popov, V., and D. Fajzieva. 2005. Environmental Health Risk III. WIT Trans actions on Biomedicine and Health, Vol 9, pp 28. Chavez, A. and A. Ramaswami, 2013. "Articulating An Infrastructure Supply Chain Greenhouse Gas (GHG) Emissions Footprint for Cities: Mathematical Relationships and Policy Relevance", Energy Policy. Vol. 54: 376 384. March 2013. http://dx.doi.org/10.1016/j.enpol.2012.10.037 Costello, A, Abbas, M, Allen, et al. 2009. "Managing the h ealth effects of climate change. UCL Institute for Global Health and Lancet Commission", The Lancet (373): 1693 733.

PAGE 276

258 De Hoog, I. Public Perspectives of Extreme Weather Events. 2011. Masters Thesis at Vrije Universiteit, Amsterdam, The Netherlands. http://www.ivm.vu.nl/en/Images/Hoog%20Iris%20de_tcm53 352238.pdf Delhi Pollution Control Committee, 2013. Real Time Ambient Air Quality Data, Govt of NCT of Delhi. http://www.dpcca irdata.com/dpccairdata/display/index.php Dinsa, G. Goryakin, Y., Fumagelli, E. and M. Suhrcke. 2012. Obesity and socioeconomic status in developing countries: a systematic review. Obes Rev. Nov 2012; 13(11): 1067 1079. doi: 10.1111/j.1467 789X.2012.01017.x Accessed Apr 2014 at http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3798095/ Falkingham, J. and C Namzie. 2001. Identifying the poor: A critical review of alternative approaches. London, London School of Economics. Falkingham, J. and C. Namzie. 2002. Measuring health and poverty: a review of approaches to identifying the poor. DFID Health Systems R esource Centre. Accessed April 2014 at: http://r4d.dfid.gov.uk/PDF/Outputs/HOppsIssuesPaperFalkingham.pdf Fox, N, Hunn, A. and N. Mathers. 2009. Sampling and Sample Size Calculation. National Institute for Health Research (NIHR). Gertstman, B. 2014. T Table. Accessed April 2014: http://www.sjsu.edu/faculty/gerstman/StatPrimer/t table.pdf Guttikunda, S. 2013. Air Quality Model for Delhi. http://www.urbanemissions.info Govt. of National Capital Territory o f Delhi. 2008. Level and Pattern of Household Consumer Expenditure in Delhi. Based on N.S.S. 64 th Round (State Sample). Directorate of Economics & Statistics. www.des.delhigotvt.nic.in Hales, S., S. J. Ed wards, and R. S. Kovats. 2003. "Impacts on health of climate extremes." Climate change and health: risks and responses. Geneva, World Health Organization http://www.who.in t/globalchange/publications/climatechangechap5.pdf IC LEI Local Governments for Sustainability: South Asia. 2013. http://www.iclei.org/index.php?id=970 Indian Ministry of Health and Family Welfar e 200 8. District Level Household Survey (DLHS) http://www.rchiips.org/ Jiang, L., Yao, L., Bond, E., Wang, Y. and L Huaung. 2011. Risk perceptions and preparedness of typhoon disaster on coastal inhabitants in China. Am H Disaster Med; Mar Apr; 6(1): 119 26. http://www.ncbi.nlm.nih.gov/pubmed/21678821

