Citation
A comparison between completers and non-completers of a psychosocial intervention for head and neck cancer patients

Material Information

Title:
A comparison between completers and non-completers of a psychosocial intervention for head and neck cancer patients
Creator:
Anderson, Derek Ryan
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
vi, 48 leaves : ; 28 cm.

Subjects

Subjects / Keywords:
Head -- Cancer -- Patients -- Counseling of ( lcsh )
Neck -- Cancer -- Patients -- Counseling of ( lcsh )
Head -- Cancer -- Psychological aspects ( lcsh )
Neck -- Cancer -- Psychological aspects ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (M.A.)--University of Colorado Denver, 2009. Clinical psychology
Bibliography:
Includes bibliographical references (leaves 42-48).
General Note:
Department of Psychology
Statement of Responsibility:
by Derek Ryan Anderson.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
465013022 ( OCLC )
ocn465013022

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

A COMPARISON BETWEEN COMPLETERS AND NON-COMPLETERS OF A PSYCHOSOCIAL INTERVENTION FOR HEAD AND NECK CANCER PATIENTS. by Derek Ryan Anderson B.S., Arizona State University A thesis submitted to the University of Colorado Denver in partial fuifiiiment of the requirements for the degree of Master of Arts Clinical Psychology 2009

PAGE 2

This thesis for the Psychology Master of Arts degree by Derek Ryan Anderson has been approved ---Evelinn Borrayo Kate DeRoche Allison Bashe, OCT 07/lr-/04 1 Date

PAGE 3

Anderson, Derek R. (M.A., Clinical Psychology) A Comparison Between Completers and Non-Completers of a Psychosocial Intervention for Head and Neck Cancer Patients. Thesis directed by Associate Professor Kristin Kilbourn, Ph.D., M.P.H. ABSTRACT There is a demonstrated need for psychosocial interventions aimed at improving symptoms and quality oflife in Head and Neck Cancer (HNC) patients. In order to develop feasible, acceptable programs, researchers and clinicians need to understand and identify the specific characteristics that differentiate those participants who will be likely to adhere to the psychosocial interventions versus those that will not (Ostroff et al., 2004). An awareness of potential barriers to intervention adherence could lead to a better understanding of the most effective recruitment approaches and successful participant retention strategies (Ostroff et al., 2004). The goal of this study was to identify characteristics ofHNC patients who successfully complete or dropout of an intervention designed to alleviate distress and improve quality of life. The current study examined selected physical, psychosocial, and demographic variables to determine if they were associated with adherence to a telephone counseling intervention for HNC patients undergoing cancer treatment. Overall, intervention completers were more likely than non completers to be older, unemployed, in a committed relationship, and to have quit using tobacco. Completers also reported better physical and functional well-being, as well as a lower level of pain disability. There were no other reportable differences between completers and non-completers due to non-significant values and small effect sizes. Having a better understanding of HNC patients' characteristics will benefit future research by providing information that can be used to design more effective recruitment and retention strategies for psychosocial interventions. This abstract accurately represents the content of the candidate's thesis. I recommend its publication.

PAGE 4

TABLE OF CONTENTS Tables ........................................................................................... vi CHAPTER 1. INTRODUCTION ............................................................................................ 1 Research Questions ............................................................................. 3 2. LITERATURE REVIEW .................................................................................. 5 Quality of Life ....................................................................................... 5 Pain ................................ : .................................................................... 6 Distress ................................................................................................ 7 Coping .................................................................................................. 8 Risk Factors: Substance Use ................................................................ 9 Social Implications .............................................................................. 1 0 intervention Adherence ....................................................................... 11 Summary of Literature Review ............................................................ 14 3. METHODS ................................................................................................... 16 EASE Intervention Study .................................................................... 16 EASE Recruitment.. ............................................................................ 17 Assessment Procedures ..................................................................... 17 Assessment Instruments ..................................................................... 18 Data Analysis ..................................................................................... 21 4. RESULTS ..................................................................................................... 24 Participants ........................................................................................ 24 Demographics .................................................................................... 26 iv

PAGE 5

Medical Variables ............................................................................... 27 Risk Factors ....................................................................................... 28 Tobacco Use ................................................................................. 28 Alcohol Consumption ..................................................................... 29 Psychological Variables ...................................................................... 30 Results Summary ............................................................................... 32 5. DISCUSSION ............................................................................................... 33 Demographic Variables ...................................................................... 33 Medical Variables ............................................................................... 35 Risk Factor Variables .......................................................................... 35 Psychological Variables ...................................................................... 36 Limitations .......................................................................................... 37 Future Directions ................................................................................ 40 BIBLIOGRAPHY ...................................................................................................... 42 v

PAGE 6

TABLES Table 1 Number of participants for demographic and medical variables ........................... 25 2 Chi square tests for demographic and medical variables ..................................... 27 3 Chi square tests for alcohol and tobacco use ...................................................... 29 4 Independent samples t test for psychosocial measures ....................................... 32 vi

PAGE 7

CHAPTER 1 INTRODUCTION Head and Neck Cancers (HNC) make up approximately 5% of all cancer malignancies in the United States. Worldwide, there are an estimated 644,000 new cases of HNC every year, and approximately 39,750 in the United States. New cases represent 3.2% of all cancers in the United States and 2.2% {12,460) of all cancer deaths. (Marur & Forestiere, 2008). The prevalence ofHNC is three times higher in men and more common among African-Americans (Jemal et al., 2005). Most Head and Neck Cancers originate in the cells that line the surface of the mouth, nose, or throat. Over 90% of all head and neck cancers are squamous cell, which can occur in the lal)nx, nasopharynx, orophal)Tix, hypopharynx, or the lip/oral cavity (Dobrossy, 2005). Numerous risk factors are attributed to an increased risk ofHNC, with alcohol and smoking being the most prevalent. Blot et al. (2007) found that up to 75% of new HNC cases can be attributed to alcohol or tobacco use. Over the last 10 years, the incidence of tongue and tonsil cancer among people under the age of 45 has significantly increased due to the Human PapillomaVirus (HPV). The Human PapillomaVirus is also a risk factor for HNC, but has higher survival rates than HNC caused by other common risk factors (e.g., smoking). (Shiboski, Schmidt & Jordan, 2005).

PAGE 8

Concurrent chemotherapy and radiation is the standard treatment for the majority ofHNC patients, and many will receive this regimen prior to surgery in order to minimize organ damage (Singh, 2002). Due to the often difficult treatment regimen for HNC, patients frequently experience intense pain and physical symptoms such as hoarseness, difficulty swallowing, communication difficulties, respiratory problems, and disfigurement (Chawla et al., 1999). Furthermore, HNC patients experience higher levels of distress, fatigue, and social impairment than other types of cancer (Semple, 2001 ). HNC patients often report psychosocial and physical symptoms during and after treatment; typically experiencing greater levels of distress and psychosocial needs compared to other types of cancer (e.g., breast and prostate) (Bolund, 1985). Although HNC patients experience a gradual lessening of acute treatment related symptoms after cancer treatment concludes, approximately 24% to 50% ofHNC patients report persistent and chronic symptoms long after treatment has ended. (Hammerlid, Silander, Homestam & Sullivan, 2001). The high level of physical and psychological symptoms experienced by most HNC patients supports the need for interventions aimed at improving symptom management and quality of life (Allison et al., 2004). Studies have shown improved quality of life for HNC patients who complete psychosocial interventions during or after treatment (Allison et al., 2004; Hammerlid, Ahlner-Elmqvist et al., 1999; Hammerlid, Silander et al., 200 1; Hammerlid, Persson, Sullivan & Thomas, 1999; Semple, Dunwoody, Kemohan & McCaughan, 2009). Understanding and identifying the specific characteristics ofHNC patients who are the most likely to participate and benefit from psychosocial programs is of critical importance to researchers and clinicians (Ostroff, Ross, Steinglass, Ronis-Tobin & 2

PAGE 9

Singh, 2004 ). An awareness of potential barriers to intervention adherence may lead to a better understanding of the most effective recruitment approaches and successful participant retention strategies (Ostroff et al., 2004). The purpose of the current study is to quantify the demographic, medical, and psychological variables associated with successful completion of a telephone counseling intervention designed to alleviate physical and psychological symptoms for HNC patients undergoing curative treatment at the University of Colorado Cancer Center. Research Questions There are few studies examining demographic, medical, or psychological differences between completers and non-completers of HNC interventions. This study used quantitative data from a psychosocial intervention baseline questionnaire (pre-intervention) of a pilot study to answer the following questions: RQ 1: Is there a significant difference between intervention completers and noncompleters based on demographic variables (i.e., age, employment, relationship status, and education)? RQ2: Is there a significant difference between intervention completers and noncompleters based on medical variables (i.e., Karnofsky Performance Status, (KPS) and tumor site?

