PSYCHOSOCIAL DISTRESS AND PERCEIVED RISK AS PREDICTORS OF
SCREENING MAMMOGRAPHY IN WOMEN AT RISK FOR FAMILIAL
Bernadette Anne Pistone
BSN, University of Missouri Columbia, 1985
M.A., Colorado Christian University, 1993
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Health and Behavioral Sciences
2005 by Bernadette Anne Pistone
All rights reserved
This thesis for the Doctor of Philosophy
Bernadette Anne Pistone
has been approved
This thesis for the Doctor of Philosophy
Bernadette Anne Pistone
has been approved
Pistone, Bernadette Anne (Ph.D., Health and Behavioral Sciences)
Psychosocial Distress and Perceived Risk as Predictors of Screening
Mammography in Women at Risk for Familial Breast Cancer
Thesis directed by Professor Joy Berrenberg
Background. First-degree relatives (FDRs) of breast cancer patients are at an
increased risk of developing breast cancer yet only 65-75% of women with FDRs
get the recommended screening mammograms. This study examined the
role that perceived risk, psychosocial distress, and perceived control played in
predicting mammography screening behavior in women with either a single FDR
(SFDR) or multiple FDRs (MFDR). Methods. This study was a secondary
analysis of data from a three-part study done by Marcus et al. in 1996 involving
1082 first-degree relatives of women with breast cancer. A factor analysis was
used to group the large number of variables from a baseline core interview into a
smaller, more manageable set of outcome measures, of which five factors were
extracted. A one-way ANOVA examined differences in psychosocial distress,
perceived risk, and perceived control between SFDRs and MFDRs. Linear and
logistical regression was used to examine the proposed role of, psychosocial
distress, perceived risk, perceived control and attitudes and beliefs about breast
cancer as mediating variables. Two-way ANOVAs were used to examine
relationships between the number of FDRs with breast cancer, psychosocial
distress, perceived risk and sociodemographic factors. Results. Women with
MFDRs reported a higher degree of perceived risk (p=. 000) and more perceived
control than women with a SFDR (p=. 003). Perceived risk was the only variable
that acted a mediator between the number of FDRs with breast cancer and having
had a screening mammogram in the last year. There were no significant
interactions between number of FDRs with breast cancer, psychosocial distress
or perceived risk and sociodemographic factors. Conclusions. More research
needs to be done regarding the role of perceived risk and the decision of women
to get the recommended mammograms. In addition, there are other factors that
need to be explored, such as sociodemographics, to help understand this issue
especially in women at an increased risk of developing breast cancer.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
This thesis is dedicated to my friends and family who provided loving support
throughout this journey. To my friends, Shari, Dave, Reggie, Tracey, and Bruce who
continued to include me in plans even if it was unlikely I could attend.
Lee has been with me throughout the process and she is one of the most caring
people I know. She is a valued friend and confidant.
Tim is one of my most faithful friends who has walked by my side from the very
beginning and has been a true gift from God. I could not have completed this
sometimes painful journey without him. He has never hesitated to hold me up even
during the very difficult times.
A special dedication goes out to my family. My mom and sisters, Angela and
Mary Ellen, consistently called to check in with me and to remind me that I could
finish this thesis. My brother Tim even proofread my entire dissertation while he was
on vacation in Denver. In spite of their busy schedules, my dad and Peter helped me
in completing some paperwork that required several signatures and footwork on their
part. Even Mick and Steve hung in there with me.
And I cannot forget my sweet nieces Sophie, Lily and Carley. Their pictures and
phone calls have helped more than they will know.
Life is nothing without friends and family that love you and want
only the best for you.
I want to begin by sincerely thanking the members of my committee who made it
possible for me to complete this long journey. Joy Berrenberg graciously agreed to
take on the role of chair of my committee after my project had already begun and
quickly became a greatly valued asset. She encouraged me to set the bar higher than I
would have chosen myself and often helped me to stay on the path even during the
times that I felt like I could no longer go on. I truly could not have completed this
degree without her unending support.
Craig Janes was my crucial contact within the Health and Behavioral Sciences
department. Even after accepting a new position in British Columbia, he followed
through with his commitment to remain on my committee and specifically flew to
Denver in order to be physically present for my final defense.
Linda Krebs experience as an oncology nurse was invaluable. I have known
Linda for several years and specifically chose her as a committee member because
she is an example of a professional and knowledgeable nurse that I would like to
Susan Dreisbach agreed to become a member of my committee just 6 months
prior to my final defense and she has played a vital role in that short amount of time.
I have benefited greatly from her participation on my committee.
I could not have completed this degree without the help of Dr. A1 Marcus and his
generous offer to use his dataset for my study. In addition, Veronica Gardner played
a critical role as a statistical consultant and Melody Johnson ensured that I maintained
my sanity during the formatting process.
Connie Steinbrunn is one of the most compassionate and giving people I know.
She was always willing to listen to me (even when she was overwhelmed with her
own workload) and never expressed any resentment in having to cover for me at
work. I feel very gratefiil to have her as a co-worker. Dr. Draznin has supported me
in completing this degree including a slight push here or there to set that final defense
date. Dr. Prochazka has been there for me in sharing his experience and offering wise
advice, even when it meant that I was not as available to help him with his research.
What a great boss! Clearly I could not have completed this thesis without the help of
Thank you all.
Early Detection and Screening...............................2
2. BACKGROUND AND LITERATURE REVIEW.............................6
Breast Cancer Overview......................................6
Breast Cancer Risk Factors..................................9
Attitudes and Beliefs......................................37
Overview of Qualitative Research...........................44
What We Know About Screening Mammography Behavior..........48
3. THEORETICAL MODELS..........................................57
Health Belief Model........................................58
Transactional Model of Stress and Coping...................64
Review of HBM and TMSC.....................................70
General Critique of Theoretical Models.....................74
4. PURPOSE AND HYPOTHESES.......................................77
Rationale for Hypotheses 1 and 2...........................78
Rationale for Hypothesis 3.................................78
Rationale for Hypothesis 4.................................79
Rationale for Hypothesis 5.................................79
Rationale for Hypotheses 6-10..............................80
Overview of Primary Analysis by Marcus et al. (in press)...83
FDR Telephone Interviews...................................86
Departures from the Marcus et al. Study (in press)...........90
Overview of Analyses.........................................94
Analysis of Variance.........................................96
Mediator Effects with Linear and Logistic Regression.........96
Data Analysis and Hypotheses................................100
Power and Sample Size.......................................104
One-Way Analysis of Variance................................118
Mediator Effects with Linear and Logistic Regression........121
Two-Way Analysis of Variance................................128
Summary of Results..........................................144
Overview of Theoretical Findings............................151
Theoretical Perspective in Predicting Screening Mammograms.164
Implications and Application..............................165
Limitations of Study......................................168
A. BASELINE CORE INTERVIEW................................177
2.1 Current Guidelines for Screening Mammography.......................20
3.1 Health Belief Model................................................62
3.1 Health Belief Model (Example)......................................63
3.2 Diagram of Transactional Model of Stress and Coping................67
3.3 Diagram of Transactional Model of Stress and Coping (Example)......68
5.1 Diagram of Mediation Model.........................................97
6.1 Results Demonstrating Perceived Risk as a Mediator Variable.......126
5.1 Study Variables and Sample Items from Baseline Core Interview and
Impact of Event Scale......................................................90
5.2 Demographic Characteristics of Participants (N=l,082).....................93
6.1 Factor Loadings for Five-Factor Solution from Principal-Components
Analysis with Varimax Rotation............................................110
6.2 Eigenvalues, Percentages of Variance, and Cumulative Percentages
6.3 Cronbachs Alpha for Four-Subscales......................................113
6.4 Means and Standard Deviations for Intrusive Thoughts, Avoidance Behavior,
Perceived Risk, Attitudes and Beliefs, and Perceived Control.............115
6.5 Intercorrelations for Intrusive Thoughts, Avoidance Behavior, Perceived Risk,
Attitudes and Beliefs, and Perceived Control N=853)...................... 117
6.6 Means, Standard Deviations, and One-Way Analyses of Variance
(ANOVAs) for SFDRs and MFDRs on Four Dependent Variables.................119
6.7 Summary of Five Separate Regression Analyses with Number of
FDRs Predicting Intrusive Thoughts, Avoidance Behavior, Perceived Risk,
Perceived Control, and Attitudes and Beliefs..........................123
6.8 Summary of Two Separate Logistic Regression Analyses with Number
of FDRs and Mediator Variables Predicting Whether Participants Had a
Mammogram in the Last Year............................................125
6.9 Two-Way Analyses of Variance for Intrusive Thoughts, Avoidance
Behavior, and Perceived Risk as a Function of Number of FDRs and Race....131
6.10 Two-Way Analyses of Variance for Intrusive Thoughts, Avoidance
Behavior, and Perceived Risk as a Function of Number of FDRs
6.11 Means and Standard Deviations for Perceived Risk as a Function of
6.12 Two-Way Analyses of Variance for Intrusive Thoughts, Avoidance
Behavior, and Perceived Risk as a Function of Number of FDRs
6.13 Means and Standard Deviations for Intrusive Thoughts as a Function
6.14 Means and Standard Deviations for Avoidance Behavior as a Function
6.15 Two-Way Analyses of Variance for Intrusive Thoughts, Avoidance
Behavior, and Perceived Risk as a Function of Number of FDRs
6.16 Means and Standard Deviations for Avoidance Behavior as a Function
6.17 Two-Way Analyses of Variance for Intrusive Thoughts, Avoidance
Behavior, and Perceived Risk as a Function of Number of FDRs and
Relationship to Index Case.........................................143
6.18 Summary of Findings for Two-Way ANOVAs.................................146
Breast cancer is the second leading cause of death in women in the United
States. It is estimated there will be 212,118 new cases of breast cancer and 41,250
deaths from breast cancer in 2005 (ACS, 2004). The strongest risk factors for breast
cancer are sex, age, and family history (ACS, 2004). Breast cancer is primarily a
disease found in women with only 1% of breast cancers diagnosed in men. Breast
cancer is very rare in women younger than 20 and even 30 years of age, however the
incidence drastically increases in women as they approach 50 (Kuller, 2001; Willett,
Rockhill, Hankinson, & Hunter, 2000).
It is well documented that a family history of breast cancer is a risk factor;
however, there are differences between a familial and a hereditary risk of breast
cancer (Collaborative Group on Hormonal Factors in Breast Cancer, 2001; Russo et
al., 2002; Thalib et al., 2004). Although many women perceive risk based on the
occurrence of breast cancer on the maternal side, this increased risk is also related to
paternal lineage (Andrykowski, Munn & Studts, 1996).
Familial breast cancer comprises about 15-20% of breast cancer in families
where there is a clustering of two or more affected blood relatives. A tendency for
familial breast cancer is usually exhibited by the presence of only two or three cases
of breast cancer in a family, older ages of onset, and absence of genetic features, such
as ovarian cancer in a family member. In contrast, hereditary or genetic breast cancer
refers to breast cancer that can be linked to genetic mutations. Genetic cases of breast
cancer account for only 5-10% of all breast cancers (DeMichele & Weber, 2000;
Narod, 2001; Rubenstein, 2001).
Early Detection and Screening
Although limited, there are tools for primary risk reduction and early detection
of breast cancer for women with an increased risk of developing breast cancer.
Primary risk reduction interventions include prophylactic bilateral mastectomy,
chemoprevention (such as tamoxifen) and prophylactic oophorectomy (removal of the
ovaries). Secondary detection tools include breast self-exam, clinical breast exam
and screening mammography (Gross, 2000).
Of prime importance is the use of screening mammography as a tool for early
detection of breast cancer. Several studies have demonstrated that screening
mammography can decrease mortality from breast cancer even though there has been
controversy regarding the value of mammography in decreasing mortality (Lee, 2002;
USPSTF, 2002). Although the greatest benefit has been seen in women older than
50, decreased mortality has also been shown in women 40-50 years of age
(Brekelmans et al., 2001; deKoning, 2000; Humphrey, Helfand, Chan, & Woolf,
Despite the potential value of mammography in detecting early stage breast
cancer, only approximately 67% of eligible women get a mammogram in accordance
with the age appropriate recommendations. However, for poorer women with a lower
income this number drops to approximately 50%. The National Cancer Institute has
set a goal of 80% of women in the general population getting a mammogram within
the applicable age guidelines by the year 2010 (NCI-PDQ Screening, 2003).
Mammography rates for first-degree relatives (FDRs) of women with breast
cancer tend to range between 65-75%. However, some studies have found that as
few as 54% of younger women (< 35 years of age) with at least one first-degree
relative (FDR) with breast cancer have obtained a baseline mammogram (Bastani,
Maxwell, Bradford, Das, & Yan, 1999; Schwartz, Taylor, Willard, & Sigel, 1999).
Women with a family history of breast cancer may want to consider starting
mammograms at an earlier age, but should discuss this and other options with their
provider (ACS, 2004).