PAGE 277

259 Kanti Paul, B. 2011. Environmental Hazards and Disasters: Contexts, Perspectives, and Management. Wiley Blackwell. Lalchandani, 2011. Northeast Distr ict; most densely populate d, poor infrastructure. Times of India. http://articles.timesofindia.indiatimes.com/2011 04 01/delhi/29369 878_1_yamuna vihar sq km population density Lo, A. 2010. China's Response to Climate Change. Environ. Sci. Technol., 44 (15), 5689 5690 http://pubs.acs.org/doi/pdf/10.1021/es101976r McGranahan, G. 2007. Improving Water and Sanitation Services in Deprived Urban Neighborhoods: Avoiding Global Distractions and Pursuing Local Priorities USAID Global Urban Poverty pdf.usaid.gov/pdf_docs/PNADO607.pdf Ministry of Urban Development (MoUD), 2012. Improving urban services through Service Level Benchmarking. http://www.urbanindia.nic.in/programme/uwss/slb/slb.htm MoUD. 2010. Servi ce Level Benchmarks Databook. http://www.urbanindia.nic.in/programme/uwss/slb/Databook/databook.pdf O'Donnell, O., Doorslaer, E., Wagstaff A. and M. Lindelow 2008. "An alyzing Health Equi ty Using Household Survey Data". The World Bank, Washington DC, 2008, www.worldbank.org/analyzinghealthequity Penn State Extension, 2006. How to Determine a Sample Size. Accessed Apr 2014 at: extension.psu.edu/evaluation/ pdf/TS60.pdf Po, J. and S. Subramanian. 2011. Mortality burden and socioeconomic status in India. PLoS One. 2011 Feb 9;6(2):e16844. doi: 10.1371/journal.pone.0016844. Accessed Apr 2014 at: http://www.ncbi.nlm.nih.gov/pubmed/21347373 Ramaswami, A. and S. Dhakal. 2011. Low carbon policies in the USA and China: why cities play a critical role. Carbon Management (2011) 2(4), 359 362 Ramaswami, Anu. 2013. "Understanding Urban Infrastructure Related Greenhouse Gas Emissions and Key Mitigation Strategies". In frastructure & Land Policies, e ds. Ingram, Gregory K., K. L. Brandt. Cambridge, MA: Lincoln Institute of Land Policy Rockefeller Foundation. 2013. Asian Cities Climate Change Resilience Network. www.acccrn.org Sperli ng, J. B., & Ramaswami, A. 2012. Exploring health outcomes as a motivator for low carbon city development: Implications for infrastructure interventions in Asian cities, Habitat International, doi:10.1016/j.habitatint.2011.12.013

PAGE 278

260 Yu, D. and A. Pugliese. 2 012. Majority of Chinese Prioritize Environment Over Economy. Gallup Survey. http://www.gallup.com/poll/155102/majority chinese prioritize environment econ omy.aspx United Nations (UN), 2001. Chapter 6: Population Growth in Cities. http://www.un.org/esa/population/publications/wup2001/WUP2001_CH6.pdf UN, 2011. World Urban ization Prospects. http://esa.un.org/unup/ UN Habitat, 2003. Slums of the World: The face of urban poverty in the new millennium? http://www .un.org/en/sustainablefuture/pdf/Rio+20_FS_Cities.pdf UN Habitat, 2006/7. State of the World's Cities. Slums: Some Definitions. http://www.unhabitat.org/documents/media _centre/sowcr2006/SOWCR%205.pdf UN, 2013. The World We Want 2015. http://www.myworld2015.org/ Chapter 6 International Insitute for Population Sciences (IIPS) and Macro International. 2007. National Family Hea lth Survey (NFHS 3), 2005 06: India: Volume 1. Mumbai: IIPS. http://www.rchiips.org/NFHS/nfhs3.shtml India Demographic and Health Survey. 2005 6. http://www.measuredhs.com/Publications/Publication Search.cfm?ctry_id=57&country=India WHO and UN HABITAT. 2010. Hidden cities: unmasking and overcoming health inequities in urban settings. Kobe: World Health Organization, The WHO Centre for Health Development, Kobe, and United Nations Human Settlements Programme http://www.hiddencities.org/ Rydin, Y., Bleahu, A., Davis, M., Davilla, J. et al. 2012. Shaping cities f or health: complexity and the planning of urban environments in the 21 st century. The Lancet. Gertstman, B. 2014. T Table. Accessed April 2014: http://www.sjsu.edu/faculty/gerstman/StatPrimer/t table.pdf Government of National Capital Territory (NCT) of Delhi. 2008a. Annual Report on Registration of Births and Deaths in Delhi. Government of National Capital Territory (NCT) of Delhi. 2008b. Report Medical Certification of Causes of Deaths in Delhi.

PAGE 279

261 Government of India Central Bureau of Health Intellig ence. 2010. National Health Profile. Accessed July, 2011: www.cbhidghs.nic.in Govt. of National Capital Territory of Delhi. 2008. Level and Pattern of Household Consumer Expenditure in Delhi. Based on N.S.S. 64 th Round (State Sample). Directorate of Economics & Statistics. www.des.delhigotvt.nic.in Gupta, K. 2009. Health and living conditions in Eight Indian Cities. Indian Ministry of Health. http://www.measuredhs.com/pubs/pdf/OD58/OD58.pdf Agarwal, S. 2003. State of Urban Health: Delhi. Urban Health Resource Centre. www.uhrc.in National Crime Records Bureau, 2008. A nnual Delhi Road Safety Report. http://ncrb.nic.in/adsi2008/home.htm Municipal Corporation of Delhi. 2013. www.mcdonline.gov.in Rajarathnam, U et al. 2011. Publi c Health and Air Pollution in Asia (PAPA): Appendix 1: Mortality Registration System in Delhi. International Institute for Population Sciences (IIPS) and Macro International. 2009. National Family Health Survey (NFHS 3), India, 2005 06: Delhi. Mumbai: II PS. Cromley, E. and S. McLafferty. 2011. GIS and Public Health. 2 nd Edition. New York, NY: Guilford Press, 2011 WHO, 2013. World Health Survey. http://www.who.int/healthinfo/survey/en/ US Census B ureau/CDC/NIH, 2012. NIH National Health Interview Surveys. http://www.cdc.gov/nchs/nhis.htm CDC, 2011. Behavioral Risk Factor Surveillance System (BRFSS) http://www.cd c.gov/brfss/ RAND Health. 2010. Medical Outcomes Study (MOS): Measures of Quality of Life Core Survey. http://www.rand.org/health/surveys_tools/mos.html Kohli, R., Sane, S. et al. 2005. M odification of medical outcome study (MOS) instrument for quality of life assessment & its validation in HIV infected individuals in India. Indian J Med Res 122, Oct 2005, pp 297 304