PAGE 10

RQ3: Is there a significant difference between intervention completers and non completers based on specific HNC risk factors (e.g., alcohol and tobacco use)? RQ4: Is there a significant difference between intervention completers and non completers based on psychosocial factors (i.e., quality of life, pain, distress, coping self-efficacy, and social support)? 4

PAGE 11

CHAPTER2 REVIEW OF LITERATURE The general need for additional research related to HNC, and specific research regarding compliance to psychosocial interventions was established in Chapter 1. The following chapter provides a review of the literature regarding issues faced by HNC patients undergoing adjuvant therapy and the importance of psychosocial interventions in alleviating physical and psychological symptoms. Quality of Life Gotay and Bottomley ( 1996) found that there are four dimensions of Quality of Life which include physical, functional, psychological, and social well-being. According to Semple, Sullivan, Dunwoody and Kemohan (2004), quality of life covers all aspects of well-being that impacts patients and may include living standards and environmental factors. Cancer treatment can cause lower quality of life in one or more aspects of the cancer experience (Gritz et al., 1999). Head and neck cancer patients often experience disabilities and altered appearances associated with their treatment, which can increase the risk for psychosocial and medical complications (Vokes, Weichselbaum, Lippman & Hong, 1993). Many studies have shown that HNC patients have reported quality of life values well below the norm at diagnosis and tend to decrease steadily during and after 5

PAGE 12

treatment (Hammerlid & Taft, 2001). Funk et al. (1997) compared quality of life between a HNC sample to an age-matched, United States population norm and found that HNC patients had reported significantly lower levels of quality of life during treatment and up to 6 months after treatment. Due to the high morbidity rates among HNC patients, assessing quality of life is particularly important (Gritz et al., 1999). A better understanding of quality of life can lead to more effective treatment and improved recovery (Hammer lid, Persson et al., 1999). Fritz (2001) reported that the primary areas of concern for HNC patients undergoing treatment were depression, pain, treatment side-effects, and image disturbance. Many studies have now shown improved quality of life for HNC patients who complete psychosocial interventions during or after treatment (Allison et al., 2004: Hammerlid, Ahlner-Elmqvist et al., 1999; Hammerlid, Silander et al., 2001; Hammerlid, Persson et al., 1999; Semple et al., 2009). Pain Researchers have shown that 30%-80% of all cancer patients experience pain and the pain experienced by cancer patients frequently creates difficulties in achieving physical and psychosocial recovery (McGuire, Yarbro & Ferrell, 1995; Whale, Lyne & Papanikolaou, 2001). Pain has been described as a multidimensional experience which includes, physiological, cognitive, behavioral, affective, sensory, and sociocultural (McGuire, 1999). For HNC patients, pain is typically categorized into severe or chronic pain (Singh, 2002). Although pain may diminish over time, pain associated with HNC treatment can last for many years after treatment concludes (Whale et al., 200 I). The authors of a pain study revealed that tumor recurrence was the cause of pain for 35% of 6

PAGE 13

HNC patients, treatment was the cause of pain for 30%, and a combination ofboth caused pain in 25% ofHNC patients (Chua, Reddy, Lee & Patt, 1999). Griepp (1992) indicated that up to 90% of cancer pain can be controlled with analgesics, but there remains evidence that pain is often incorrectly assessed and undermanaged. The complex nature of surgical and radiotherapy treatment can often make it difficult to correctly assess pain levels for HNC patients (Wong, Wood & McLean, 1998). Additionally, HNC patients are often reluctant to communicate the severity of their pain due to physical communication impairments, or a belief that pain is something that should be dealt with on a personal level (Rodriguez, McMillan & Yarandi, 2004). A review by Bair, Robinson, Katon & Kroenke (2003) reported high rates of pain in 65% of depressed individuals, indicating a strong association between pain and depression. There is also evidence that depressed patients being treated for pain will be less likely to recover from their pain compared to patients not experiencing concurrent symptoms of depression and pain (Bair et al., 2003). Some of the characteristics that are associated with acute, uncontrolled pain are lack of knowledge, fear of dependence on pain medications, and the uncertainty regarding the meaning of the pain and/or poor medical management (Haisfeld et al., 1994). Untreated or poorly managed pain can slow the recovery process and further other symptoms, such as depression or anxiety (Whale et al., 2002; Bair et al., 2003). Distress Cancer patients experience high levels of distress in the form of an."Xiety and depression. Stefanak, Derogatis and Shaw (1987) found that 23.8% of oncology patients with varying diagnoses, reported moderate depression, and 8% reported severe depression. 7

PAGE 14

Depression is even more pronounced in HNC patients. Significant depressive symptoms have been reported in 20% to 40% ofHNC patients (Duffy et al., 2002; Hammerlid, Ahlner-Elmqvist et al., 1999). Cancers ofthe tongue and pharynx alone account for 20% of all suicides among male cancer patients (Farberow, Ganzler, Cutter & Reynolds, 1971). There is ample evidence that psychological distress is often times left untreated in cancer patients unless treatment is specifically sought out (Semple et al., 2004). Untreated anxiety and depression can persist for HNC survivors long after the completion of treatment. Petruson, Silander and Hammerlid (2003) found that at a three year follow-up after treatment, 25% ofHNC patients were still experiencing significant anxiety and 20% were experiencing depression. In addition to distress caused from HNC treatment, certain addictive behaviors, such as smoking, have been found to be related to distress in HNC patients (Duffy et al., 2007). Depression and anxiety are prevalent among cancer patients, but are often times more pronounced in HNC patients due to a persistent and prolonged occurrence of distress after conclusion of treatment (Petruson et al., 2003). The high level of distress typically observed in HNC patients demonstrates the importance of treating depression and anxiety in this population. Coping Most HNC patients struggle with uncertainty, while also attempting to cope with the challenges associated with symptom management and life disruption (Semple et al., 2009). Maladaptive coping techniques (e.g., denial, behavioral disengagement, avoidance) are often used by HNC patients to cope with stressful events which can make the cancer experience especially challenging. (Sherman, Simonton, Adams, Vural & Hanna, 2000). 8

PAGE 15

Researchers have shown that coping style is associated with psychological adjustment to cancer (Kugaya, Akechi, Okamura, Mikami & Uchitomi, 1999; Schnoll, MacKinnon, Stolbach & Lorman, 1995). According to Glanz and Lerman (1992), about 50% of the variance in psychological adjustment to cancer is attributed to coping skills. A few psychoeducational interventions have been developed to improve coping skills for HNC patients and have revealed that interventions focusing on coping strategies such as relaxation, problem-solving, goal setting, and cognitive coping can significantly improve quality of life, social functioning, physical symptoms, sleep, and distress. (Allison et aL 2004; Newell, Sanson-Fisher & Savolainen, 2002; Vivela et al., 2005). Education and training in adaptive coping skills can be a statistical powerful tool for cancer patients who are undergoing treatment and for those who are adjusting to the challenges of cancer survivorship. Risk Factors: Substance Use According to Blot et al. (2007), alcohol and tobacco use are the greatest risk factors for developing head and neck cancer. The majority ofHNC patients (85%90%) report past or present tobacco use (Harari, O'Connor, Fiore & Kinsella, 1995) and 35% to 46% ofHNC patients continue smoking following diagnosis and/or treatment (Ostroff et al., 2004; Terrell, Fisher & Wolf, 1998). Researchers also suggest that 41%-54% ofHNC patients continue to consume alcohol after treatment concludes (Allison et al., 2004; Deleyiannis, Thomas, Vaughn & Davis, 1996). Continued use of alcohol and tobacco after diagnosis of HN C has been found to significantly increase the likelihood of medical complications and recurrence (Gritz, Vidrine & Fingeret, 2007; Schwartz et al., 1994). 9

PAGE 16

Alcohol abuse in conjunction with invasive head and neck surgery can severely compromise cognitive functioning, rendering the patient unable to properly understand the scope of their disease (McCaffrey et al., 2007). There are also a number of demographic and medical characteristics that are associated with smoking cessation following diagnosis and treatment. A more extensive tumor stage and certain HNC sites (e.g., larynx, pharynx) were are associated with higher rates of smoking cessation for HNC patients (Ostroff et al., 2004 ). The high rates of alcohol and tobacco use by HNC patients before and after treatment highlight the need for psychoeducational interventions that implement coping based strategies and referrals, as needed, to substance abuse and tobacco cessation programs. In addition to recurrence rates, Duffy et al. (2002) showed that smoking and alcohol cessation can have a positive impact on quality oflife for HNC patients. A study conducted by Duffy et al. (2006) utilized a CBT-based intervention for HNC patients and reported a significant decrease in concurrent use of alcohol and tobacco, but no significant change for those who only used alcohol or tobacco nonconcurrently. The study by Duffy et al. (2006) is important because many HNC patients smoke and problem-drink concurrently, and these are found to work synergistically, rather than additively, on cancer recurrence (Maier, Dietz, Gewelke, Heller & Weidauer, 2005). When alcohol and tobacco use are combined, the risk for HNC is increases by15-fold. (Tuyns et al., 1988). Social Implications HNC patients are faced with many challenges during post-treatment and survivorship phases of their diagnosis. Often times, the symptoms and disfigurement that 10