As with most behaviors, there are factors that influence adherence to
recommended screening behaviors. The health psychology literature asserts that the
willingness of a person to participate in a screening behavior is related to her
perception of risk and the associated psychosocial distress (Neise, Rauchfuss, Paepke,
Beier, & Lichtenegger, 2001).
In a meta-analysis done by McCaul, Branstetter, Schroeder, and Glasgow
(1996), women with a family history of breast cancer were more likely to be
screened, especially those women who perceived themselves as being vulnerable to
developing breast cancer. In addition, women with a family history of breast cancer
and greater worry were also more likely to be screened. Other research has shown
that more positive beliefs in personal control and beliefs in the efficacy of
mammography predict higher mammography rates (Aro, deKoning, Absetz, &
Schreck, 1999; Audrain-McGovem, Hughes, & Patterson, 2003).
However, several studies have shown mixed results when examining the
relationship of risk perception and screening behavior (including mammography) in
women with a family history of breast cancer. FDRs of women with breast cancer
have been found to either overestimate their risk of getting breast cancer (Neise et al.,
2001) or underestimate their risk resulting in increased adherence or decreased
adherence to screening behaviors respectively (McCaul et al., 1996).
Of interest is the significant difference in women with one affected first-
degree (SFDR single first-degree relative) versus those women with two or more
affected first-degree relatives (MFDRs multiple first-degree relatives) with breast
cancer in terms of having had a mammogram in the last twelve months. In a primary
intervention study done by Marcus et al. (in press), 70.9 % of women with a SFDR
and 79.7% of women with MFDRs reported having a mammogram in the last twelve
months (p value = 0.023). Although one might question the accuracy of self-report of
screening mammography, there is evidence that self-report can be quite accurate
(Aiken, Fenaughty, West, Johnson, & Luckett, 1995).
The primary purpose of this study was to reanalyze the Marcus et al. (in press)
data to examine and compare the role of psychosocial distress, perceived risk,
perceived control, and attitudes and beliefs as mediators in screening mammography
behavior in women with a SFDR versus MFDRs with breast cancer. Differences in
intrusive thoughts, avoidance behavior, perceived risk, and perceived control in
women with a SFDR versus women with MFDRs were examined. In addition, the
relationship of demographic variables, number of FDRs and psychosocial distress and
perceived risk were also explored.
Although not a primary purpose, the Health Belief Model (HBM) and the
Transactional Model of Stress and Coping (TMSC) were evaluated in terms of what
one would expect each theory to predict. An ongoing evaluation of theories is an
important part of the theoretical process in that theories and models help to explain
behavior and provide guidance in developing interventions to encourage behavior
change (Glanz, Lewis, & Rimer, 1997).
BACKGROUND AND LITERATURE REVIEW
Breast Cancer Overview
The incidence of breast cancer has been rising in the United States since the
1930s. Although some of the increased incidence can be accounted for by the use of
screening mammography that became available in the 1980s, this phenomenon does
not explain the entire increase (Willett et al., 2000).
A lifetime risk of breast cancer is often reported as 1 in 8 or 12.5%. This ratio
can be misleading since this number refers to the cumulative risk of developing breast
cancer from birth. In actuality, a 40-year-old woman has less than a 1 in 60 chance of
developing breast cancer in the next ten years, while a 70-year-old woman has a 1 in
25 10-year risk (ACS, 2003; Heilman & Harrison, 2000; NCI-PDQ Screening,
There are also differences among ethnic groups, in particular Caucasian and
African American women. According to the Surveillance, Epidemiology, and End
Results (SEER) Program registries, Caucasian women in the United States have a
13.1% lifetime risk of developing breast cancer, whereas African American women
have a 9.6% lifetime risk. Interestingly, breast cancer rates for white women are
much greater than for African American or Hispanic women (ACS, 2004; Ashing-
Giwa, 1999; Clark, 2000; Salazar, 1996).
However, it should be noted that overall mortality from breast cancer is higher
in African American and Hispanic women. In addition, the increased lifetime risk of
breast cancer in Caucasian women is associated with a higher socioeconomic status in
comparison to African American women. The decreased incidence of breast cancer
in African American women may be related to reproductive factors such as an earlier
age at first lull-term pregnancy and earlier age at menopause (ACS, 2003; Ashing-
Giwa, 1999; Clark, 2000; Salazar, 1996; Yood et al., 1999).
Breast cancer begins as a localized malignant mass that can spread to the rest
of the body via the lymphatic system. The size of the mass and number of local and
regional lymph nodes involved determine the prognosis of breast cancer at the time of
diagnosis. Metastasis to distant sites (liver, brain, bone, etc) is most common in
advanced breast cancer, but may occur as late as 10 years or longer after the initial
diagnosis of any stage breast cancer. Finally, the estrogen status of a breast cancer is
also used as a predictive factor in the diagnosis and treatment decisions for breast
cancer in women with tumors that are estrogen receptor positive (ER +) (Clark, 2000;
Elledge & Fuqua, 2000; Harris, 2000).
Women with a tumor that is ER+ tend to respond better to treatment and have
increased disease-free survival in comparison to women with a tumor that is ER
negative (ER -). Some researchers have hypothesized that over expression of
estrogen receptors results in a better response to therapy (ER+) (Clark, 2000; Elledge
& Fuqua, 2000).
Given that estrogen is thought to promote tumor growth, pre-menopausal
women typically have a less favorable outcome compared to women who develop
breast cancer after menopause because they produce more estrogen than post-
menopausal women. Furthermore, older women are more likely to die from causes
other than breast cancer so the absolute risk of dying from breast cancer declines with
increasing age as a result of these competing causes of death (Clark, 2000; Elledge &
Staging breast cancer allows health professionals to group patients according
to the extent of their disease as well as to determine treatment and prognosis. Several
factors are used to determine the stage of breast cancer. The American Joint
Committee on Cancer (AJCC), which is a collaboration of the American Cancer
Society and the American College of Surgeons currently determine the guidelines for
staging. The AJCC system is based on the TNM system (T = tumor, N = nodes, and
M= metastasis) that uses a clinical and pathologic staging system. Stages range from
Stage 0 (pre-cancerous) to stage IV (most advanced with distant metastasis). Both
clinical and pathologic staging are used to stage and determine treatment for breast
cancer (Harris, 2000).
The stages used for breast cancer are as follows: Stage 0: carcinoma in situ
(also referred to as pre-cancerous); Stage I: tumor 2cm or smaller with no lymph node
involvement; Stage II: tumor 2-5cm with/without nodal involvement (Stage ILA may
involve a site with no primary tumor but with nodal involvement); Stage III: no
evidence of primary tumor to a tumor that has invaded the chest wall with some
degree of nodal involvement, but no metastasis; Stage IV: any size tumor and nodal
involvement as well as distant metastasis (Harris, 2000).
Survival rates for breast cancer vary according to extent of disease and
ethnicity. Five-year survival rates are as follows: Localized disease (Stage I): 97%
for Caucasian women and 90% for African American women; Regional Disease
(Stage 2-3): 77% for Caucasian women and 61% for African American women; and
Distant Disease (Stage IV): 21% for Caucasian women and 17% for African
American women. Although there is a decreased incidence of breast cancer in
African American women, African American women tend to present with more
advanced breast cancer and thus the mortality rate is higher in this population. This
may be related to cultural differences and lack of access to healthcare in African
American women (Harris, 2000).
Breast Cancer Risk Factors
Risk factors for breast cancer include older age, positive family history,
environmental factors, lifestyle (i.e. alcohol, diet, activity), hormonal and
reproductive factors, and a previous biopsy (risk increases with number of biopsies
related to suspicious findings and scarring can obscure a potentially malignant change
in breast tissue). Furthermore, women with a previous history of breast cancer and
women with a history of radiation to the thoracic region may be at an increased risk
of developing breast cancer. Family history and increasing age are two of the most
significant risk factors known to increase a womans chances of developing breast
cancer (Gail et al., 1989; Gross, 2000; Isaacs, Peshkin, & Lerman, 2000).
The risk for first-degree relatives (FDRs) of women with breast cancer is
greatest in those families where the affected relative was diagnosed at a young age.
In addition, the risk is greater when there are an increased number of affected
relatives and the biologic relationship is closer (Willett et al., 2000). In a pooled
analysis the relative risk of breast cancer conferred by a FDR relative with breast
cancer was found to be 2.1 (95% confidence interval 2.0 2.2) (Pharoah, Day, &
Duffy, 1997). In addition, Claus, Risch, and Thompson (1990) found a sharp increase
in risk associated with women with two or more FDRs. Assessing a womans risk of
breast cancer based on a family history is difficult. Incorrect information may be
given or family history information may be unknown especially if the relatives are
distant, died early, or the woman was adopted.
BRCA1 and BRCA2 are two gene mutations that have been studied and are
known to increase a womans chances of developing breast cancer. Although genetic
testing has been refined over the past several years, only 5-10% of breast cancer cases
can be attributed to a genetic mutation. Consequently, this does not guarantee an
accurate assessment of the true risk of breast cancer. In addition, the presence of both
breast and ovarian cancer in a family may indicate the presence of a genetic mutation
(Narod, 2001; Rubenstein, 2001).
A carrier of BRCA1 will not necessarily develop breast or ovarian cancer.
However, carriers of the BRCA1 and BRCA2 genes have approximately a 56-85%
risk of developing breast cancer and/or 10-60% chance of developing ovarian cancer
in their lifetimes (Audrain-McGovem et al., 1995). This risk of breast cancer exceeds
the 12-13% lifetime risk seen in the general population (ACS, 2004).
A geneticist will generally follow two steps if a BRCA1 and/or a BRCA2
gene mutation is suspected. First, a detailed pedigree of the womans family is
compiled to determine if she meets the criteria for a genetic contribution. The criteria
include age at which the cancer occurred in the family member (the younger the age
the cancer occurs, the more likely it is to have a genetic tendency), the number of
affected relatives, the gender of the affected person (men are more likely to be a
carrier of the BRCA2 gene), the bilateral occurrence of breast cancer and the presence
of multiple cancers in one person.
The second step is to assess the personal cancer risk of the person initiating
the consultation (Huiart et al., 2002). A family will then be rated as having a high,
moderate, or low breast cancer risk. The criteria used most often to make these
judgments are based on descriptive epidemiological data that have been transformed
into models (Eccles, Evans, & Mackay, 2000).
Genetic mutations can be related to an autosomal dominant pattern of
inheritance, an inherited inactivation of a tumor suppressor gene, or result from a
hereditary breast cancer syndrome (DeMichele & Weber, 2001). Breast cancer
related to a genetic mutation will often first appear in younger, pre-menopausal
women (median age =42), develop as a secondary breast cancer in the contralateral or
opposite breast (estimated 5-year rate of 30%), or present as ovarian cancer in another
family member (Narod, 2001; Rubenstein, 2001).
The percentage of women interested in genetic testing ranges between 43-
89%. Most of the women interested in this option are younger women with a positive
family history and women older than 50 years of age with breast cancer
(Andrykowski et al., 1996). Genetic testing is available at academic medical centers,
research groups, and commercial laboratories. However, it is important to remember
that more than 50% of breast cancer cases occur in women with no known risk factor
(Humphrey et al., 2002).
Thus, genetic testing is not appropriate for the majority of women at an
increased risk of developing breast cancer; therefore other methods of determining
risk may be helpful. In addition to assessing a womans risk of breast cancer based
on family history, there are other objective methods used to determine a womans
objective risk of breast cancer. The Gail and Claus models are two methods used to
assess an individuals risk of breast cancer in clinical practice and research settings
(Claus et al., 1990; Gail et al., 1989).
The Gail model uses the age of an individual and a set of risk factors to
determine an individuals risk of developing breast cancer in 5 years and during
his/her lifetime. The risk factors used in the Gail model include: age at menarche,
number of children, and/or age at birth of first child, number of previous biopsies,
number of FDRs with breast cancer, and the presence of atypical hyperplasia on
biopsy (Gail et al., 1989).
A software program from the NCI and a calculator from Astra Zeneca
(manufacturer of tamoxifen) are available to clinicians to calculate breast cancer risk
using the Gail model. The Gail model has been criticized because it does not take
into account paternal lineage or number of second degree relatives. However, it is the
model currently in use by the National Cancer Institute (Gail et al., 1989; McTieman
et al., 2001).
Another model used to determine an individuals objective risk of developing
breast cancer is the Claus model developed in the population-based, case-control
study in the Cancer and Steroid Hormone (CASH) study. This model uses the
number of first and second-degree relatives with cancer (maternal and paternal
lineage) and the affected relatives age to calculate a risk assessment in 10-year
increments from age 29-79 (Claus et al., 1990; McTieman et al., 2001).
There are primary risk reduction measures and early detection tools available
that may help detect some breast cancers at an earlier stage. Primary risk reduction
measures include prophylactic mastectomy, prophylactic oophorectomy, and
chemoprevention. Tools for early detection include breast self-exams, clinical breast
exams, and screening mammography.