PAGE 280

262 Government of India, Ministry of Health and Family Welfare. 2010. Annual Report to the People on Health. http://mohfw.nic.in/WriteReadData/l892s/9457038092AnnualReporthealth.pdf Census India. 2011. Annual Health Survey. http://censusindia.gov.in/2011 common/AHSurvey.html Agarwal, S. 2010. Maternal Child Health and Nutrition Survey in Urban Slums of Delhi: A Women's Questionnaire. Friel S Kjellstrom T McGranahan G Monge M, Roy J (2010). Urban health inequity: the added stress of climate change. Working Paper Global Research Network on Urban Health Equity. Bartlett, J. Kotrlik, J, and C Higgins. 2001. Organizational Research: Determining Appropriate Sample Size in Survey Research. Information Technology, Learning, and Performance Journal, Vol. 19, No. 1, Spring 2001 Fox, N, Hunn, A. and N. Mathers. 2009. Sampling and Sample Size Calculation. National Institute for Health Research (NIHR). Penn State Extension, 2006. How to Determine a Sample Size. Accessed at: extension.psu.edu/evaluation/pdf/TS60.pdf Gupta, K. Arnold, F. and H. Lhungdim. 2009. Health and Living Conditions in Eight Indian Cities. National Family Health Survey (NFHS 3), India, 2005 06. Mumbai: Internati onal Institute for Population Sciences; Calverton, Maryland, USA: ICF Macro. Bartlett, J. Kotrlik, J, and C Higgins. 2001. Organizational Research: Determining Appropriate Sample Size in Survey Research. Information Technology, Learning, and Performance J ournal, Vol. 19, No. 1, Spring 2001 Chapter 7 Bharati, P, Bharati, S. et al. 2008. Chronic Energy Deficiency Among Indian Women by Residential Status. Ecology of Food and Nutrition. http://www.measuredhs.com/publications/journal details.cfm?article_id=613&C_id=26&T_ID=0&P_ID=0 Cairncross, S., Hunt, C., et al 2010. Water, sanitation and hygiene for the prevention of diarrhea. International Journ al of Epidemiology; 39: i93 i205. Oxford University Press. DOI: 10.1093/ije/dyq035

PAGE 281

263 Chavez, A., Ramaswami, A., Dwarakanath, N., Ranjan, R., & Kumar, E. (2011). Implementing Expanded Geographic Based Greenhouse Gas Accounting for Delhi, India. Data Availabi lity and Methods. Journal of Industrial Ecology. Cropper, M. L., Simon, N. B., Alberini, A., & Sharma, P. K. (1997). The health effects of air pollution in Delhi, India. PRD working paper 1860 (unpublished). New Ideas in Pollution Regulation. Washington, D.C.: World Bank. Accessed November, 2011. www.econ.umd.edu/wcropper/publications/jc36.pdf Gunther, I. and G. Fink. 2010. Water, Sanitation and Children's health: Evidence from 172 DH S Surveys. Policy Re search Working Paper; WPS 5275. Lamberti, L., Walker, C. and R. Black. 2012. Systematic review of diarrhea duration and severity in children and adults in low and middle income countries. BMC Public Health 2012, 12:276 Rajarathnam U, Seghal M, Nairy S, Patnayak RC, Chhabra SK, Kilnani, Santhosh Ragavan KV. 2011. Part 2. Time series study on air pollution and mortality in Delhi. In: Public Health and Air Pollution in Asia (PAPA): Coordinated Studies of Short Term Exposure to Air Pollut ion and Daily Mortality in Two Indian Cities. Research Report 157. Health Effects Institute, Boston, MA. Rushton, L. 2003. Health hazards and waste management. Br Med Bull (2003) 68 (1): 183 197.