PAGE 17

can occur as a result of surgery, radiation, and chemotherapy lead to a number of functional and psychosocial adaptations. Some of these functional hardships include difficulties eating, breathing, and communicating (Semple et al., 2004). Today, there are advanced developments in the treatment ofHNC such as reconstruction, conservation, speech therapy, prostheses, and specialist multi-disciplinary teams, which lessen the negative impact of the disease (Llewellyn, McGurk & Weinman, 2005). Nevertheless, many HNC survivors experience facial disfigurement and functional disability, leading to social isolation and withdrawal (Ostroffet al., 2004; Watt-Wattson & Graydon,l995). According to Head et al. (2004), social isolation can have a negative impact on adherence to medical regimens, which can affect long-term recovery and survival for HNC patients. F iegenbaum ( 1981) examined the efficacy of social skills training for a group of individuals that had recently experienced facial disfigurement. The intervention group showed significant improvements in anxiety and self-confidence which were maintained at the two-year follow-up. The high level of social dysfunction caused by HNC proves that there is a definite need for developing and implementing psychosocial interventions that address issues of stigma, psychological debilitation, and social isolation (Head et al., 2004). Intervention Adherence Although there is evidence suggesting that psychosocial interventions may improve many of the physical and emotional symptoms associated with medical treatment, researchers have documented high dropout rates for HNC patients who were recruited to participate in psychosocial interventions (Allison et al., 2004; Head et al., 2004; Ostroff et 11

PAGE 18

aL 2004; Van Den Brink et al., 2006). The Multiple Family Group (MFG) intervention study for HNC patients conducted by Ostroff et al. (2004 ), implemented rigorous recruitment and accommodation methods, but still showed very high dropout rates. Only 80 of 174 (46%) eligible families returned completed surveys. Ofthe remaining 80 families that returned surveys and were informed of the MFG workshops, only 15 ( 19%) of the families completed the six-hour workshop. Patients who did not participate in the study stated that they did not need family support, or they could not attend due to logistical reasons, such as geographical distance (Ostroff et al., 2004). Due to the difficulty in retaining participants for HNC interventions, researchers have begun implementing new methods of delivering the intervention content. Meyer and Mark (1995) called for more research to examine methods of reducing costs oftreatment delivery and implementation. Among some ofthe new methods of intervention delivery is the use of the telephone. Telephone interventions have allowed cancer patients to receive the intervention content at home, which has increased convenience by eliminated geographic barriers (Gotay & Bottomley, 1998; Marcus et al., 2002). Allison et al. (2004) carried out a psycho-educational intervention for HNC patients that offered three options of delivering the content of the intervention which included small group, one-on-one, and home format options. Aside from examining the differences in intervention modality, Allison hoped that this method would decrease attrition. Unfortunately, only 35 of 66 (53%) remained in the study from beginning to end. A recent HNC intervention study testing a "user-friendly" telehealth intervention system designed to give automated answers and advice to patients experiencing treatment related symptoms reported that only 42 of the 12

PAGE 19

75 (56%) patients who consented for the study completed the intervention (Head et al., 2004). There are many difficulties keeping HNC patients involved in psychosocial interventions, but there are also challenges to consenting patients to participate. A psychosocial intervention study by Duffy et al. (2006) focusing on smoking, alcohol use, and depression, reported significant recruitment challenges. Although they were able retain 83% of their participants throughout the course of the study, only 42% of those that were originally approached actually agreed to participate in the research. Despite new "user friendly" modalities of delivering psychosocial interventions, (e.g., telephone, computer) many HNC patients decline to participate, or elect to dropout of studies before completion. There are a variety of reasons that could account for high attrition rates. Duffy et al. (2006) found that HNC patients recruited from a VA Hospital were more likely to participate in their study than those recruited from the University hospital. The researchers postulated that the higher participation level at the VA Hospital may be due to veterans having fewer resources, and thus, greater motivation to participate in an intervention. In the same study, smokers were found to be more likely to participate than non-smokers. In an intervention study performed by Vivela et al. (2006), participants who elected to drop out were more likely to be less educated and to have earlier stage cancer. Van den Brink et al. (2006) carried out a HNC telemedicine intervention study in which 9 of20 refusals oftheir intervention group were due to a "computer phobia" and 15 of 25 control-group refusals were due to the high amount of effort in completing questionnaires. Since HNC patients tend to be older males, it is likely that more modern 13

PAGE 20

intervention modalities may be causing older participants to dropout or refuse to participate in interventions. Many studies have reported that younger participants are more likely to complete participation in these studies (Ostroff et al., 2004; Vi vela et al., 2006). All of these factors are important to consider when designing a HNC psychosocial intervention study. Overall, age, education, alcohol and tobacco use, social support, proximity, and intervention modality have all been shown to affect the outcome of attrition for HNC psychosocial interventions. Summary of Literarture Review In conclusion, there is a demonstrated need for psychosocial interventions aimed at improving symptoms and quality of life in HNC patients. In order to develop feasible, acceptable programs, researchers and clinicians need to understand and identify the specific characteristics that differentiate those participants who will be likely to adhere to the psychosocial interventions versus those that will not (Ostroff et al., 2004). An awareness of potential barriers to intervention adherence could lead to a better understanding of the most effective recruitment approaches and successful participant retention strategies (Ostroff et al., 2004). The goal of this study is to identify characteristics ofHNC patients who successfully complete or dropout of an intervention designed to alleviate distress and improve quality of life. The current study examined selected physicaL psychosocial, and demographic variables to determine if they were associated with adherence to a telephone counseling intervention for HNC patients undergoing cancer treatment. Having a better understanding ofHNC patients' characteristics will benefit 14

PAGE 21

future research by providing information that can be used to design more effective recruitment and retention strategies for psychosocial interventions. 15

PAGE 22

CHAPTER3 METHOD The following chapter addresses the methods and procedures for the current study, including an overview of intervention procedures, participant recruitment, assessment procedures, assessment instruments, and data analysis. EASE Intervention Study This study is a secondary data analysis of The Ease and Alleviating Symptoms Everyday (EASE) study, which examined the feasibility of a telephone-based counseling intervention for recently diagnosed HNC patients (PI: Kilbourn, NCI R-21; CA11535401). The goal of EASE was to conduct a feasibility study of a telephone counseling intervention to improve symptom management and psychosocial care among HNC patients who were undergoing curative cancer treatment at the University of Colorado Cancer Center. The program utilized counselor-initiated telephone calls targeted to HNC patients at strategic points during cancer treatment. The intervention phase of the study included up to 10 telephone counseling sessions delivered over a 16-24 week period. The intervention was structured into three categories: (1) Assessment and Appraisal/Coping Skills Training_ (2) In-Treatment Counseling, and (3) Re-entry Counseling, which corresponded with key phases in the illness treatment continuum, including before, during, and slightly after treatment. 16

PAGE 23

EASE Recruitment Patients diagnosed with HNC were recruited from the Radiation Oncology Clinic at The University of Colorado Cancer Center (UCCC). In order to be eligible for the study, patients had to: 1) have been diagnosed with a squamous cell carcinoma ofthe head or neck, 2) be receiving or planning to receive radiation treatment for their cancer, 3) be less than halfway through their radiation treatment, 4) be 18 years of age or older, 5) be able to read and sign an informed consent form in English. A member of the clinical medical staff identified potential participants and alerted a study team member (e.g., research assistant) about the patient's appointment time. The research assistant approached the individual in the reception area and briefly explained the study. If the potential participant was interested in the study, the team member met with them in a private clinic room after the appointment was complete to gather consent, provide an EASE orientation packet, and schedule the initial session. There were 27 HNC patients approached to participate in the study and 24 participants consented to participate in the study. Three participants dropped out of the study before taking their first telephone assessment, leaving 21 total participants. Assessment Procedure Questionnaire assessments were conducted over the telephone at baseline (prior to beginning the intervention) and follow-up (after the intervention) by highly trained telephone interviewers who work in AMC Cancer Research Center's Computer-Assisted Telephone Interviewing (CATI) Unit. The AMC Cancer Research Center of Denver specializes in cancer prevention and control research, as well as research for therapies and 17