Prophylactic bilateral mastectomy is one option for primary risk reduction of
breast cancer, whereby a woman at an increased risk of breast cancer, without a
diagnosis of breast cancer, can choose to remove both breasts. Although early data
suggest that bilateral prophylactic mastectomy may significantly reduce an
individuals chance of developing breast cancer, there are significant risks and
psychosocial side effects associated with this option (Isaacs et al., 2000).
In addition, women with a history of cancer in one breast may decide to
remove the contralateral breast or opposite breast as a preventive measure. It is
important to note that this procedure does not completely eliminate the possibility of
breast cancer since it is impossible to remove all breast tissue in any procedure.
There are also data to suggest that prophylactic oophorectomy may reduce the risk of
breast cancer in women with hereditary breast cancer (Isaacs et al., 2000).
Chemoprevention has recently become an option for women with an increased
risk of breast cancer. The Breast Cancer Prevention Trial (BCPT) involved 13,388
women, 75% of whom had at least one relative with breast cancer. This trial was
stopped earlier than planned after finding that women with an increased risk of breast
cancer (1.66 5-year risk using the Gail model) could decrease their risk by 49% by
taking tamoxifen for five years. The benefit was greatest for women less than age 50
since the side effects related to tamoxifen increased in women older than 50 years of
age. These side effects include endometrial cancer, venous thrombosis, pulmonary
embolus, and cataracts. Due to serious side effects, the risk-benefit ratio for
tamoxifen must be assessed for each individual (Gail et al., 1989).
The Study of Tamoxifen and Raloxifene (STAR) trial may introduce another
option for chemoprevention in women at an increased risk of breast cancer. The
STAR trial is currently underway to measure the effectiveness and adverse events of
tamoxifen as compared to Raloxifene to determine the effectiveness and adverse
events between the two drugs. Raloxifene does not appear to have the adverse events
associated with tamoxifen; however, it is not clear if it is as effective as tamoxifen in
decreasing the risk of breast cancer in pre-menopausal women. Because Raloxifene
is only approved for use in post-menopausal women, pre-menopausal women are not
currently eligible to participate in the STAR trial (Dunn & Ford, 2001).
Tools for early detection described in the May 2003 American Cancer Society
(ACS) guidelines include breast self-exams (BSEs), clinical breast exams (CBEs)
performed by health professionals, and screening mammography. In 1997, the
American Cancer Society recommended BSE monthly starting at age 20, for women
at an average risk of breast cancer, however, the effectiveness of this recommendation
has been brought into question, primarily based on a study done by Thomas, Gao, and
Ray (1997). This randomized study done in Shanghai concluded that BSE did not
reduce breast cancer mortality and may have led to an increased number of benign
breast biopsies. However, the data on this issue remain inconclusive and the ACS
currently recommends BSE as an option beginning at age 20. One potential
advantage of this recommendation is a means for a woman to become familiar with
her breasts so that she might notice changes sooner than if she did not perform BSE
Clinical Breast Exams (CBEs) are a second option for secondary detection
tools. The ACS recommends CBEs approximately every three years for women in
their 20s and 30s and annually for women age 40 and older. A CBE should be
performed by a trained health care professional and in some cases may prove to have
additional benefit if done prior to screening mammography so that any abnormal
finding during a CBE can be more closely followed up with mammography.
Undergoing a CBE also allows a woman an opportunity to discuss noted changes as
well as risk factors and methods for early detection of breast cancer with her provider
Ultrasound and MRI of the breasts are also being evaluated as screening tools
for the early detection of breast cancer; however, the data evaluating cost
effectiveness are not available at this time. Women at an increased risk for breast
cancer should discuss these procedures with their primary care provider to determine
the benefit of ultrasound and/or MRI as an early detection measure (ACS, 2004).
Screening mammography is a common screening mechanism used to assess
the presence of an abnormality in a womans breast. Although there has been a great
deal of controversy surrounding the issue of screening mammography and its effect
on breast cancer mortality, new evidence demonstrates that mammography can be
effective in detecting breast cancer at age 40 (versus age 50 that was previously being
promoted). As a result, ACS guidelines continue to recommend screening
mammography annually for women starting at age 40 (ACS, 2004).
One of the advantages of screening mammography is that it can identify
ductal carcinoma in situ (DCIS) that is considered a pre-cancerous condition and
breast cancers too small to be detected by palpation. Screening mammography uses
ionizing radiation to image breast tissue. In this procedure, the breast is compressed
between a plastic plate and an x-ray cassette and then films are taken using two
different positions. The use of two views decreases the rate of false positive findings.
To ensure the quality among all mammography centers, the Mammography Quality
Standards Act was enacted by Congress in 1992 requiring all facilities performing
mammography to be certified by the FDA (Smith & DOrsi, 2000).
The sensitivity of mammography (ability of mammography to detect breast
cancer when it is present) is dependent on several factors and is of great importance
since the failure to diagnose breast cancer is a common cause of litigation. The
sensitivity of mammography has been shown to range from 54-58% in women under
the age of 40, to 81-94% in women over the age of 65. It should be noted that the
sensitivity of mammography is lowest for women aged 30-49 mostly related to the
increased breast density that makes identification of abnormalities difficult (NCI-
PDQ: Screening, 2004; Smith & DOrsi, 2000). Because the sensitivity of
mammography is highly dependent on the skill of the radiologist in interpreting the
study, more radiologists are specializing in reading and interpreting mammograms
(Smith & DOrsi, 2000).
The Journal of the National Cancer Institute conducted an observational study
using information collected by seven mammography registries across the United
States to examine differences in the discovery of late-stage disease (positive lymph
nodes or metastases) in mammograms performed yearly versus every two years.
Women aged 40-49 years were 1.35 times more likely to have late-stage disease with
a two-year screening interval than those with a one-year screening interval. There
were no differences in women aged 50 and older (White et al., 2004).
In terms of mortality, a 17% reduction in mortality has been shown for women
who begin mammography screening at age 40. There is also evidence suggesting that
regular mammography screening in women aged 50-59 years may result in a 25-30%
reduction in breast cancer mortality. Other studies have found that annual screening
mammography can decrease breast cancer mortality 23% in women between forty
and fifty years of age and 45% in women over fifty (Timins, 2005).
As mentioned earlier, mammography as a screening tool has been criticized,
primarily in terms of its value in decreasing mortality. A meta-analysis done in 1997
as part of the Cochrane Collaboration Review criticized the use of mammography as
a screening tool. There were two major issues put forth by Olsen and Gotzche. First
of all, based on the randomized clinical trials available at that time, the authors
concluded that there was no evidence that breast cancer screening reduced mortality.
Secondly, these authors believed that screening procedures led to more aggressive
and sometimes unnecessary treatments including mastectomies, especially hue in
women 40-49 years old (ACS, 2003). Despite the credibility of Cochrane reviews,
the data presented by these two authors has faced extreme scrutiny and been
challenged by other breast cancer experts, to the point that most experts continue to
recommend screening mammography as an effective early detection tool in spite of
the findings by Olsen and Gotzche (ACS, 2003; Fletcher & Elmore, 2003).
The U.S. Preventive Services Task Force (USPSTF) consists of a group of
health experts who review published research and then make recommendations about
preventive health care. The USPSTF found evidence that women 40-69 years of age,
who undergo screening mammography may die less frequently than women who do
not get screened. They believe that the findings of Olsen and Gotzsches use of all-
cause mortality obscures the important finding of a 31% reduction in mortality in
breast cancer as a result of screening mammography. However, the USPSTF
cautions that there is a risk of false positive findings that can result in psychosocial
distress and unnecessary procedures such as breast biopsies. Therefore, the risks and
benefits of screening mammography must be evaluated by women on an individual
basis (USPSTF, 2002). The current guidelines for screening mammography are as
Figure 2.1 Current Guidelines for Screening Mammography
Organization Age Screening Should Begin Interval
American Cancer Society 40 Yearly
National Cancer Institute 40 Every 1-2 Years
United States Preventive Services 40 Every 1-2 Years
(Lee, 2002; USPSTF, 2002)
One of the problems with screening mammography is that best practice for
mammography screening remains unestablished. Although screening mammography
is not perfect, it cannot be dismissed since it remains one of the best tools currently
available in the early detection of breast cancer (USPSTF, 2002).
It is important to remember that the recommendations for early detection of
breast cancer provided above are intended for women with an average risk of breast
cancer. Women with an increased risk of breast cancer should be advised to discuss
the possibility of initiating any or a combination of these screening tools sooner or
with increased frequency with their primary care provider. Because none of these
early detection tools are 100% accurate in detecting breast cancer, women at an
increased risk of breast cancer may need to ask additional questions as they make
decisions regarding screening behavior (ACS, 2003; Smith & DOrsi, 2000).
Some of these women may also choose to be more diligent in performing
breast self-exams, having a CBE performed on a regular basis by a healthcare
professional and/or having an MRI or ultrasound done as an additional screening tool.
For example, ultrasound and/or MRI of the breast, while not recommended for use as
a general screening tool, may have benefit in conjunction with other screening
methods for women at high risk for breast cancer. In addition, a MRI may be done
following a positive mammogram in women at an increased risk of developing breast
cancer (ACS, 2003).
Women at an increased risk of developing breast cancer may want to pursue
more intensive surveillance such as getting mammograms at an earlier age than is
recommended for women in the general population. Some experts recommend that
women with a family history of breast cancer begin getting mammograms 5-10 years
prior to the age that their youngest relative was diagnosed with breast cancer.
Women known to have a BRCA 1 or BRCA 2 mutation may want to consider
initiating screening mammography beginning at age 25 or younger (Smith et al.,
Finally, high-risk women have the option of risk reduction measures such as
chemoprevention or prophylactic bilateral mastectomy as described previously.
Therefore, it is important that women at an increased risk of breast cancer discuss
these options with their primary care provider as early as possible (ACS, 2004).
However, there are other important issues associated with being at an increased risk
of developing breast cancer including psychosocial distress, perceived risk, perceived
control, and attitudes and beliefs as well as the relationship of these factors with
Thus, assessment of levels of psychosocial distress, perceived risk, perceived
control, and attitudes and beliefs towards breast cancer are critical pieces that should
be evaluated in women with an increased risk of developing breast cancer. Likewise,
these factors should be considered when evaluating screening practices for the early
detection of breast cancer, including breast self-exam, physical exam by a primary
care provider, and mammography as well as available primary risk reduction
Marshall (1996) has described three levels of adverse effects related to
preventive screening programs. These three levels include the screening procedure
itself, the investigation of abnormal findings, and the treatment of abnormal findings
including false-positive results. In addition to experiencing psychosocial and other
forms of distress related to screening, a womans personal attitudes and beliefs in
terms of the effectiveness of a preventive behavior will also influence her adherence
to prevention behavior. These factors will be discussed in some detail below.
One must keep in mind that although risk reduction methods can be very
useful, they can also cause physical, psychosocial, social and/or ethical harm to those
choosing to participate in early detection methods, especially in women with a family
history of breast cancer. For example, although some studies have shown a decrease
in psychosocial distress following a prophylactic mastectomy (Frost et al., 2000),
other studies have shown significant distress in high-risk women choosing to undergo
a prophylactic mastectomy. In addition, some studies have shown that women may
regret choosing this option (Lodder et al., 2002).
Many studies have found that the presence of high levels of perceived risk and
psychosocial distress such as intrusive thoughts, avoidance behavior, or anxiety in
women with a family history of breast cancer may be associated with decreased
adherence to regular mammography screening (Aro, Absetz, van der Ploeg, & van der
Kamp, 2000; Kash, Holland, Osborne, & Miller, 1995; Valdimarsdottir et al., 1995;
Zakowski et al., 1997).
Other studies have found that moderate distress or anxiety and/or increased
perceived risk may increase a womans adherence in getting screening mammograms
(Diefenbach, Miller, & Daly, 1999; McCaul et al., 1996). However, others have not
found a significant relationship between a womans level of perceived risk and
psychosocial distress and mammography screening behavior (Bowen, Alfano,
McGregor, & Anderson, 2004).
Psychosocial distress in women at an increased risk of breast cancer is an
important area that should be considered in terms of its impact in the lives of these
women as well as the association with screening behavior (NCI-PDQ-Psychosocial,
2003). Psychosocial distress can include intrusive thoughts, avoidance behavior,
worry, anxiety, and depression. Psychosocial distress factors may not always be
detrimental to a womans life or to her decision to adhere to breast cancer screening
guidelines, however, these factors can result in increased levels of distress and
decreased adherence to screening procedures for breast cancer including
mammography (McCaul et al., 1996).
For example, avoidance behavior is a form of psychosocial distress that can
result in positive or negative behaviors. If a woman believes that eating a healthy diet
and exercising may decrease her risk of developing breast cancer, using these acts as
a strategy to get away from the sources of the stress can be healthy. However,
escape-avoidance as described by Folkman and Lazarus (2005) that entails wishful
thinking, where the woman may wish that if she believes or wishes hard enough,
the problem will go away or if she chooses to avoid getting mammograms so she will
not have to worry about the results, may result in more depression and anxiety and
decreased adherence to screening guidelines.