PAGE 24

patient care. The CA TI specializes in computer-assisted telephone interviews. A member of CA TI set up the appointment for the baseline assessment after recruitment and consent were complete. There was an average of9.30 days from the time of recruitment until the baseline assessment was conducted. The baseline assessments were completed prior to the beginning of the counseling intervention. Participants received up to $60 in gift cards as compensation for completing their baseline assessment, follow-up assessment, and a follow-up focus group ($20 cards for each). Assessment Instruments Each participant completed two questionnaires, which took place prior to beginning the first telephone session (baseline) and just after completing the final session (post-intervention). However, the current study only examined computer-assisted telephone interview data collected baseline. The measurement/assessment tools included questions regarding demographics and substance use, as well as psychosocial scales that measured quality of life. pain, distress, coping, and social support. Furthermore, medical variables (e.g., Karnofsky Performance Status) were gathered by physicians and nurses from the medical charts. The following describes the eight measures examined in the current study. Demographics Participants reported their gender, age, education, employment status, and relationship status on their baseline questionnaire. 18

PAGE 25

Medical Variables The medical variables used in this study to discern intervention adherence will be tumor site and the Karnofsky Performance Status (KPS) (Shag, Heinrich & Ganz, 1984). Tumor site was categorized into oral (e.g., tongue) and non-oral (e.g., larynx, phal)'ILx). A KPS score is a temporal assessment of quality of life for cancer patients. The score is reported by patients' doctors in order to measure a patient's medical requirements, physical activity, and level of independence. The score is reported in increments of 10 from 0 to I 00, with I 00 being the highest level of functioning. The KPS is frequently assessed before, during, and after exposure to various interventions in order to stratify patients into subsets and determine the physical abilities of the patients before and after the intervention (Chawla et al., 1999). Risk Behaviors Participants were first asked if they currently use tobacco. If not, they were asked if they have ever used tobacco. Tobacco use was then stratified into a "yes" or "no" category. The same procedure was used to assess alcohol use, but was stratified into four categories: none, low (l-2 drinks per week), moderate (3-9 drinks per week), and high (10 or more drinks per week) use. Due to the small sample size, the four categories were reasonably based on the proper distribution of the participants throughout the four categories. Quality of Life The Functional Assessment of Cancer Therapy-Head and Neck (FACTH&N Version 4) (Cella et al., 1993) is a reliable and valid measure of Quality of Life (QOL). It combines generic questions about physical, emotional, functional, and social well-being 19

PAGE 26

with specific questions about head and neck cancer symptoms, side effects, and concerns. It has a 28-item generic core (FACT-G) plus 9 items specific to head and neck cancer and a high internal consistency (alpha= .89) (List et al., 1996). Patients rate all items on a 5point rating scale ranging from "not at all" to "very much". Pain Disability The Pain Disability Index (POI) was used to assess the degree to which pain interferes with functioning (Tait, Chibnall & Krause, 1990). The POI is a 7-item, self report scale with a high internal consistency (alpha= .86) (Chibnall & Tait, 1994). Participants were asked to rate their level of disability from 0 ("no pain interference") to 10 ("total pain interference") in the following functional areas: family/home responsibilities, recreation, social activity, occupation, sexual behavior, self-care, and life support activity. Cancer Specific Distress The Impact of Events Scale (IES) is a 15-item self report scale designed to measure two major psychological responses to stressful life events: avoidance and intrusion (Horowitz, 1979). Participants rated each question from 0 ("not at all") to 3 ("often") based on experiencing either avoidance or intrusion of specified thoughts during the past 7 days. Subscale scores are calculated for Intrusion (7 items; alpha= .78) and Avoidance (8 items; alpha= .82). Split-half reliability of the total scale is .86 (Horowitz, 1979). The current study used the experience of cancer as the stressful life event. Decision Self-Efficacy The Decision Self-Efficacy Scale (DSES) is an 11-item measure of confidence and belief in one's ability to make an informed decision, and has shown a high level of internal 20

PAGE 27

consistency (alpha=.84) (Bunn & O'Connor, 1996). Participants report scores on a scale from 0 ("not at all confident") to 10 ("very confident") based on the extent to which they believe in their own abilities to make decisions and obtain information about treatment options and expressing their concerns. Social Support Social support was measured using the Interpersonal Support Evaluation List -12 (Cohen & Hoberman, 1985; Cohen, Mermelstein, Karmack & Hoberman, 1985). Participants were rated on a scale from 0 ("definitely false") to 3 ("definitely true") based on "first person" statements concerning the perceived availability of potential social resources. The ISEL-12 is a well-validated measure of perceived social support which has demonstrated high internal consistency (alpha= .90) and is stable over time (r =.70) (Brookings & Bolton, 1988). Data Analysis Data was examined for outliers and normality. Skewness and kurtosis was determined to be the best venue for examining normality with a small sample size. Instead of using an absolute value of+ 1. 96, which would typically be used with a larger sample, an alpha level of0.01 with an absolute value of .58 was used to determine significant levels of skewness and kurtosis (Field, 2005). Independent t-tests were conducted to determine whether a significant difference between completers and non-completers were obtained for the five psychosocial measures. Also, Cohen's d was computed for comparison of group difference to determine the magnitude of difference between completers and non-completers for all psychological 21

PAGE 28

variables. The guidelines for effect size were 0.2 for a small effect, 0.5 for a moderate effect, and 0.8 for a large effect size (Cohen, 1988). All categorical data, including demographic, medical, alcohol, and tobacco variables, were examined using a chi-square test of independence. Variables were separated into two groups due to the chi-square requirement ofN 2: 5 in each category. Education was categorized into "an associates degree or higher" or "less than an associates degree". Relationship status was separated into "in a committed relationship" (i.e., married, committed relationship) or "not in a committed relationship" (i.e., single, divorced, widowed, legally separated). Employment was categorized into "employed" (i.e., full-time employed), or ''unemployed" (i.e., unemployed, retired). For medical variables, the KPS was separated into "90 or above" (i.e., 90, 100), or "below 90" (i.e., 70, 80), in order to meet statistical assumptions. Due to the low amount of current tobacco users (N = 2), only past tobacco use (yes or no) and tobacco cessation were compared with intervention completion. Current and past alcohol use (yes or no), as well as alcohol cessation was compared with intervention completion. Once the new variables were established, the chi square was conducted in order to determine if completers or non-completers were more likely to fall into one category of each variable. Finally, odds ratios were computed for each significant chi-square test in order to clearly emphasize the relevance of the findings. With a small sample size (N = 21 ), there was a high likelihood that analysis would not show significant values due to a lack of statistical power. The discrepancy between effect size and significance was due to the small sample and low amount of statistical power. A large effect size without significance indicates a high likelihood that a larger 22

PAGE 29

sample would have produced a significant difference. Odds ratios were reported for every chi square test of independence, regardless of significance, in order to demonstrate a more representative indication of the difference between completers and non-completers. Odds ratio cutoff values were determined based on practical use, with effect sizes of small (1 -3), moderate (3 5), and large (above 5). An odds ratio of I implies that the event is equally likely in both groups and anything over 1 implies that either completers or non-completers are that many times more likely to fit into a category. 23

PAGE 30

CHAPTER4 RESULTS Participants A total of 24 participants consented to participate in the study. There were 19 of the 24 participants that received at least one counseling session. The mean number of counseling sessions for participants was 4.26 and the standard deviation was 2.42. Completion of the intervention was defined as completing both the baseline and follow-up telephone assessments. It was determined that completion of both assessments was a good indicator of full involvement in the study because using counseling sessions to determine completion was not appropriate due to the variability in treatment paths and recruitment times (e.g., some were recruited before treatment and some up to two weeks after treatment began). Twenty-one completed the baseline assessment only (non-completers) and 11 completed both the baseline questionnaire and follow-up assessment (completers). All "completers" of the intervention had at least two telephone counseling phone calls. Table I provides an overview of the number and percentage of completers and non-completers in each of the categorical variables that include demographic, medical, and risk factor variables. 24