In addition, avoidance behavior as a coping mechanism can affect vigilance
that may increase or decrease distress. Increased vigilance can increase distress in
situations where nothing can be done such as waiting for the results of a biopsy.
However, increased vigilance may also decrease distress and increase perceived
control if a woman has an increased understanding of her risk and her options such as
risk reduction methods and tools for early detection of breast cancer (Folkman &
A greater level of worry has been associated with a positive family history of
breast cancer. Specifically, FDRs may have increased levels of nonspecific distress
in addition to avoidance behavior and intrusive thoughts related to breast cancer in
comparison to women without a family history of breast cancer (McCaul, Branstetter,
ODonnell, Jacobson and Quinlan, 1998). In some cases, 27-53% of women at an
increased risk of developing breast cancer have demonstrated clinically significant
increased numbers of intrusive thoughts and more avoidance behavior (Gurevich et
al., 2004). However, in a study by Brain, Norman, Gray, and Mansel (1999) family
history was weakly correlated with psychosocial factors such as anxiety and worry.
Hailey, Carter, and Burnett (2000) found that women with a FDR with breast
cancer experienced more negative attitudes and expressed more anxiety related to
breast cancer than women without a family history. In another study, women at an
increased risk of breast cancer had significantly higher levels of depressive symptoms
and emotional alienation in comparison to a control group. Furthermore, some of
these women had enough psychosocial distress to warrant a referral for counseling
(Kash et al., 1995).
Studies have also shown that the relationship of a woman at an increased risk
of developing breast cancer to the woman diagnosed with breast cancer can influence
the level of distress. For example, Zakowski et al. (1997) showed that women at an
increased risk of developing breast cancer had increased levels of psychosocial
distress if her parents had died of cancer. In addition, daughters of mothers that had
died from breast cancer, who were an adolescent at the time of their mothers death or
experienced changes in their life plans secondary to their mothers breast cancer,
were also shown to have higher levels of distress (Wellisch, Gritz, Schain, Wang, &
The relationship between psychosocial distress and genetic counseling for
breast cancer has also been studied. Lloyd et al. (1996) and Watson et al. (1999)
found a significant difference between women who received genetic counseling and
those who did not in terms of cancer-specific distress. Women with an increased risk
of breast cancer experienced more worry after undergoing genetic counseling. This
finding may be related to the fact that women felt they had few options for prevention
and a sense of less control if they were found to have a genetic risk factor. However,
other researchers have described decreased levels of worry (Bish et al., 2002) and
lowered levels of depression symptoms (Lerman et al., 1996) following genetic
The impact of psychosocial distress on screening behaviors is another
important consideration; however the relationship is far from clear. Several studies
have been done examining the association of psychosocial distress and cancer
screening behavior; however, the results have been inconsistent (Absetz, Aro, &
Sutton, 2003; Diefenbach et al., 1999; Drossaert, Boer, & Seydel, 1996).
Absetz et al. (2003) found that women who had personally experienced a
family or friend with breast cancer reported increased levels of perceived risk
resulting in increased screening behavior. Brain et al. (1999) found increased anxiety
led to hypervigilence that in turn resulted in over-adherence to breast self exams.
However, those women who scored higher on a measure of breast cancer anxiety
were less likely to adhere to screening recommendations.
A positive association between breast cancer worry and screening
mammography behavior has been seen in some studies. For example, McCaul et al.
(1998) found that thinking and worrying about breast cancer was modestly related to
increased adherence to screening behavior.
In contrast, some studies have found that increased levels of distress may
impede screening behavior. Specifically regarding screening mammography, some
researchers have found serious psychosocial distress to be associated with decreased
adherence to screening mammography (Audrain-McGovem et al., 2003; Drossaert et
al., 1996; Schwartz, Taylor, & Willard, 2003). Andersen, Smith, Meischke, Bowen,
and Urban (2003) reported an inverted u-shaped pattern in the association of cancer
worry and mammography. Women with mild or severe levels of cancer worry were
less likely to adhere to annual screening mammography, while women with moderate
levels of worry were most likely to adhere.
Other studies have demonstrated that high levels of cancer worry may
negatively affect a womans choice to participate in genetic counseling programs and
screening behaviors such as monthly self breast exam, clinical breast exams, and
screening mammography (Audrain-McGovem et al., 2003; Bowen et al., 1999;
Diefenbach et al., 1999; Lerman et al., 1996). In addition, some research has shown
that high-risk women with a false-positive finding on a mammogram followed by a
breast biopsy have increased anxiety and breast cancer worries towards future
screening mammograms, even when the finding turns out to be normal (Gilbar et al.,
Although studies have been done gathering information on the number of
affected FDRs, very few studies have analyzed the data controlling for this variable
(Chalmers, Maries, Tataryn, Scott-Findlay, & Serfas, 2003; Julian-Reynier et al.,
2001). In addition, some studies that account for the different number of affected
FDRs often involve a specific group of women pursuing or receiving genetic
counseling (Duric et al., 2003; Huiart et al., 2002).
The literature describing psychosocial distress in women at an increased risk
of developing breast cancer appears to be inconsistent. Therefore, it would appear
that more research is needed to better understand the association of psychosocial
distress and screening mammography behavior for women at an increased risk of
developing breast cancer due to a family history of breast cancer in a first-degree
Understanding the role of risk perception and breast cancer is an important
aspect of understanding the impact of being at an increased risk of developing breast
cancer, particularly in relation to a womans decision to adhere to screening
mammography guidelines. Several studies have demonstrated that women, including
FDRs of women with breast cancer, tend to incorrectly estimate their risk of breast
cancer either by overestimating or underestimating their risk (Lipkus, Iden,
Terrenoire, &Feaganes, 1999).
Some studies have identified an association between increased anxiety and
breast cancer worries in women who overestimated their risk of developing breast
cancer (Brain et al., 1999; Hopwood, 2000; Lipkus et al., 1999; Watson et al., 1999).
In addition, the inaccuracy of these womens perceived risk for themselves, whether
overestimated or underestimated, may be associated with increased cancer related
distress (Vernon, 1999),
Other studies have found no significant relationship between perceived risk
and distress (Audrain et al., 1995; Hopwood, Shenton, Lalloo, Evans, & Howell,
2001; Kash et al., 1995). This maybe a result of lack of sensitive measures of
perceived risk; therefore new methods of characterizing perceived risk may be useful
(Loescher, 2003). Another possible explanation is the role of optimistic bias, where
some women may believe that their risk of developing breast cancer is less than the
risk of other women their age (Clarke, Lovegrove, Williams, & Machperson, 2000;
Weinstein, & Lyon, 1999).
The impact of perceived risk has been studied in women who have
experienced breast cancer in friends or family. For example, women without a family
history of breast cancer were more likely to overestimate their risk of developing
breast cancer if they had a friend who had experienced breast cancer (Drossaert et al.,
1996; Montgomery, Erblich, DiLorenzo, & Bovbjerb, 2003). In addition, Hopwood
(2000) showed that women who had lost a mother to breast cancer during their
adolescence were more likely to overestimate their risk of developing breast cancer
and they had higher levels of cancer worry. Finally, Marcus et al. (in press) found
that younger FDRs who were sisters and daughters of women with breast cancer, with
a higher income and more education, were most likely to correctly assess their
increased risk of breast cancer.
A few studies have examined risk perceptions in minorities. For example,
African American women are more likely to underestimate their risk of breast cancer
than Caucasian women even when a family history exists. Specifically, 40% of FDRs
with a lower income and of African American ethnicity incorrectly perceived their
risk of breast cancer to be lower than women without a family history (Hughes,
Lerman, & Lustbader, 1996; Lipkus et al., 1999).
Perception of risk as related to genetic testing may also influence a womans
decision to pursue testing. For example, women from families with a history of
breast cancer may overestimate their risk of developing breast cancer and thus be
more likely to pursue genetic testing (Hopwood, 2000). Although genetic risk
counseling has been shown to increase a womans accuracy regarding her perception
of risk for breast cancer, approximately 33% of women continue to have inaccurate
perceptions (either higher or lower) of their risk of breast cancer following genetic
counseling (Lipkus et al., 1999). It is unclear if this is a result of a poor
communication process, difficulty on the part of the woman to comprehend, or
remember the information and/or the effect of a womans personal beliefs regarding
susceptibility (Hopwood, 2000).
In fact, while underestimating ones personal risk of developing breast cancer
may result in less precautionary behavior such as screening mammography
(Weinstein & Lyon, 1999), some studies have shown that a greater perception of
personal risk tends to increase mammography behavior. For instance, some studies
have shown that women, regardless of a family history of breast cancer, are more
likely to be screened for breast cancer if there is a sense of personal vulnerability
(Hopwood, 2000; McCaul et al., 1996).
In a meta-analysis done by McCaul et al. (1996) on the relationship of risk for
breast cancer and mammography screening, 35 of 38 studies showed a positive
relationship between perception of risk and increased adherence to screening
behavior; However, the explanation for this association could not be determined from
this study. This is partly related to flaws of a meta-analysis in that a researcher
usually collects quantitative information using key words from several studies; thus
most meta-analyses cannot provide the explanations for the results. In this case, the
meta-analysis done by McCaul et al. (1996) cannot explain why risk was associated
with increased mammography behaviors, but rather only that this relationship was
present in some of the 200 studies reviewed.
Other studies have shown opposite results. Bastani et al. (1999) and Neise et
al. (2001) found no association between having a family history of breast cancer,
increased level of perceived risk of developing breast cancer, and increased adherence
to screening mammography. In addition, Neise et al. (2001) found that women with a
family history of breast cancer were less likely to get a mammogram if they perceived
themselves to be at an increased risk of developing breast cancer. Finally,
Calvocoressi et al. (2004) reported women being less likely to get a mammogram if
they perceived themselves to have an increased risk of developing breast cancer. This
remained true even when controlling for race, age, and family history of breast
cancer. Of concern is that small groups of women who greatly overestimate their risk
of developing breast cancer (by as much as 50%) may consider or undergo a
prophylactic mastectomy, even though this may an inappropriate option for these
women based on their true risk (Metcalfe & Narod, 2002).
Some of the reasons for these inconsistencies can be attributed to smaller
sample sizes, the means by which the sample is recruited (i.e. studies that recruit
women with an increased risk of breast cancer versus a study that recruits women
with and without an increased risk of breast cancer), the time period in which the
variables were assessed and the design of the study (cross sectional versus
prospective) (Hopwood, 2000; McCaul et al., 1996; Vernon, 1999).
The method of measuring perceived risk must also be considered. Perceived
risk is usually assessed either by Likert style questions (1-5) or asking women to use
a number or percentage between 1 and 100 to describe their perception of their risk in
developing breast cancer (numeracy). Numeracy is described as a persons comfort
level with the use and interpretation of numbers that may influence their
understanding and expression of their risk of developing breast cancer (Frost et al.,
2000; Katapodi, Lee, Facione, & Dodd, 2004; Lipkus et al., 2000). Thus, numeracy
may play a role in womens understanding of their risk.
A womans perception of her risk may also be influenced by the way a
question measuring risk is asked; that is on a scale of 0-100 or 0-100% versus Likert
style questions where a woman compares her risk to that of women with and without
a family history (Davids, Schapira, McAuliffe, & Nattinger, 2004; Woloshin,
Schwartz, Black, & Welch, 1999). Black, Nease, and Tosteson (1995), conducted a
study to determine how women aged 40-50 years of age perceived their risk of breast
cancer and the effectiveness of screening. In this study, numerate women
overestimated their risk of dying from breast cancer less than did innumerate women.
Katapodi, et al. (2004) performed a meta-analysis and determined that the use
of a numerical scale (0% 100%) resulted in women overestimating their risk, while
Likert-type scales more often resulted in optimistic bias. Davids et al. (2004) found
that both numerate and innumerate women age 40-50 overestimated their risk of
developing breast cancer, however, less numerate women overestimated their risk to a
greater degree than more numerate women. Moreover, the method and information
communicated concerning risk used to describe a womans risk of developing breast
cancer (such as a 1 in 8 lifetime risk of developing breast cancer ) varies greatly
and very likely plays an important role in a womans self perception of risk
(Hopwood, 2000). The inconsistency of these data warrants further study regarding
the impact of risk perception in women with an increased risk of developing breast
cancer and the choice to adhere to mammography screening given the potential value
of mammography as an early detection tool.
Locus of control most often includes powerful others (such as the influence of
primary care providers on health outcomes), internal or perceived control where a
person believes they have control and chance control, where a person believes that
health outcomes are a result of chance or fate (Tittle, Chiarelli, McGough, McGee, &
The role of perceived control has been examined in studies of women with
breast cancer. Osowiecki and Compas (1999) conducted a longitudinal study
examining individual coping and adjustment in women with breast cancer. They
found that having a sense of increased perceived control with the use of problem-
focused coping such as active problem-solving, was associated with lower levels of
perceived distress at the time of diagnosis. However, this did not result in the
continued decrease of distress over the course of dealing with breast cancer. There
was no association with perceived control and emotion-focused coping, such as
avoidance of emotions, or problem-focused disengagement coping, such as avoidance
behavior and wishful thinking.