PAGE 31

Table 1. Number of participants for demographic and medical variables. Employment Status Employed Unemployed Education Level Variable Associates Degree < Associates Degree Relationship Status Committed Relationship Not in a Committed Relationship KPS Score 80 <80 Tumor Site Oral Non-Oral Tobacco Use Past Tobacco Use No Past Tobacco Use Tobacco Cessation No Tobacco Cessation Alcohol Use Current Alcohol Use No Current Alcohol Use Past Moderate/High Use No Past Moderate/High Use Akohol Cessation No Alcohol Cessation 25 # of Completers (%) 3 (27%) 8 (73%) 4 (36%) 7 (64%) 8 (73%) 3 (27%) 4 (57%) 3 (43%) 4 (40%) 6 (60%) 7(64%) 4(36%) 7(88%) 1(12%) 6(55%) 5(45%) 6(55%) 5(45%) 5(45%) 6(55%) # ofNon Completers (%) 7 (70%) 3 (30%) 6 (60%) 4 (40%) 2 (20%) 8 (80%) 4 (44%) 5 (56%) 5 (50%) 5 (50%) 5(50%) 5(50%) 3(60%) 2(40%) 3(30%) 7(70%) 5(50%) 5(50%) 6(67%) 3(33%)

PAGE 32

Demographic Variables The number of intervention completers (N = 11) and non-completers (N = 1 0) was nearly equivalent. An independent t-test was conducted to determine whether age was associated with completing the EASE intervention. Age did not significantly differ between intervention completers and non-completers t(l9) = l.61,p = .12, d = .70. Although the finding was not significant, the effect size suggests a large difference with the mean age of completers (M = 62.82, SD = 9. 70) higher than non-completers (M = 56.20, SD = 9.09). For the remaining demographic variables, 2 x 2 chi square tests of independence were conducted to determine if employment status, education level, or relationship status were significantly associated with intervention completion. Table 2 provides an overview of X2 and p values for all categorical demographic variables. There was a significant association between intervention completion and employment status, with intervention participants 6.21 times more likely to complete the intervention if they were unemployed. There was no significant association between intervention completion and education level, although there was a small effect size showing that non-completers were 2.63 times more likely to have an associate's degree or higher than completers. There was a significant association between intervention completion and relationship status, with intervention participants 10.67 times more likely to complete the intervention if they were in a committed relationship. Overall, intervention completers were more likely than non completers to be older, unemployed, and in a committed relationship. 26

PAGE 33

Table 2. Chi square tests for demographic and medical variables Variable x2 p-va!ue Employment Status 3.83 .05* Education Level 1.17 .28 Relationship Status 5.84 .02* KPS (pre-radiation) 0.04 .84 KPS (mid-radiation) 0.25 .61 Tumor Site 0.20 .65 Note: *p S05 Medical Variables Kamofsky Performance Status (KPS) and tumor site were tested for significant associations with intervention completion using 2 x 2 Chi square tests of independence. Table 2 provides an overview of X2 and p values for all categorical medical variables. There was no significant association between intervention completion and KPS pre-radiation, or KPS mid-radiation. Very small effect sizes showed that completers were 1.25 times more likely to have a higher pre-radiation KPS score (indicating better functioning) and 1.66 times more likely to have a higher mid-radiation KPS score. Tumor site also displayed no significant association with intervention completion and a small effect of completers being 1.50 times more likely to have a non-oral tumor site. As shown in Table I. the number of completers and non-completers based on KPS and tumor site is nearly even. Due to the 27

PAGE 34

similarity between completers and non-completers, odds ratios show small effect sizes. In conclusion, any small difference is likely due to chance and there are no reportable differences in completion based on medical variables. Risk Factors Tobacco Use Due to the small sample size in the EASE study, all risk factor variables were separated into dichotomous variables. Two by two chi square tests of independence were conducted for all adequate risk factor variables to determine if alcohol or tobacco use were significantly associated with intervention completion. Twelve out of 21 (57%) participants reported using tobacco before being diagnosed with cancer. At baseline assessment, only 2 of the 21 reported using tobacco, which represents an 83% cessation rate of tobacco use after diagnosis. Table 3 provides an overview of x: and p values for all tobacco-related variables. Although the majority of the participants quit using tobacco after diagnosis, cessation was not significantly associated with intervention completion. Computation of an odds ratio shows that although this was not a significant finding, there was a moderate effect size of completers being 4.67 times more likely to have ceased using tobacco after diagnosis. There was not a significant association between intervention completion and past tobacco use. The effect size was small, showing that past tobacco users were 1.75 times more likely to complete the intervention. Current tobacco use could not be analyzed for this study because of the small amount of current tobacco users. Overall, the marginal 28

PAGE 35

significance and moderate effect size of smoking cessation shows completers were more likely to have quit using tobacco after diagnosis. Table 3. Chi square tests for alcohol and tobacco use. Variable Past Tobacco Use 0.40 Tobacco Cessation 3.36 Past Moderate/High Alcohol Use 0.04 Current Alcohol Use 1.29 Alcohol Cessation 0.90 Note: *marginal significane Alcohol Consumption n r .53 .07* .85 .26 .34 There were 11 out of21 (52%) participants who reported using moderate to high amounts of alcohol before being diagnosed with cancer, but only 4 ofthe 21 (19%) continued using moderate or high amounts of alcohol after diagnosis. After diagnosis, 17 of the EASE participants reported consuming low or no alcohol. There were 55% of participants who reported ceasing alcohol completely after being diagnosed with cancer. Table 3 provides an overview of X2 and p values for all alcohol-related variables. Although over half of the participants stopped alcohol consumption after diagnosis, there was not a significant 29

PAGE 36

association between intervention completion and alcohol cessation. Odds ratio shows that there was a small effect size with completers 2.40 times more likely to have continued consuming alcohol after being diagnosed. There was not a significant association between intervention completion and moderate or high alcohol consumption in the past, with a small effect size showing that completers were 1.2 times more likely to have consumed moderate or high levels of alcohol in the past. There was also no significant association between intervention completion and current alcohol consumption. Despite a non significant finding for moderate or high alcohol consumption, the effect size was near the cutoff between small and moderate, with completers 2.80 times more likely to currently consume alcohol. Due to the non-significant findings and small effect sizes, there are no reportable differences in completion based on alcohol use. Psychological Measures All psychosocial measures were normally distributed with one exception. The Social Well-Being subscale (FACTGS) ofthe FACTHN was the only measure found to be non-normally distributed; therefore, a non-parametric Mann -Whitney test was conducted. Completers were not significantly different than non-completers regarding their social well being based on the Social Well-Being subscale (FACTGS) of the FACT HN, U= -.8l,p = .41). All other psychosocial variables were analyzed using independent sample t-tests to determine whether there was a significant difference between completers and non completers. As shown in Table 4, Physical (FACT-GP), Functional (FACTGF), and Head and Neck Specific (FACTHN) Well-Being were all significantly higher in 30

PAGE 37

completers of the intervention and showed large effect sizes. Emotional Well-Being was not significantly different among completers and non-completers, but the effect size for the FACT GE was moderate to large. The combined total of all of the FACT subscales displayed significantly higher well-being among completers and showed a large effect size. The PDI analysis revealed a significantly higher level of pain disability among non completers. Table 4 shows that there was no significant difference between intervention completers and non-completers based on the IES A viodance, IES Intrusive, DSES (decisional self-efficacy), and the ISEL -12 (social support). Overall, intervention completers had significantly higher levels of physical well-being, functional well-being, head and neck specific well-being, and a significantly lower level of pain disability. 31

PAGE 38

Table 4. Independent samples t-tests for psychosocial measures. Completers Non-Completers Independent Samples t-test Mean SD Mean SD p-value Effect Size (d) FACT-HN 161.62 23.00 136.10 17.76 2.82 .011 1.25*** Total FACT-HN 29.41 6.17 22.61 6.61 2.44 .025* 1.06*** GP FACT-HN 26.89 2.49 24.45 4.72 1.50 .149 0.67** GE FACT-HN 27.79 5.61 22.87 4.98 2.12 .048* 0.92*** GF FACT-liN 47.73 9.59 36.40 7.88 2.94 .008* 1.29*** H&N PDI 11.18 20.04 29.17 14.82 2.32 .032* 1.03*** IES-13.27 5.35 13.90 6.26 -0.25 .807 0.10 Avoidance IES-11.36 4.13 14.00 6.00 1.18 .252 0.52** Intrusion DSE 42.00 5.18 39.20 6.41 1.11 .282 0.48 ISEL -12 12.45 4.97 13.82 2.60 -0.78 .446 0.36 Note: *p < .05, ** .80 > d > .50 (moderate effect), *** d > .80 (large effect). Results Summarv Overall, intervention completers were more likely than non-completers to be older, unemployed, in a committed relationship, and to have quit using tobacco. Completers also reported better physical and functional well-being, as well as a lower level of pain disability. There were no other reportable differences between completers and noncompleters due to non-significant values and small effect sizes. 32

PAGE 39

CHAPTERS DISCUSSION The primary goal of the current study was to determine if there were any significant differences between completers and non-completers of a psychosocial intervention for HNC patients undergoing curative treatment at the University of Colorado Cancer Center. Previous research studies have revealed some of the characteristics of intervention completers and non-completers (Allison et al., 2004; Head et al., 2004; Ostroff et al., 2004; Van Den Brink et al., 2006). However, no studies to date have focused on a broad range of demographic, medical, substance use, and psychosocial variables that might differentiate completers and non-completers of a psychosocial intervention study for HNC patients. Demographic Variables Contrary to previous research stating that younger participants were more likely to complete HNC interventions (Ostroff et al., 2004; Vivela et al., 2006), we found the opposite relationship. Although the difference was not significant (p = .12), the average age of a completer was 6.62 years older than non-completers and the effect size (d = .70) for this finding was large. The large effect size indicates that there is a high likelihood that a bigger sample would produce a significant difference. The discrepancy between effect 33

PAGE 40

size and significance is due to the small sample and lack of statistical power, which will be addressed in more detail in the limitations section. One reason for the inconsistency with previous research could be due to older participants having less work responsibilities (e.g., retired or semi-retired), and therefore, may have more time than younger participants to speak to intervention counselors. In support of these ideas, we found that unemployed participants were 6.21 times more likely to complete the intervention than employed participants. We also noted interesting findings related to education. Although the difference was not significant, non-completers were 2.63 times more likely than completers to have an associate's degree or higher. This finding is contrary to previous studies (Vi vela et al., 2006) reporting that non-completers were more likely to be less educated. The effect size for education is small and the difference was not significant, making it likely that this contradictory finding could be due to chance. Nevertheless, this finding suggests the importance of examining the impact of education on recruitment and adherence in larger intervention trials. In terms of relationship status, completers were I 0.67 times more likely to be in a committed relationship than non-completers, which is contrary to the limited amount of previous research (Ostroff et al., 2004) showing no significant differences in relationship status between completers and non-completers. This finding could be due to the higher level of accountability that may be placed on the participant by their spouse or significant other to finish the study. Patients with HNC have been shown to experience higher levels of social withdrawal and isolation than patients with other forms of cancer (Watt-Wattson & Graydon, 1995). Participants in a committed relationship may experience less social 34

PAGE 41

isolation, which may help to eliminate treatment difficulties and increase intervention adherence. It is important to note that completers were more likely to be in a committed relationship, but did not report higher levels of social support (ISEL). This contradiction may be the result of individuals in committed relationships possessing certain personality traits (e.g., determination) conducive to completion. Medical Variables There were no significant differences between completers and non-completers based on tumor site or KPS, which supports Ostroff's finding that HNC intervention adherence was not associated with medical variables (Ostroff et al., 2004). A HNC intervention study conducted by Vivela et al. (2006) found that completers were more likely to have an earlier stage cancer, but tumor stage analysis could not be conducted with our sample because of the lack of variability, in that 86% ( 18 of 21) of participants had stage IV cancer. Risk Factor Variables Of the 12 past tobacco users in the sample, only two continued to use tobacco after being diagnosed with a HNC. Although the association between tobacco cessation and intervention adherence held only marginal significance (p = 07), there was a moderate effect size of completers being 4.67 times more likely have quit using tobacco than non completers. Cessation of tobacco or alcohol could be indicative of a higher level of determination and discipline regarding recovery, which could also affect one's desire to complete psychosocial steps toward a more effective recovery. Duffy et al. (2006) found that current smokers were more likely to complete a psychosocial intervention than non-35

PAGE 42

smokers, but this sample showed no significant association (p =.52) and a small effect size (1.75) between tobacco use and intervention adherence. Over half of EASE participants stopped alcohol consumption after diagnosis, but the association between intervention completion and alcohol cessation was not significant (p = .34). Current alcohol consumption was also not significantly associated with intervention adherence (p = .25), but the effect size was very close to the cutoff of a moderate effect showing that completers were 2.80 times more likely to complete the intervention. The greater likelihood of completers to currently consume alcohol could be due to a greater need for adaptive coping skills that alcohol users may lack. Although past tobacco use suggested a higher likelihood of completion, past alcohol consumption was not significantly association (p = .83) with intervention completion and was nearly even between the two groups. Intervention completers are more likely to have quit using tobacco and currently consume alcohol even after being diagnosed with HNC. The fact that completers are more likely to cease tobacco use, but not alcohol use may have important implications for future research. Despite these somewhat contradictory results, these findings emphasize the importance of assessing and monitoring past and current substance use given the potential impact on recruitment and adherence. Psychosocial Variables Completers ofthe EASE intervention had significantly higher levels of physical (FACT GP), Functional (FACT GF), and Head and Neck Specific (FACT HN) Well Being, as well as significantly lower levels of pain disability (POI). There were no significant differences between completers and non-completers on the emotional (FACT36

PAGE 43

GE) and social well-being scales (FACT-GS). Although Ostroffet al. (2004) found no significant differences between HNC participants in any domain of the FACT-HN, our finding suggests that the impact of cancer treatment on the physical and functional domains of quality of life can impact intervention adherence. This could be due to the fact that physical and functional deficiencies or disabilities may present barriers to participants' ability to speak and communicate effectively with their counselor. Hammerlid, Person et al. ( 1999) hypothesized that patients may feel too weak and be too busy with their treatment regimen to complete a psychosocial intervention during treatment. Understanding that physical and functional deficiencies play a larger role than emotional and social problems could help future researchers in designing more effective interventions that have a greater level of sensitivity and accommodate those that have diminished physical or functional capabilities. Limitations As mentioned earlier, this study was part of a NCI-funded pilot study (R-21) with the aim of assessing the feasibility of a telephone-based counseling intervention for recently diagnosed HNC patients. Self-selection bias was an inherent limitation to this type of study. Highly motivated patients were more likely to agree to participate in a psychosocial intervention, making the sample unrepresentative of a true HNC population. Most other limitations in the current study were related to our small sample size (N = 21 ), and therefore, these findings should be interpreted with caution. Much of the analysis was conducted using the same dependent variables, creating a potential for familywise error. In order to reduce familywise error, future researchers should consider conducting log linear 37

PAGE 44

analysis, as opposed to chi square tests of independence. Normality and a lack of statistical power were the major concern with this study prior to beginning analysis, but surprising, there were many significant findings, large effect sizes, and almost all variables were normally distributed. One of the limitations for this study was that some chi square tests of independence could not be conducted due to the minimum requirement of N 2: 5 in each category. Some specific examples of this were gender and tumor stage. Out of21, there were only 4 total females in the study and all but 3 ofthe 21 individuals had stage IV cancer, which prohibited any chi square analysis to be conducted on these variables. Not only did the minimum requirement inhibit examining certain important aspects of intervention adherence, but it also required that many of the categorical data be compressed to two categories in order to fulfill the minimum requirement of N 2: 5 (e.g., ''In a Committed Relationship = "Married" and "Committed Relationship"; Not In a Committed Relationship= "Divorced", "Widowed", "Legally Separated", and "Single"). When a broad range of answers to categorical questions is compressed, certain information is lost and cannot be accounted for. There were varied start times for participants, which may have affected the accuracy of various psychosocial measures. Most participants began the intervention before beginning treatment, but there were some who did not begin the intervention until after beginning treatment. Demographic, medical, and risk factor variables were not affected by this variation, but some psychosocial measures (e.g., pain disability) may be 38

PAGE 45

skewed if the questions were answered after beginning treatment. Future researchers should attempt to have all participants begin the intervention prior to starting treatment. Limiting the definition of "completers" as having finished both assessments is an issue of concern that may need to be modified in future studies. Although all "completers" had at least two counseling sessions, some may define a participant completing only one or two counseling sessions a "non-completer", regardless of having finished both assessments. The small sample size required us to develop our definition of"completers" based primarily on participants that completed their baseline and follow-up assessments. To increase the distinction between "completers" and "non-completers", future researchers should attempt to incorporate a specific amount of counseling sessions that would help to distinguish adherence. Another limitation was the lack of information received from those that did not complete the intervention. There were a few follow-up interviews conducted with those that had completed, but it was very difficult to conduct interviews with participants after they had dropped out. This information could have been very important in determining if there were similarities between non-completers regarding their reasons for not completing the study. Adding follow-up data to our understanding ofnon-completers baseline characteristics may have given a more complete description of these individuals. The major limitation regarding the analysis and results was that the small sample size created frequent discrepancies between significance and effect sizes. The low amount of statistical power meant that some differences may have been missed or caused by confounding factors. The effect size analysis did help to give a more detailed overview of 39