The degree of perceived control may also vary by the types of early detection
methods used by women at an increased risk of breast cancer. For example, women
can perform self-breast self-exams (BSEs) at any time, which may increase a
womans sense of perceived control. Barron, Houfek, & Foxall (1997) found that
women who perceived themselves as having more control in detecting breast cancer
at an early stage were most likely to perform BSEs. Some studies have shown that
women with increased distress related to their high risk for breast cancer may perform
excessive BSEs. However, excessive performance of BSEs can lead to an increase in
false-positive findings that can also be problematic in terms of women undergoing
unnecessary breast biopsies as well as experiencing an increased amount of
psychosocial distress (Brain et al., 1999).
Aro et al. (1999) found that women who perceived themselves to have more
control over detecting breast cancer were more likely to undergo a screening
mammogram. However, a number of structural and internal barriers can influence
women getting mammograms such as location of the mammography center,
transportation, cost, need for referral, and interpretation by a radiologist, thus
potentially resulting in women feeling less perceived control and therefore less likely
to get recommended mammograms (Holm, Frank, & Curtain, 1999).
In a study done by Lloyd et al. (1996) several women acknowledged their
increased risk of developing breast cancer but were not clear on their options to
manage this increased risk. In addition, women with less education did not believe
there was much they could do to prevent breast cancer or decrease their risk of dying
from breast cancer (Absetz et al., 2003). Finally, older women tended to be more
externally motivated in terms of following the recommendation of a health care
provider to obtain a mammogram while also believing that developing breast cancer
is related to chance (Tittle et al., 2002).
It is important to note that one of the problems in understanding perceived
control is that only one or two questions measuring perceived control are commonly
used in studies, thus there is a potential gap in having a validated and standardized
tool to measure perceived control (Krause & Shaw, 2000; Osowiecki & Compas,
Attitudes and Beliefs
In general, a persons attitudes and beliefs influence health behaviors. By
definition, attitudes cannot exist without a reference to some object. In other words,
one does not arbitrarily assign like or dislike; the attitude must be directed at someone
or something. In addition, attitudes play a key role in the intake and processing of
external information (Sherman & Fazio, 1975). Ajzen and Fishbeins model of the
Theory of Reasoned Action (1980) demonstrated the importance of situational factors
and the effect on behavioral intentions that lead to a specific behavior. Leventhal,
Kelly and Leventhal, (1999) believe that a persons motivation for participating in a
certain health promoting behavior is related to the significance or meanings that
underlie a risk appraisal rather than the persons assessment of the risk itself.
Specifically regarding attitudes and beliefs towards breast cancer screening,
social and cultural factors often influence attitudes and beliefs (Vernon, 1999). For
example, the influence of social support systems may need to be considered in
minority women with an increased risk of developing breast cancer since family
relationships may be an important aspect for some families in this population. Glanz,
Grove, Lerman, Gotay, and Marchand (1999) reported family social support to be a
significant factor in Caucasian, Japanese, and Hawaiian women. In addition, African
American and Hispanic women are thought to strongly rely on family support and
fewer sources of external support (Donovan & Tucker, 2000; Hughes et al., 1996).
Other attitudes and beliefs related to breast cancer may also vary among
different ethnic groups. For example, Hispanic women tend to experience more
embarrassment towards the screening procedure (in terms of exposing their breasts)
and have more fatalistic attitudes towards breast cancer in believing that cancer is
Gods punishment for some type of past immoral behavior (Austin, McNally, &
Stewart, 2002). African American women have also been shown to have a more
fatalistic attitude towards prevention and treatment of breast cancer in comparison to
Caucasian women (Hughes et al., 1996).
The relationship of women at an increased risk of developing breast cancer
toward the relative who has been diagnosed with breast cancer can also influence a
womans attitudes and beliefs towards breast cancer. Rutton & Iannotti (2003) found
that perceived benefits and barriers of mammography were similar in women with a
relative with breast cancer and those who did not have a relative with breast cancer.
However, women with a relative with breast cancer were more likely to perceive
themselves as susceptible and thus undergo mammography screening.
A womans experience with friends or family with breast cancer and her
personal life experience with breast cancer can also influence her health beliefs (Tittle
et al., 2002). For example, women with close family or friends with breast cancer
reported an increased level of perceived risk in developing breast cancer themselves.
However, it has been suggested that older women might also be less fearful of
developing breast cancer if they had experienced friends and family with breast
cancer having been cured or at least still living with the disease (Tittle et al., 2002).
In a study done by Hailey et al. (2000) women with FDRs with breast cancer
were compared to women without a FDR with cancer. The women with a FDR with
breast cancer had more negative attitudes about breast cancer including more anxiety.
In addition, they believed they had a greater risk of developing breast cancer,
although they underestimated their actual risk. Unfortunately, this study did not
provide information on the status of the FDR. In spite of their negative attitude these
women were more likely to undergo screening for breast cancer.
Marcus et al. (in press) found that positive attitudes and beliefs towards breast
cancer screening and fewer perceived barriers were highly predictive of current and
future intentions to obtain a mammogram. In contrast, barriers to screening and
cognitive deficits related to incorrect beliefs about breast cancer may also negatively
affect adherence to screening behavior (Kash et al., 1995; Lerman et al., 1994).
Womens attitudes and beliefs about health care providers are another
important factor associated with a womans decision to participate in preventive
health screening. Physician recommendation has consistently been shown to be an
important factor for women (especially older women) in choosing to participate in
screening, particularly mammography. For example, Simoes et al. (1999) found that
women are more likely to have had a recent mammogram if they had seen a physician
in the last year. This may be related to the level of trust many women have in their
primary care providers or a sample of health compliant women.
An understanding of attitudes and beliefs of women towards breast cancer and
the screening methods available is critical to designing effective interventions to
increase screening behavior, particularly mammography. This becomes even more
important when directing effective interventions to women at an increased risk of
developing breast cancer.
A small number of studies have been done examining the relationship of
sociodemographic factors and perceived risk associated with obtaining or not
obtaining a mammogram. Champion and Skinner (2003) found that a womans
perceived risk decreased as her age and education increased. Droseaart et al. (1996)
reported age as the only sociodemographic factor correlated with risk perception, in
that women reported lower perceived risk as they aged. In terms of age, both younger
and older women tended to overestimate their risk of dying from breast cancer
(Bottorff et al., 2002).
A relationship between level of education and perception of risk is unclear.
Women with less education have been found to incorrectly overestimate their risk,
especially when a numerical risk is given, rather than the use of the concept of low,
medium, or high risk (Woloshin et al., 1999). Other studies have not found a
relationship between education and estimation of risk (Erblich, Bovbjerg, Norman,
Valdimarsdottir, & Montgomery, 2000; Katapodi et al., 2004).
The association of sociodemographic factors and mammography behavior has
also been studied. Rahman, Dignan, and Shelton (2003) reported greater adherence
to mammograms in older women, while younger women with a family history of
breast cancer were less likely to be adherent in getting mammograms than younger
women without a family history of breast cancer. Calvocoressi et al. (2004) reported
that women ages 40-49 were less likely to adhere to screening mammography
guidelines if they perceived their risk of developing breast cancer as low. However,
women in this age group who perceived themselves to have a moderate risk of
developing breast cancer were more likely to get mammograms.
Other researchers have also found predictors associated with undergoing
screening mammography include being Caucasian, younger age, having a higher
income, more education, having health insurance, and attitudes and beliefs that
mammography is beneficial as a screening tool (Clemow, Costanza, & Haddad, 2000;
Juarbe et al., 2005). Coleman and OSullivan (2001) also found that physicians were
more likely to recommend mammograms if women were white, married, had more
than a high school education and an income greater than $20,000. Moreover,
minority women may be less likely to get mammograms even when socioeconomic
status and education are controlled for or do not differ (Magai, Consedine, Conway,
Neugut, & Culver, 2004).
Bottorff et al. (2002) examined the influence of sociodemographic factors and
knowledge in relation to genetic testing. After controlling for age, education, breast
cancer status, and knowledge of genetic testing, women with at least one or two FDRs
with breast cancer were most likely to be interested in genetic testing.
Studies examining the impact of psychosocial distress, perceived risk,
perceived control, and attitudes and beliefs in women with a family history of breast
cancer have shown mixed results. Some women with a family history of breast
cancer experience greater psychosocial distress such as intrusive thoughts, avoidance
behavior, worry, or anxiety while other women at an increased risk of developing
breast cancer do not. In addition, some women overestimate their risk of developing
breast cancer, while others underestimate their risk, both of which can result in an
inaccurate perception of their true risk that can lead to increased levels of
psychosocial distress for some women and possibly decreased adherence to
recommended guidelines for screening mammography.
The findings have also been inconsistent when examining the association of
these variables to genetic testing and screening behavior such as BSEs and
mammography. For example, some studies have found that women who manifest
greater levels of psychosocial distress or perceive themselves to have an increased
risk of developing breast cancer are more likely to have screening mammograms,
however, other studies have found that greater psychosocial distress and increased
perceived risk impede screening behavior.
Although sociodemographic factors have been studied in relationship to
screening mammography, few studies could be found examining the relationship of
sociodemographic factors such as age, education, and income in relation to
psychosocial distress and perceived risk. Therefore, this is an area that needs further
study if one is to understand the association of sociodemographic factors and
psychosocial distress and perceived risk in this group of women at an increased risk
of developing breast cancer, especially in relation to screening behaviors for breast
cancer such as mammograms.
Given the inconsistency of the literature looking at women with an increased
risk of developing breast cancer as a group, there remains an additional need to
examine these issues comparing women with a SFDR versus those women with
MFDRs with breast cancer, since there could be important differences that are not
evident when looking at women at an increased risk of developing breast cancer as
one group. Finally, this knowledge could be helpful in developing effective
interventions to increase adherence to screening mammography, especially for
women known to be at an increased risk of developing breast cancer.
Overview of Qualitative Research
Qualitative research is a method of research that allows a researcher to use an
inductive and interactive process of inquiry between the researcher and his/her data
(Morse, 1992, p.2). Qualitative research methods are not designed to provide
quantitative estimates of survey results; rather qualitative methods allow researchers
to gather data that cannot be obtained from quantitative methods. Examples of
qualitative methods include the use of focus groups, structured, or semi-structured
interviews and open-ended questions.
Focus groups are a methodology often used to clarify issues and perceptions
and to probe decision-making processes through the use of group dynamics (Covello
& Peters, 2002). Focus group data are usually collected from homogeneous or
heterogeneous groups of 6 12 participants per group. This allows researchers to
explore knowledge, attitudes, and beliefs among a particular group through the use of
dynamic and interactive exchange. Focus groups are considered a unit of analysis
and can be used as the sole method to gather information or explain quantitative
results or inconsistencies (Crabtree & Miller, 1999). It is important to note that the
small sample sizes used in focus groups may encourage even greater bias in data due
to group dynamics, thus the applicability of results from focus groups should be
treated with care.
Structured and semi-structured interviews are used in other types of
qualitative research. The type of interview to be used is dependent on the specific
project and the information desired. Structured interviews consist of a rigidly
structured interview schedule with open-ended questions that direct the interview. A
structured interview is commonly used to conduct telephone interviews (Crabtree &
Semi-structured interviews are guided, concentrated, focused, and open-ended
communication events. The probes, prompts, and questions are written in the form of
a flexible interview guide (Crabtree & Miller, 1999). Semi-structured interviews are
designed to ask all participants the same questions; however the responses may vary
among the participants so that further probing by the researcher can be done if desired
Open-ended questions are often used in conjunction with structured
questionnaires. Open-ended questions allow researchers to gather information that
may not be obtained from the pre-determined set of responses used on many
questionnaires. The data gathered from all of these methods are often audiotaped or
videotaped and then usually transcribed verbatim with input from the entire research
team. This is especially true during the coding and thematic analysis processes.
There are a variety of coding methods used in the interpretation of qualitative
data. For example, when coding for focus groups, each investigator will review the
transcripts looking for key words and phrases, sometimes making notations of
emerging themes during the process. At this point the research team will often gather
to discuss their individual findings in an effort to ensure that there is a consensus in
the gathering of the data. The final goal is to cluster the data under major themes
until there are no additional themes emerging, a process called saturation (Crabtree &
For other qualitative methods such as interviewing, one method of organizing
the text is the use of codes from a coding manual. Development of the coding manual
is dependent on the level of detail to be coded, the goal of the research and the level
of previous understanding. Once the coding manual is prepared the text is
categorized by themes and content and then arranged in a manner that allows
interpretation of the data. One area of controversy regarding coding in general is that
of performing a literature review prior to the coding or performing the coding first.
The former allows a researcher to have some background knowledge prior to
developing a code, while the latter allows a researcher the freedom to create a code
based on the data in hand (Crabtree & Miller, 1999).