PAGE 46

the differences between completers and non-completers, but future research should attempt to accrue a greater sample size and carry out larger intervention trials in order to eliminate questions regarding the impact of significance values. Future Directions Although a larger sample size may have been useful for data analysis, there were still many interesting differences found between completers and non-completers. The significantly lower level of physical and functional well-being for non-completers suggests that researchers should consider conducting an intervention before or after treatment. Hammerlid, Persson et al. (1999) questioned whether HNC patients have the motivation to complete a psychosocial intervention during their medical treatment. Patients involved with an intervention before or after treatment may experience less physical and functional symptom interference, allowing the participant to focus more thoroughly on the intervention. In many ways the non-completers ofthis sample have the greatest level of need for a psychosocial intervention. There could be a benefit for future researchers to design a focus group with patients who match a non-completer profile and determine their specific needs, including what they believe to be the best intervention design. With a greater understanding of the characteristics of a non-completer, this may be feasible for future research. Overall, there is still much research to be done regarding psychosocial intervention adherence. This study displayed that there are definite differences between completers and non-completers of a psychosocial intervention involving HNC patients undergoing curative 40

PAGE 47

treatment. Understanding these disparities could provide useful insight for future researchers in designing and implementing more effective psychosocial interventions for HNC patients. 41

PAGE 48

BIBLIOGRAPHY Allison, P.J., Nicolau, B., Edgar, L., Archer, J., Black, M. & Hier, M. (2004). Teaching head and neck cancer patients coping strategies: results of a feasibility study. Oral Oncology, 40, 538-544. Bair, M.J., Robinson, R.L., Katon, W. & Kroenke, K. (2003). Depression and pain comorbidity. Archive of Internal Medicine, 163, 2433-2434. Blot, W., McLaughlin, J.K., Winn, D.M., Austin, D.F., Greenberg, R.S., Preston-Martin, S., et al. (2007). Smoking and drinking in relation to oral and pharyngeal cancer. Cancer Research, 48(11), 3282-3287. Bolund, C. (1985). and cancer: II. Medical factors in suicide by cancer patients in Sweden from 1973 to 1976. Psychosocial Oncology, 3, 17-30. Brookings, J.B. & Bolton, B. (1988). Confirmatory factor analysis ofthe interpersonal support evaluation list. American Journal a_( Community Psychology. 16(1), 137147. Brouha, X., Tromp, D., Hordijk, G.J., Winnubst, J. & DeLeeuw, R. (2005). Role of alcohol and smoking in diagnostic delay of head and neck cancer patients. Acta Otolaryngologica, 125, 552-556. Bunn, H. & O'Connor, A (1996). Validation of client decision-making instruments in the context of psychiatry. Canadian Journal of Nursing Research, 28(3), 13-27. Cassano, P. & Fava, M. (2002). Depression and public health: an overview. Journal of Psychosomatic Research, 53, 849-857. Cella, D.F., Tulsky, D.S., Gray, G., Sarafian, B., Linn, E., Bonomi, A, et al. (1993). The functional assessment of cancer therapy scale: development and validation of the general measure. Journal ofClinical Oncology, 11, 570-579. Chawla, S., Mohanti, B.K., Rakshak, M., Saxena, S., Rath, G.K. & Bahadur, S. (1999). Temporal assessment of quality of life of the head and neck cancer patients receiving radical radiotherapy. QualityofLife Research, 8(1), 73-78. 42

PAGE 49

Chibnall, J.T. & Tait, R.C. (1994). The pain disability index: factor structure and normative data. Archives o_(Physical Medical Rehabilitation, 75, 1082-1086. Chua, K.S., Reddy, S.K., Lee, M.C. & Patt, R.B. (1999). Pain and loss of function in head and neck cancer survivors. Journal o_(Pain Symptom Management, 18(3), 193-202. Cohen, J. (1988). Statistical statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Earlbaum Associates. Cohen, S. & Hoberman, H. (1983). Positive events and social supports as buffers of life change stress. Journal o.f Applied Social Psychology, 13, 99-125 Cohen, S., Mermelstein, R., Karmack, T. & Hoberman, H. (1985). Measuring the functional components of social support. In I. G. Sarason & B.R. Sarason (Eds.), Social support: theory, research, and application. Holland: The Hague Martinus Nijhoff. Deleyiannis, F.W., Thomas, D.B., Vaughan, T.L. & Davis, S. (1996). Alcoholism: independent predictor of survival in patients with head and neck cancer. Journal of the National Cancer Institute, 88(8), 542-549. Dobrossy, L. (2005). Epidemiology of head and neck cancer: magnitude of the problem. Cancer and Metastatis Reviews, 24, 9-17. Duffy, S.A., Ranis, D.L., Valenstein, M., Fowler, K.E., Lambert, M.T., Bishop, C., et al. (2007). Depressive symptoms, smoking, drinking, and quality of life among head and neck cancer patients. Psychosomatics, 48(2), 142-148. Duffy, S.A., Ranis, D.L., Valenstein, M., Lambert, M.T., Fowler, K.E., Gregory, L., et al. (2006). A tailored smoking, alcohol, and depression intervention for head and neck cancer patients. Cancer Epidemiology, Biomarkers & Prevention, 15( 11 ), 22032208. Duffy, S.A., Terrell, J.E., Valenstein, M.S., Ranis, D.L., Copeland, L.A. & Connors, M. (2002). Effect of smoking, alcohol, and depression on the quality of life of head and neck cancer patients. General Hospital Psychiatry, 24, 140-147. Farberow, N.L., Ganzler, S., Cutter, F. & Reynolds, D. (1971). An eight-year survey of hospital suicides. Threatening Behaviors, 1(3), 184-202. Fiegenbaum, W. (1981). A social training program for clients with facial disfigurations: a contribution to the rehabilitation of cancer patients. International Journal o.f Rehabilitation Research, 4(4), 501-509. 43

PAGE 50

Field, A. (2005). Discovering Statistics Using SPSS (2nd ed.). London: Sage Publications. Fritz, D. (2001). Life experiences of head and neck cancer survivors: a pilot study. Head and Neck Nursing, 19, 99-102. Funk, G.F., Kamell, L.H., Dawson, C.J., Means, M.E., Colwill, M.L., Gliklich R.E., et al. (1997). Baseline and post-treatment assessment ofthe general health status ofhead and neck cancer patients compared with United States population norms. Head and Neck, 19(8), 675-683. Gillison, M.L., Koch, W.M., Capone, R.B., Spafford, M., Westra, W.H., Wu, L., et al. (2000). Evidence for a causal association between human papillomavirus and a subset of head and neck cancers. Journal of the National Cancer Institute, 92(9), 709-720. Glanz, K. & Lerman, C. (1992). Psychological impact ofbreast cancer: a critical review. Annals of Behavioral Medicine, 14, 204-212. Gotay C. & Bottomley A. (1998). Providing psychosocial support by telephone: what is its potential in cancer patients? European Journal of Cancer Care, 7. 225-231. Griepp, M.E. (1992). Undermedication for pain: an ethical model. Advances in Nursing Science, 15, 44-53. Gritz, E.R., Carmack, C.L., de Moor, C., Coscarelli, A., Schacherer, C.W., Meyers, E.G. et al. (1999). First year after head and neck cancer: quality oflife. Journal a_( Clinical Oncology, 17(1 ), 352-360. Gritz, E.R., Vidrine, D.J. & Fingeret, M.C. (2007). Smoking cessation a critical component of medical management in chronic disease populations. American Journal of Preventive Medicine, 33(6), 414-422. Haisfield, M.E., McGuire, D.B., Krumm, S., Shore, A.D., Zabora, 1. & Rubin, H.R. (1994). Patients' and health care providers' opinions regarding advance directives. Oncology Nursing Fomm, 21(7), 1179-1187. Hammerlid, E., Ahlner-Elmqvist, M., Bjordal, K., Biorklund, A., Evensen, J., Boysen, M. et al. (1999). A prospective multicenter study in Sweden and Norway of mental distress and psychiatric morbidity in head and neck cancer patients. British Journal ofCancer, 80(5-6), 766-774. Hammerlid, E., Persson, L.O., Sullivan, M. & Thomas, W. (1999). Qualitv-of-life effects of psychosocial intervention in patients with head and neck cancer. The Journal o.f Otolaryngology -Head and Neck Surgery, 120(4), 507-516. 44

PAGE 51

Hammerlid, E., Silander, E., Homestam, L. & Sullivan, M. (2001). Health-related quality of life three years diagnosis of head and neck cancer-a longitudinal study. radiotherapy. Head and Neck, 23, 113-125. Hammerlid E. & Taft C. (200 I). Health-related quality of life in long-term head and neck cancer survivors: a comparison with general population norms. British Journal of Cancer, 84(2), 149-156. Harari, P.M., O'Connor, N.J., Fiore, M.C. & Kinsella, T.J. (1995). Radiation oncologists can assist head and neck cancer patients with smoking cessation. International Journal of Radiation Oncology, Biology, and Physics, 31(3), 645-649. Head, B.A., Studts, J.L., Bumpous, J.M., Gregg, J.L., Wilson, L., Keeney, C., et al. (2004). Development of a telehealth intervention for head and neck cancer patients. Telemedicine and e-health, 15(1), 44-52. Horowitz, M., Wilner, N. & Alvarez, W. (1979). Impact of event scale: a measure of subjective stress. Psychosomatic Medicine. 41 (3), 209-218. Jemal, A., Murray, T., Ward, E., Samuels, A., Tiwari, R.C., Ghafoor, A., et al. (2005). Cancer statistics, 2005. CA A Cancer Journalfor Clinicians. 55(1), 10-30. Kugaya, A., Akechi, T., Okamura, H., Mikami, I. & Uchitomi, Y. (1999). Correlates of depressed mood in amulatory head and neck cancer patients. Psycho-oncology. 8, 494-499. List, M.A., D'Antonio, L.L., Cella, D.F., Siston, A., Mumby, P., Haraf, D. et al. (1996). The performance status scale for head and neck cancer patients and the functional assessment of cancer therapy-head and neck scale. Cancer, 77(11), 2294-2301. Llewellyn, C. D., McGurk, M. & Weinman, J. (2005). Are psycho-social and behavioural factors related-quality of life in patients with head and neck cancer? A systematic review. Oral Oncology. 41,440-454. Maier, H., Dietz, A., Gewelke, U., Heller, W.D. & Weidauer, H. (1992). Tobacco and alcohol and the risk of head and neck cancer. The Clinical Investigator, 70(3-4), 320-327. Marcus, A., Garrett, K., Kulchak-Rham, A., Barnes, D., Dortch, W. & Juno, S. (2002). Telephone counseling in psychosocial oncology: a report from the cancer information and counseling line. Patient Education Counseling. 46. 267-275. 45

PAGE 52

Marur, S. & Forastiere, A.A. (2008). Head and neck cancer: changing epidemiology, diagnosis, and treatment. Mayo Clinic Proceedings, 83(4), 489-501. McCaffrey, J.C., Weitzner, M., Kamboukas, D., Haselhuhn, G., Lamonde, L. & Booth Jones, M. (2007). Alcoholism, depression, and abnormal cognition in head and neck cancer: A pilot study. The Journal of OtolaryngologyHead and Neck Surgery, 136(1), 92-97. McGuire, D.B. (1999). Measuring Pain. In M.F. Stromborg, S. Olson, S. Instruments for Health-care Research: (pp. 528-564). Sudbury, Mass: Jones & Bartlett. McGuire, D.B., Yarbro, C.H. & Ferrell, B.R. (Eds.). (1995). Cancer Pain Management, 2nd Edn. Boston: Jones and Bartlett. Meyer, T. & Mark, M. (1995). Effects of psychosocial interventions with adult cancer patients: a meta-analysis of randomized experiments. Health Psychology, 14, 101108. Newell, S.A., Sanson-Fisher, R.W. & Savolainen, N.J. (2002). Psychological impact of breast cancer: a critical review. Journal of the National Cancer Institute, 94(8), 558-584. Ostroff, J., Ross, S., Steinglass, P., Ronis-Tobin, V. & Singh, B. (2004). Interest in and Barriers to Participation in Multiple Family Groups Among Head and Neck Cancer Survivors and Their Primary Family Caregivers. Family Process, 43(2), 195-208. Petruson, K.M., Silander, E.M. & Hammerlid, E.B. (2003). Pain experience following radical treatment for head and neck cancer. Head and Neck, 25(7), 576-584. Rodriquez, C.S., McMillan, S. & Yarandi, H. (2004). Pain measurement in older adults with head and neck cancer and communication impairments. Cancer Nursing, 27(6), 425-433. Schag, C.C., Heinrich, R.L. & Ganz, P.A. (1984). Karnofsky performance status revisited: reliability, validity, and guidelines. Journal ofClinical Oncology, 2(3), 187-193. Schnoll, R.A., MacKinnon, J.R., Stolbach, L. & Lorman, C. (1995). The relationship between emotional adjustment and two factor structures of the mental adjustment to cancer (MAC) scale. Psycho-oncology, 4, 265-272. Schwartz, L.H., Ozsahin, M., Zhang, G.N., Touboul, E., De Vataire, F., Andolenko, P., et al. ( 1994). Synchronous and metachronous head and neck carcinomas. Cancer. 74(7), 1933-1938. 46

PAGE 53

Semple, C.J. The role of the CNS in head and neck oncology. Nursing Standard, 15(31), 39-42. Semple, C.J., Dunwoody, L., Kernohan, W.G. & McCaughan, E. (2009). Development and evaluation of a problem-focused psychosocial intervention for patients with head and neck cancer. Support Care Cancer, 17, 379-388. Semple, C.J., Sullivan, K., Dunwoody, L. & Kernahan, W.G. (2004). Psychosocial interventions for patients with head and neck cancer. Cancer Nursing. 27(6), 434441. Sherman, A, Simonton, S., Adams, D., Vural, E. & Hanna, E. (2000). Coping with head and neck cancer during different phases of treatment. Head and Neck, 22(8), 787793. Shiboski, C.H., Schmidt, B.L. & Jordan, R.C.K. (2005). Tongue and tonsil carcinoma: increasing trends in the U.S. population ages 20-44 years. Cancer, 103(9), 18431849. Singh, B. (2002). Rehabilitation and quality of life assessment in head and neck cancer. In J.P. Shah, ed., Cancer of the head and neck. Ontario: B.C. Decker. Inc. Stefanek, M.E., Derogatis, L.P. & Shaw, A (1987). Psychological distress among oncology outpatients. Prevalence and severity as measured with the brief symptom inventory. Psychosomatics. 28(10), 530-539. Tait, R.C., Chibnall, J.T. & Krause, S. (1990). The pain disability index: psychometric properties. Pain. 40, 171-182. Terrell, J.E., Fisher, S.G. & Wolf, G.T. (1998). Long-term quality oflife after treatment of laryngeal cancer. Archives ofOtolaryngol Head & Neck Surgery, 124(9), 964-971. Terrell, J.E., Nanavati, K.A, Esclamado, R.M., Bishop, J.K., Bradford, C.R. & Wolf, G.T. (1997). Head and Neck cancer-specific quality of life. Archives ofOtolaryngol Head & Neck Surgery, 123(10), 1125-1132. Tuyns, A.J., Esteve, J., Raymond, L., Berrino, F., Benhamou, E., Blanchet, F., et al. (1988). Cancer ofthe larynx/hypopharynx, tobacco, and alcohol: IARC international case-control study in Turin and Varese (Italy), Zaragoza and Navarra (Spain), Geneva (Switzerland), and Calvados (France). International Journal of Cancer. 41(4), 483-491. 47

PAGE 54

Van Den Brink. J.L., Moorman, P.W., De Boer, M.F., Hop, W.C.J., Pruyn, J.F.A., Verwoerd, C.D.A., et al. (2006). Impact on quality of life of a telemedicine system supporting head and neck cancer patients: A controlled trial during the postoperative period at home. Journal ofthe American Medical Informatics Association, 14(2), 198-205. Vivela, L.D., Nicolau, B., Mahmud, S., Edgar, L., Hier, M., Black, M., et al. (2006). Comparison of psychosocial outcomes in head and neck cancer patients receiving a coping strategies intervention and control subjects receiving no intervention. The Journal ofOtolaryngology, 35(2), 88-96. Vokes, E.E., Weichselbaum, R.R., Lippman, S.M. & Hong, W.K. (1993). Head and neck cancer. The New England Journal ofMedicine, 328(3), 184-194. Watt-Wattson, J. & Graydon, J. (1995). Impact of surgery on head and neck cancer patients and their caregivers. The Nursing Clinics of North America, 30(4), 659-671. Whale, Z., Lyne, P.A. & Papanikolaou, P. (2001). Pain experience following radical treatment for head and neck cancer. European Journal of Oncology Nursing, 5(2). 112-120. Weymuller, E.A., Yueh, B., Deleyiannis, F.W., Kuntz, A.L., Alsarraf, R. & Coltrera, M.D. (2000). Quality oflife in head and neck cancer. Laryngoscope, 110(3), 4-7. Wong, J.K., Wood, R.E. & McLean, M. (1998). Pain preceding recurrent head and neck cancer. Journal ofOrofacial Pain, 12(1), 52-59. 48