There are disadvantages in the use of qualitative research methods. One of the
more obvious differences is the increased amount of time required to gather the data
as well as the limited sample sizes that can be studied using qualitative methods. In
addition, because of the nature of the qualitative method that results in large amounts
of data, several hours of transcription and coding for common themes and content are
required. Fortunately there are software packages available that can help in managing
qualitative data (Morse, 1992).
Finally, the analysis of qualitative data can also be somewhat subjective,
which is often a criticism of this method of study. One way of dealing with the
problem of subjectivity is to have equal input from all members of the research team
throughout the analysis and interpretation of the data. For example, the triangulation
method is one way of dealing with the problem of subjectivity. In this method, three
investigators initially read and code four transcripts to identify specific content areas
and themes. The coding structure is reviewed by all of the researchers as a group to
be further refined for the coding of the other transcripts. The researchers then code
the further transcripts individually using the final coding structures to ensure a level
of validity and reliability of the information obtained from participants (Nekhlyudov,
Rong, & Fletcher, 2005).
Although subjectivity can be a problem in the evaluation of data from
qualitative research, this type of research is valuable in that it provides a deeper level
of data from participants than what can be elicited from survey or closed end
questions alone. Thus, this richer quality of data can provide researchers with
valuable information that can be used for further research often in conjunction with
What We Know About Screening Mammography Behavior
Qualitative methods have been used in several studies for breast cancer
including risk perception. For example, Covello and Peters (2002) used 6 focus
groups, composed of 8-10 female participants 45 to 54 years old, to examine
perceptions of risk for a variety of health concerns including breast cancer. The
women in these groups were asked general questions about health issues and then
questions about specific health issues that included breast cancer. The results of this
study indicated the need for a better understanding on the part of physicians of the
risk factors for breast cancer as perceived by women and how these relate to decision-
making on the part of the woman. For example, when asked What health problem or
disease would you say you fear the most at this point in your life? 58% of the
women in this study responded with breast cancer while only 9% stated heart disease,
which statistically has a higher mortality for women (Covello & Peters, 2002).
Ryan and Skinner (1999) conducted four focus groups to examine risk beliefs
and interest in genetic counseling among a group of FDRs of women with breast
cancer. In general, these FDRs did not understand risk factors for developing breast
cancer aside from having a family history and were only interested in being counseled
if there was something they could do about lowering the risk (p.104). Without
options for dealing with this increased risk, many of these women were ambivalent
about the value of knowing their genetic status: If they tell you are definitely going
to get it [breast cancer], what do you know? That sooner or later you are going to
have cancer? What are you going to do to prevent it? They dont have any prevention
yet (Ryan & Skinner, 1999,102).
In addition, the researchers found that these women felt strongly that their risk
of developing breast cancer was related to lifestyle risk factors but did not understand
the influence of epidemiological risk factors. For example, one woman whose mother
was diagnosed with breast cancer when she was pregnant with her first child, stated:
Oh good, Im pregnant. I am going to breastfeed. That is going to lower my risk
(Ryan and Skinner, 1999,101). Another woman stated: If you get a mammogram
before 30, your hormones are not.... completely developed and getting a
mammogram .... before 30 could spark something. (Ryan & Skinner, 1999,101).
In another study, Frost et al. (2000) used a sample of 78 subjects self-referred
or referred by their physician to the Familial Cancer Program of the Mayo Clinic to
examine perceived risk. All of these women had a family history of breast and/or
ovarian cancer and 39 of these women had a personal history of one of these cancers.
The instruments used in this study consisted of standard questionnaires as well as
For the 39 women without a personal history of cancer, information from
medical professionals (46%), support from family and friends (or lack of support)
(33%) and availability of screening (26%) were the most important factors related to
the subjects ability to adjust to being at an increased risk for cancer. These answers
were in response to the question: What do you find to be most difficult and most
helpful in your efforts to adjust to your perceived risk status? (Frost et al., 2000).
In responding to a second open-ended question, What impact has knowledge
of your increased risk for cancer had on your family?(p. 74), the primary concern of
44% of the women without a personal history of breast cancer was emotional distress
in terms of difficulty for the family in dealing with this knowledge as well as worry
about what the future would bring (Frost et al., 2000).
Chalmers & Luker (1996) used semi-structured, tape-recorded interviews to
examine the role of information, support and communication in helping FDRs of
women with breast cancer adapt to their status as women at an increased risk of
developing breast cancer. The women in this study clearly wanted information about
breast cancer, in part to help them in coping with their relatives illness as well as
adapting to their perception of risk for developing breast cancer themselves.
I did my duty of informing him [primary care provider] that there is a
high risk of breast cancer [in my family]. He should, at that point, take
charge and help me get going and taking care of it, you know: taking care
of myself to prevent it. (Chalmers & Luker, 1996, 206).
Friends and the family member with breast cancer were also perceived as
important and trusted sources of information; however, there was often a lack of
support: My friends -1 could tell them but I couldnt really tell them how it felt.
Other people would ask about my mom, but it was just social chit-chat. (Chalmers &
Luker, 1996, 208). This is important because the high risk womans perception of her
own risk was greatly influenced by the source and amount of information she
received from her family and friends as well as the health care system (Chalmers &
Qualitative studies have also been used to examine the level of psychosocial
distress of FDRs of woman with breast cancer. Wellisch et al. (1991) interviewed 60
daughters whose mothers had died of breast cancer using structured, open-ended
questions presented by an interviewer. Lengthy interviews included questions on
psychological functioning and emotional issues.
One finding suggested that subjects who were adolescents during their
mothers illness experienced greater levels of distress during their mothers illness, in
comparison to those subjects who were children or adults during their mothers illness
(Wellisch et al., 1991). In addition, the daughter perceived her mothers illness to
affect her own long-term plans in areas such as career options and/or choice of a
spouse. For example, one daughter expressed concern in finding a husband who
would be the type of man whom she could depend on to take care of her when she
was diagnosed with breast cancer (Wellisch et al., 1991,177).
DiProspero et al. (2001) used a focus group involving 8 women at increased
risk of developing breast cancer to obtain information on the effect of testing positive
for the presence of the BRCA 1 or BRCA 2 gene mutation. The most common
concern was the length of time between testing and receiving the results (up to 3
years). In addition, 75% of the women indicated that their level of worry increased
following their diagnosis: Now I see cancer in my future whereas before I would
have seen it in my past. I feel I am enveloped in a black cloud that never goes away.
I fear for my children in a way I never did before. (Di Prospero et al., 2001,1006).
Another study done by Nekhlyudov et al. (2005) used semi-structured
interviews via the telephone to evaluate the role of distress in the decision making
process for mammography in women younger than age 50. Sixteen Caucasian and
African American women between the ages of 38-45, who were receiving care at a
large New England medical practice, were randomly selected to participate in this
Although these women believed mammography to be a valuable tool for the
early detection of breast cancer, even in women younger than age 50, some women
expressed a level of distress. One woman who stated she had intentions to get a
mammogram, was still concerned that when women are diagnosed with breast cancer,
their whole worlds fall apart... everything changes, their relationships with their
husbands, their children and with others. (Nekhlyudov et al., 2005, 184).
All of the women trusted their provider for information related to breast
cancer as well as the providers recommendations for the timing of screening
mammography. However, all of the women brought up at least one issue related to
psychosocial factors including self-efficacy or being proactive about my health, a
focus on health and prevention or attaining peace of mind (Nekhlyudov et al., 2005).
Even though these women did not have an increased risk of developing breast
cancer they experienced a great deal of fear related to the possible development of
breast cancer. These women based this fear on their personal interactions with young
women who had died of the disease. In addition, these women perceived the
prevalence of breast cancer to be high based on the statistics commonly used such as
a 1 in 8 chance of developing breast cancer. It just seems so much more
pronounced than it was to me, in my own experiences. In the last couple of years it
just seems that I cant go anywhere without knowing somebody [with breast cancer].
(Nekhlyudov et al., 2005,186).
Issues surrounding the use of screening mammography and other early
detection options are another area in which qualitative studies have been done.
Bailey, Erwin, & Belin (2000) interviewed 60 African American women in 5 focus
groups to understand and apply cultural data to the development of an intervention
designed to increase the use of mammography. The qualitative data demonstrated the
importance of spirituality for African Americans: Its jus like Job in the Bible: some
folks think that youre bein punished for somethin youve done if you get cancer.
(Bailey et al., 2000, p.139) and the preference for role models in dispelling the fear of
fatalism: I came [to get a mammogram] because Murdie asked me to. (Bailey et al.,
The intervention associated with this study was titled the Witness Project. In
a follow-up survey of this intervention 67% of the women had undergone a
mammogram since the focus group and all of the women stated that they would be
willing to have their mammogram at the next scheduled visit (Bailey et al., 2000).
This is especially important in this group of women because they often have
advanced disease when they are diagnosed partly related to their lack of participation
in screening for the early detection of breast cancer.
Another study used semi-structured interviews to examine breast self-care
practices of women at increased risk of developing breast cancer. Most of the women
interviewed in this study felt a greater amount of anxiety in performing breast self
exams primarily related to their perception of doing it right or the anxiety related to
possibly finding a lump and how they would react: What does it feel like if you
should find something? What does it feel like if your breasts are hard but there is
nothing there.. .is it normal to feel that? Why does it hurt? (Chalmers & Luker,
However, most of these women felt more control over their feelings of the
threat of breast cancer by undergoing clinical breast exams and mammograms. This
was partly related to the fact that these two procedures were not dependent on their
ability or their expertise so that a negative result decreased their anxiety regarding
risk and psychosocial distress related to their increased risk of developing breast
cancer: I realize that sometimes things dont show up on mammography that women
have found themselves, but it [mammography] would be like insurance. (Chalmers
& Luker, 1996,1215).
The studies described above have provided a greater understanding in the
study of risk perception and psychosocial distress related to breast cancer. Although
quantitative research has determined that these two areas are important in discussions
related to developing breast cancer, especially in women at an increased risk, the
qualitative studies discussed here have provided more specific and a richer level of
data in understanding these issues.
For example, Chalmers & Luker (1996) found that many women at an
increased risk of developing breast cancer depend on family, friends and the health
care system for information regarding risk. Although many quantitative studies will
ask women to rate their perception of their risk for developing breast cancer, in most
cases quantitative studies are not designed to gather additional information as to how
a woman determines her level of risk including the sources of her information. The
information gained from the qualitative studies could be very helpful if a researcher
wanted to develop an intervention and/or a new quantitative survey or questionnaire.
As evidenced by the research described above, qualitative studies can provide
valuable information to researchers through the use of methods that allow participants
to share information and thoughts that may not be evident by using a closed end
survey or questionnaire. In addition, qualitative methods add to the understanding of
an issue allowing researchers to assess aspects of situations as they naturally occur
rather than explaining it (Polit & Hungler, 1993). However, qualitative methods are
often best used in conjunction with quantitative methods to ensure that the most
complete data are obtained. Quantitative studies allow a researcher to gather a large
amount of data in a reasonable amount of time as well as limit the influence of
subjectivity, which can be a problem in qualitative research. Larger representative
samples then allow for increased generalizability of the knowledge gained. Thus,
both methods play an important role in research and should be considered depending
on the research question being asked.
A theory involves concepts that provide a systematic view of relationships
between variables that can be used to explain and predict behaviors. Although
theories are abstract by nature, there should be some degree of operationalization that
allows a theory to be tested (Glanz et al., 1997). Theories are helpful in the
behavioral and social sciences where theoretical models are used to develop and
implement health promotion, disease prevention and behavior change interventions.
Theoretical models can be helpful in understanding why people or populations
engage in certain behaviors as well as how they might adopt a health protective
Gochman (1982) proposed a working definition of health behavior as those
personal attributes such as beliefs, expectations, motives, values, perceptions, and
other cognitive elements; personality characteristics, including affective and
emotional states and traits; and overt behavioral patterns, actions, and habits that
relate to health maintenance, to health restoration, and to health improvement (Glanz
et al., 1997, 32). Thus, models for predicting health behavior should take some or all
aspects of this definition into account.
Health Belief Model
The Health Belief Model (See Figure 3.1) is one of the earliest models
developed to explain and predict behavior. The HBM was developed in response to
the failure of a large number of adults to participate in tuberculosis screening
programs. Although these free screening programs were conducted in mobile units in
various neighborhoods to increase convenience and accessibility, very few people
participated in the screening program. In 1952, Hochbaum (1982) began studying
samples of the adults that had participated in the screening and assessed their
readiness to participate in this screening. He found that the decision to participate in
a screening program was strongly associated with perceived susceptibility and
perceived benefits (Rimer, 1997). According to this model, decisions regarding
adoption or change of a health behavior are based on three interacting variables that
include the following: perceived susceptibility, perceived severity, and perceived
benefits and barriers of the desired health behavior.
Perceived susceptibility is a persons subjective perception of risk or
vulnerability to a health threat. Perceived severity includes an individuals
assessment of the seriousness and evaluation of medical, clinical, and social
consequences of contracting an illness or leaving it untreated, as well as the impact on
others. Perceived benefits include an individuals belief that an action will be
efficacious, whereas, perceived barriers are an individuals assessment of the costs
involved in following through with a behavior change (Strecher & Rosenstock, 1997).
In 1974, Becker summarized findings from HBM research describing why
individuals do or do not choose to engage in recommended behaviors and found
considerable support for the HBM. An updated critical review of all studies done
using the HBM between 1974 and 1984 showed continued empirical support for the
use of the HBM (Stretcher & Rosenstock, 1997). Perceived barriers and perceived
susceptibility were the most powerful predictors of an individuals decision to
practice a specific health behavior, while perceived severity was the least powerful
predictor (Stretcher & Rosenstock, 1997).
The health belief model has been shown to be a good predictor of health
behaviors such as cessation of cigarette smoking (Weinburger, Greene, Mamlin, &
Jerin, 1981), AIDS- protective behavior (Catania, Kegeles, & Coates, 1990) and safe
sex (Basen-Engquist, 1992). The HBM has also been used in studies looking at BSE
and mammography. For example, Benedict, Goon, Hooman, & Holder (1997)
conducted a study to determine the frequency of BSE, CBE and mammography in
adult daughters of women with breast cancer. Questionnaires were mailed to women
participating in a breast cancer support group or being seen in one medical oncology
practice and these women with breast cancer were asked to have their adult daughters
complete the survey. Fear of breast cancer was found to inhibit performance of BSE
while talking to their mother acted as a cue to action to perform BSE.
Rutton and Iannotti (2003) looked at the different constructs of the HBM as
well as issue involvement with breast cancer and salience of breast cancer family
history. A total of 378 women with and without a family history of breast cancer
related to mammography behavior were assessed in this study. Perceived benefits
and barriers were similar in women with and without a family history of breast
cancer. However, cues to action, susceptibility, issue involvement, and salience of
family history increased mammography behavior in women with a history of breast
cancer in their family.
Finally, Rutton and Iannotti (2003) did another study looking at the
relationship of perceived benefits and barriers, perceived susceptibility, and cues to
action to screening mammography adherence. A total of 97 adherent women and 213
non-adherent women were included. Adherent women with a family history of breast
cancer reported greater benefits of mammography and greater response to cues to
action than women without a family history of breast cancer. Non-adherent women
with a family history of breast cancer expressed fewer perceived benefits related to
mammography and perceived breast cancer to be less severe than women without a
family history of breast cancer.
One of the strengths of the HBM is explication of the relationships between
perceived susceptibility or risk of the health threat, the perceived severity of the
potential health threat, and the perceived barriers and benefits of behavior change.
However, a major criticism of the HBM has been the lack of consistent measurement
of HBM concepts. In addition, the HBM has just recently considered the role of self-
efficacy in health behavior change that can be an important variable when studying
long term behavior change (Strecher & Rosenstock, 1997).
The concept of self-efficacy addresses an individuals belief that he or she is
able to follow through with the desired behavior changes despite challenges in many
different contexts (Bandura, 1977). Self-efficacy had not previously been considered
an important variable in the HBM because the HBM has been used most often to
explain simple, one-shot preventive actions. However, as the HBM is being used to
explain more lifestyle behaviors that require long-term changes, self-efficacy is
becoming a more critical component (Glanz et al., 1997).
Another potentially important variable in the HBM is cues to action. Cues to
action increase readiness to move towards the recommended behavior and refer to
any type of reminder about a health threat or health behavior that might encourage
behavior change. For example, advice from friends, family, and the media may
influence the likelihood that a person will act (Glanz et al., 1997; Harrison, Mullen, &
Green, 1992). In addition, some physician recommendations can be considered a cue
to action. It is important to note that cues to action have not been empirically studied
to a point that allows them to be definitively included as a variable in the HBM.
Figure 3.1 Health Belief Model
Individual Perceptions Modifying Factors Likelihood of Action
barriers to behavior
Perceived SusceDtibilitv/ Severitv of Disease - Perceived threat of disease - Likelihood of behavior change
Cues to Action
*Bolded items are constructs that included variables in the data analysis
Figure 3.2 Health Belief Model (Example)
Likelihood of Action
Perceived benefits minus perceived
barriers to behavior change
If I have BCa, it is more likely to be
diagnosed early if I get a mammogram
Getting yearly mammograms does not
make a difference so why should I put
myself through the anxiety I will
experience every time I get a
I do not have health insurance
Transactional Model of Stress and Coping
The Transactional Model of Stress and Coping (TMSC See Figure 3.3 and
3.4) allows researchers to evaluate the factors associated with coping reactions to
stress such as a health threat. The key concepts in the TMSC include stress, appraisal
(primary and secondary), coping effort (problem-focused and emotion-regulation),
person factors, and environmental factors. These processes are followed by short and
long-term outcomes (Lazarus & Folkman, 1984).
Stress is one of the primary variables used in this model and is based on a
number of subjective cognitive judgments that result from an interplay between
person and environment. In this model, situations are not considered stressful in and
of themselves, rather it is an individuals appraisal of the situation as harmful,
threatening, or stressful that determines the level of stress. This appraisal is based on
both internal and external factors that interact with each other (Zakowski, 1997).
Psychological stress is one possible outcome of person-environment interplay
in the face of a stressor. Specifically, Lazarus and Folkman (1984) view
psychological stress as a particular relationship between the person and environment
that is appraised by an individual as taxing or exceeding his or her resources and
endangering his or her well-being.
Primary Appraisal and Secondary Appraisal are two types of appraisal
processes in the TMSC model. Primary appraisal involves the assessment of stressful
situations. When an individual is faced with a threat or harm the individual will
respond with thoughts or actions specific to themselves. In addition, the individual
will assess the threat to determine his/her options to manage the situation and the
emotional reactions (psychological distress and risk perception) associated with the
Secondary appraisal involves the evaluation of the threat or challenge and
what might and can be done to manage the situation, as well as the costs and benefits
of a particular decision. Both primary and secondary appraisals are dependent on
cognitive processes that are involved in the decision-making process.
Coping is another important construct in this model. Lazarus and Folkman
(1984) define coping as constantly changing and behavioral efforts to manage
specific external and/or internal demands that are appraised as taxing or exceeding the
resources of the person.
The TMSC describes coping as consisting of two primary categories:
problem-focused coping and emotion regulation. Problem-focused coping strategies
are similar to problem solving strategies; however, this type of coping involves an
analytic thought process that is focused on the environment as well as self. These
strategies include evaluating the problem, developing alternative solutions, weighing
the risks and benefits, and selecting the most advantageous solution (Lazarus &
Emotion regulation coping is the second primary category of coping. This
form of coping can lead to defensive processes used in many types of stressful
encounters and can include actions such as avoidance, minimization, distancing,
selective attention, perception of positive value from negative events and emotional
distress (Lazarus & Folkman, 1984).
Although limited, the TMSC has been tested in some research studies. One of
the areas studied was the role of coping and breast self-examination. In the latter
study, a convenience sample of 269 women was recruited from an employee list at a
medical center and from a professional nurses group. These women were asked to
complete a survey booklet that included questions on BSE frequency, coping style,
and questions from the HBM scale based on HBM constructs (Balneaves & Long,
Figure 3.3 Diagram of Transactional Model of Stress and Coping
Perceived susceptibility Coping Effort
Perceived severity Problem Management
Motivational relevance Emotional Regulation
Perceived control over
Perceived control over
Dispositional coping style
Figure 3.4 Diagram of Transactional Model of Stress and Coping (Example)
I have a family history of BCa
My relatives who got BCa before
menopause did not do as well as
those who got BCa after
Problem Management: I need to talk to my PCP about
getting a mammogram every year
I am not going to get mammograms because I do not want
to deal with possible abnormal findings, especially since
my (FDR) was very sick and died anyway
My sleep is worse if I think about my relative having
I feel guilty that I do not have the BRCA gene when my
two sisters do
I get a mammogram every
year or as recommended by
Fear of developing BCa
affects my daily life in a
Perceived Control over Outcomes
I can get a mammogram every year to increase my
chances that BCa will be caught early
I have health insurance that covers yearly mammograms
There are follow-up tests that can be done if something
abnormal is found on a mammogram
Dispositional coping style -I am a very fearful person that prevents me
from taking action
I will do whatever I can in difficult situations
Social support: My family/ffiends are very close and we talk about how it
affects us to be at a higher risk of developing breast cancer
My family/'friends do not talk about these things
Whatever is meant to happen will happen regardless of
what I do (ex: God is in control)
Coping style was found to affect BSE practices in that women categorized as
defensive high anxious women (awareness of high anxiety even with attempts to
defend against it) and repressors (high anxiety but unaware of it) reported higher
BSE practice than did true high anxious (high anxiety level in everyday life) and
true low anxious women (low anxiety and no need for defensiveness) (Barron et al.,
The strength of the TMSC is the consideration of appraisal in a womans
perception of the health threat as well as strong consideration of multiple cognitive
and psychosocial factors as related to decisions to change a health behavior. In
addition, this model includes self-efficacy as a major variable. However, one of the
weaknesses of the TMSC is that the variables can be perceived as complex in terms
of their intended meaning and the relationship of the variables to each other
(Balneaves & Long, 1999).
As described in the literature review of this paper, FDRs of women with
breast cancer may experience significant psychological distress and inaccurate
perceptions of risk that can influence their decision to adhere to screening
recommendations. Given the level of psychological distress seen in some women
with an increased risk of developing breast cancer, it is important to consider how
these women perceive their level of risk and cope with the stress of a potential life-
threatening disease such as breast cancer, especially in relation to their participation
in screening practices. This may be even more important for women with a
hereditary or familial history of breast cancer. Thus, the TMSC is a model that could
be helpful in predicting mammography behavior in this group of high-risk women
and therefore will be examined in this study.
Review of HBM and TMSC
The HBM and the TMSC are two models that include some constructs
applicable to examining the role of psychosocial distress factors, perceived risk,
perceived control, and attitudes and beliefs in explaining the decision of a woman
with one or more FDRs with breast cancer having had a screening mammogram in the
Both models share the constructs of perceived susceptibility, perceived
severity, and benefits and barriers, however, there are also unique constructs in each
model. The HBM adds sociodemographics and more recently cues to action as
constructs with the current consideration of self-efficacy as an additional construct.
The TMSC includes Primary Appraisal and Secondary Appraisal as constructs that
include cognitive actions that may or may not be conscious on the part of the person
appraising the situation (Lazarus, 1999). In addition to the constructs of perceived
susceptibility and perceived severity shared with the HBM, motivational
relevance (how will this affect my well-being or what is at stake) and causal focus
(self-blame) are included in the Primary Appraisal. The Secondary Appraisal in the
TSMC includes perceived control and self-efficacy, as well as the construct of
Coping Effort that includes problem management and emotional regulation.
None of these constructs are included in the HIM. Sociodemographic information is
an important factor in understanding a persons perception and behavior in the
presence of a stressor. Although the HBM includes sociodemographic factors as a
construct it is not as clear a construct in the TMSC.
The TMSC is an important theory in that coping is not just a response to a
stressor but is strongly affected by the appraisal of the meaning of the stressor to the
individual. Thus this model stresses the role of cognition and assignment of
perceived meaning in the emotion process of coping, which most models fail to
address, including the HBM (Folkman & Lazarus, 2005). According to Lazarus
(1999), stress cannot be explained by a stimulus alone because the degree of the stress
response to the same stressor varies among and within people depending on perceived
Coping generally acts as a mediating rather than a moderating variable
because it is generated in the event and changes the original relationship of the
stressor and the outcome variable. A moderator is a variable present prior to the
appearance of the stressor (such as a personality trait) that interacts with other
conditions that result in a specific outcome (Folkman & Lazarus, 2005).
Although the HBM can help to explain the relationship between perceived
risk and getting a mammogram for women at an increased risk of breast cancer, this
model does not allow for the impact of the cognitive and emotional factors associated
with the behavior of mammography. The TMSC acknowledges the role of the coping
process as a mediator in the relationship of knowing one is at an increased risk of
breast cancer and the decision to have had a mammogram in the last year that goes
beyond understanding that one is at an increased risk of developing breast cancer.
In terms of perceived control, an important factor is that knowing one has an
increased risk of developing breast cancer is something a woman must learn to
manage. On the other hand, a mammogram can serve as a concrete behavior
associated with the possibility of early detection that may allow a woman to feel an
increased level of perceived control. Assessment of perceived control is present in
the TMSC but not the HBM.
Psychosocial distress, which can include avoidance behavior and intrusive
thoughts, must also be considered when examining the relationship of a stressor and a
behavior, in this case the stressor being at an increased risk of developing breast
cancer and the behavior of having had a mammogram in the last year. The role of
psychosocial distress in the process of coping is an important piece of emotional
management in the coping effort as described in the TMSC that is not present in the
HBM. The source of the stressor is also important. For example, denial that one is at
an increased risk of developing breast cancer is different than being in denial that
getting a mammogram will affect the outcome of detecting the disease at an early
Finally, the outcome is also different in these two models. In the HBM, the
outcome is the Likelihood of Behavior Change however; the outcome in the TMSC
is Adaptation, which includes emotional well-being and functional status in
addition to health behaviors. The outcomes of the TMSC would appear to
encompass a more holistic view of a goal of health promotion than behavior change
Although the HBM and TMSC share important constructs, the TMSC
acknowledges the critical role of examining coping effort in the relationship of being
at an increased risk of developing breast cancer and having had a mammogram in the
last year. As a result, the TMSC may help researchers and health care providers to
better understand the process of a womans decision to get mammograms by
addressing the role of coping on a womans response to the stressor of being at an
increased risk of developing breast cancer. Thus the TMSC takes the important role
of perceived risk as an additional step.
Unfortunately, neither of these theories account for all of the factors that might
influence a woman at an increased risk of developing breast cancer in choosing to
adhere to recommended screening mammograms. Although both models have a
construct including social support, neither model has a clear construct for attitudes
and beliefs. Regardless, theories are useful in guiding the development of hypotheses
that can aid the development of interventions for health promotion such as screening
mammography for the early detection of breast cancer, especially in women with a
family history of breast cancer.
General Critique of Theoretical Models
Although theoretical models are often used to explain and guide health
behaviors the use of these models has been criticized. Jane Ogden (2003) performed
a meta-analysis of 923 articles published between 1997 and 2001 in four primary
health related journals. These 47 articles focused on well-known models (i.e. HBM,
TRA, TPB, and PMT). Her finding was that the models used most often in health
behavior research fell short in terms of the authors described intention in using these
models to generate hypotheses and test the data (Ogden, 2003).
Second, Ogden (200) and others have found that many of these theories
explain very little of the variance. The models in the articles reviewed by Ogden
demonstrated a lack of support for the predicted associations and the models ability
to predict the chosen outcome. In addition, rather than rejecting the model the
authors provide explanations as to why the data may not have fit the model.
Thirdly, although several of these theories and models have been used to
explain behavior and develop health promotion programs, many of these popular
models focus on the outcome of behavior change with little description of how a
person or population reaches that outcome.
Finally, the most popular models are rarely tested against one another.
Weinstein (1993) searched the PsychLit database and found 205 articles written
between 1974 and 1991, which mentioned at least one theory in the title, abstract, or
index terms. Only 10 articles listed more than one theory and only four of the articles
involved an empirical comparison. The theories reviewed by Weinstein included the
HBM, subjective expected utility theory, the PMT, and the TRA.
These models have several similarities; however, there are important
differences. Even though these theories use different terms (i.e. perceived severity,
negative utility, and negative evaluation), all of the models assume that concern about
a negative outcome and a desire to avoid this outcome can create the motivation to
change a particular health behavior. As a result the individual may believe that this
change can decrease the chances of the negative outcome. The biggest difference
between these theories is in the combination of the variables used to predict a specific
action or outcome.
Regardless, the real goal in using any theory should be to decide which
variables and features in these theories best improve the understanding of health
protective behavior (Weinstein, 1993). Finally, researchers should be cognizant that
models may too simplistic in predicting and explaining some behaviors, especially
those behaviors that involve complex cognitive processes.
As described above, there are gaps in our understanding of the association of
psychosocial distress, perceived risk, and perceived control in women with an
increased risk of breast cancer. This knowledge is important since there could be an
important relationship between the variables listed and a womans general coping
behavior, that in turn affect her decisions, such as participation in screening
behaviors. This is especially important in the presence of the threat of being at an
increased risk of developing breast cancer (Audrain-McGovem et al., 2003; Fry &
Moreover, the relationship of sociodemographic factors and psychosocial
distress, perceived risk, and perceived control is not clear (Drageset & Lindstrom,
2005). Understanding these associations is critical for researchers in designing
interventions to increase mammography behavior in this group of high risk women,
given that psychosocial distress, perceived risk, and perceived control variables are
more than likely greatly impacted by sociodemographic factors but also modifiable
(Magai, Consedine, Conway, Neugut, & Culver, 2004). Finally, there is a significant
gap in understanding the relationships described above when comparing women with
one FDR versus those women with two or more FDRs and their increased risk of
developing breast cancer.
PURPOSE AND HYPOTHESES
The primary purpose of this study will be to reanalyze the Marcus et al. data
(in press) to examine and compare the role of psychosocial distress, perceived risk,
perceived control, and attitudes and beliefs as mediators in screening mammography
behavior in women with a SFDR versus MFDRs with breast cancer. Differences in
intrusive thoughts, avoidance behavior, perceived risk, and perceived control in
women with a SFDR versus women with MFDRs will also be examined. In addition,
the relationship of demographic variables, number of FDRs, psychosocial distress,
and perceived risk will also be explored.
Although not a primary purpose, the Health Belief Model (HBM) and the
Transactional Model of Stress and Coping (TMSC) were evaluated in terms of what
one would expect each theory to predict. It was expected that the constructs of
perceived risk and sociodemographic variables (as modifiers) in the HBM will be
supported by the data as predictors of screening mammography. In terms of the
TMSC, it was expected that there would be data to support perceived susceptibility
under Primary Appraisal, perceived control under Secondary Appraisal, and
psychosocial distress under Emotional Regulation in Coping Effort in predicting
Specifically, this study tested the following 10 hypotheses.
Rationale for Hypotheses 1 and 2
These hypotheses are based on previous studies that have shown that women with a
family history of breast cancer have increased distress including intrusive thoughts
and avoidance behavior (Aro et al., 2000; Zakowski et al., 1997). Hypotheses 1 and 2
stating that there will be a difference between women with a SFDR and women with
MFDRs have been developed based on the clinical experience of the researcher.
Hypothesis 1: Women with MFDRs will report a greater number of intrusive thoughts
than women with a SFDR.
Hypothesis 2: Women with MFDRs will report more avoidance behavior than women
with a SFDR.
Rationale for Hypothesis 3
There is literature supporting a relationship between perceived risk and having a
familial history of breast cancer (Lipkus et al., 1999; Montgomery et al., 2003).
However, there is a gap in the literature regarding differences in perceived risk
between women with a SFDR and women with MFDRs; therefore this hypothesis is
partially based on the clinical experience of the researcher.
Hypothesis 3: Women with MFDRs will report greater perceived risk than women
with a SFDR.
Rationale for Hypothesis 4
There is very little data examining the role of perceived control and having FDRs
with breast cancer, however, some literature suggests perceived control may play a
role in a womans decision to adhere to screening mammography recommendations
(Absetz et al., 2003; Osowiecki & Compas, 1999). The literature regarding
differences based on the number of FDRs could not be found; therefore this portion of
the hypothesis is based on the clinical experience of the researcher.
Hypothesis 4: Women with MFDRs will report lower levels of perceived control than
women with a SFDR.
Rationale for Hypothesis 5
The literature suggests a relationship of more psychosocial distress (Audrain-
McGovem et al., 2003; Julian-Reynier et al., 2001), greater perceived risk (McCaul,
et al., 1996; Hopwood, 2000), more perceived control (Aro et al., 1999; Lloyd et al.,
1996) and positive attitudes and beliefs towards detecting breast cancer (Donovan &
Tucker, 2000; Hughes et al., 1996) to greater adherence to screening mammography
guidelines. However, there is a gap in the literature examining these factors as
mediators between women with a SFDR and women with MFDRs with breast cancer
and adherence to screening mammography.
Hypothesis 5a: Intrusive thoughts will mediate mammography practices.
5b. Avoidance behavior will mediate mammography practices.
5c: Perceived risk will mediate mammography practices.
5d: Perceived control will mediate mammography practices.
5e: Attitudes and beliefs towards breast cancer screening will mediate mammography
5f: This effect will be stronger in women with MFDRs compared to women with a
Rationale for Hypotheses 6-10
Although sparse, most of the supporting literature for the relationship of
sociodemographic variables and intrusive thoughts, avoidance behavior, and
perceived risk has been embedded in research done with another primary focus
(Champion & Skinner, 2003; Katapodi et al., 2004). Given the limited research
regarding sociodemographic factors, hypotheses 6-10 are primarily based on the
findings from the Marcus et al. study (in press).
Hypothesis 6a: Ethnic females (i.e., non-Caucasians) will report a greater number of
intrusive thoughts than Caucasian women.
6b. Ethnic females (i.e., non-Caucasians) will report more avoidance behavior than
6c. Ethnic females (i.e., non-Caucasians) will report greater perceived risk than
6d: This difference will be enhanced for women with MFDRs.
Hypothesis 7a: Younger women will report a greater number of intrusive thoughts
than older women.
7b. Younger women will report more avoidance behavior than older women.
7c. Younger women will report greater perceived risk than older women.
7d: Having MFDRs will enhance this difference.
Hypothesis 8a: Less educated women will report a greater number of intrusive
thoughts than more educated women.
8b: Less educated women will report more avoidance behavior than more educated
8c: Less educated women will report less perceived risk than more educated women.
8d: This difference will be enhanced for women with MFDRs.
Hypothesis 9a: Women with lower incomes will report a greater number of intrusive
thoughts than women with higher incomes.
9b: Women with lower incomes will report more avoidance behavior than women
with higher incomes.
9c: Women with lower incomes will report less perceived risk than women with
9d: This difference will be enhanced for women with MFDRs.
Hypothesis 10a: FDRs with a daughter with breast cancer will report a greater
number of intrusive thoughts than FDRs with a mother or sister with breast cancer.
1 Ob: FDRs with a daughter with breast cancer will report more avoidance behavior
than FDRs with a mother or sister with breast cancer.
10c: FDRs with a daughter with breast cancer will report less perceived risk than
FDRs with a mother or sister with breast cancer.
lOd: This difference will be enhanced for women with MFDRs.
Overview of Primary Analysis by Marcus et al. (in press')
Marcus et al. (in press) conducted an extensive, three-part randomized trial in
1996 that examined a number of factors influencing the psychosocial well-being and
breast cancer screening practices of FDRs of breast cancer patients. This randomized
trial was designed to increase screening mammography among a large cohort of
female FDRs of breast cancer patients.
The authors examined the relationship of psychosocial distress, perceived risk,
attitudes and beliefs, and perceived barriers to having had a screening mammogram in
the last twelve months or intentions to get a mammogram in the next twelve months
in FDRs of women with breast cancer.
The original data set collected in 1996 by Marcus et al. (in press) contains
information from telephone interviews of 1,082 women with one or more first-degree
relatives (FDRs) diagnosed with breast cancer. Included in the telephone interview
were several measures to assess the psychosocial factors associated with the diagnosis
of breast cancer in a close relative.
The purpose of the Marcus et al. study (in press), part I was threefold: 1) to
characterize the level of psychosocial morbidity and perceptions of risk among a
sample of FDRs with one or more relatives with breast cancer, 2) to examine the
interrelationships between psychosocial morbidity and risk perceptions in this sample
of women at an increased risk of developing breast cancer, and 3) to identify the
significant predictors of psychosocial morbidity and perceived risk based on
sociodemographics and family history. The analysis of part I was extended in part II
to focus on the relationships between the measures of psychosocial morbidity and
perceived risk and self-reported screening mammography, both current practices and
future intentions. In part III the relationship between breast cancer screening
attitudes, beliefs and current screening mammography practices and future intentions
In the Marcus et al. (in press) study women with breast cancer (referred to as
index cases) were identified using information from 30 hospitals and local healthcare
systems nationwide. These index cases received an introductory packet of
information in the mail with a letter informing them that they were being contacted on
behalf of their physician whose name was on the letter. The index cases were then
contacted and interviewed (Appendix A) by study personnel at which time informed
consent was obtained from the index case (woman with breast cancer) to contact their
FDRs who did not have cancer (women at an increased risk of developing cancer).
This method of obtaining informed consent from the index case to contact their FDR
would not be allowed today given the strict HIPAA regulations enacted in 2003.
The index case interview was conducted to obtain a list of eligible FDRs
without breast cancer for study enrollment. The index case was asked to identify
which of their FDRs had been diagnosed with cancer and to identify the cancer site if
known. This information was used to construct a family history of cancer as reported
by the index case and was then coded to reflect the number of FDRs with breast
cancer in each family unit.
A total of 2,428 index cases were identified from hospital records of which
872 were ineligible, based on the eligibility criteria (described below). Of the
remaining 1,556 index cases, 841 interviews were completed (54%) with 34%
refusing to participate. The other 12% of women that agreed to participate in the
study could not be reached. A total of 1,346 eligible FDRs were identified from the
841 completed interviews. About 11% declined to participate and 8% could not be
reached even after multiple attempts. A total of 1,082 FDRs were eligible for the
study and chose to participate. The eligibility criteria for the FDRs included the
1) the FDR was female and 40+years of age
2) the FDR was already aware of the cancer diagnosis of the index case