Facilitating research participation and improving quality of life for African American prostate cancer survivors and their intimate partners

A pilot study of telephone-based coping skills training

Abstract

African American men experience worse prostate cancer outcomes compared with those of Caucasian men, not only in incidence and mortality rates, but also in coping with the side effects of treatment. Unfortunately, African American men have been significantly under-represented in research evaluating the efficacy of psychosocial interventions for improving coping in prostate cancer survivors. This pilot study explored the feasibility and efficacy of coping skills training (CST), an intervention developed to enhance coping with treatment side effects in a sample of African American prostate cancer survivors and their intimate partners. The intervention was delivered in a telephone-based format designed to facilitate research participation. A total of 40 couples were randomized to either 6 sessions of CST or usual care. Survivors completed measures of disease-specific quality of life (QOL) related to urinary, sexual, bowel, and hormonal symptom domains, as well as measures of global QOL (i.e., physical functioning and mental health). Partners completed measures of caregiver strain, mood, and vigor. Analysis of data from 30 couples (12 couples in CST, 18 couples in usual care) indicated that CST produced moderate to large treatment effects for QOL related to bowel, urinary, sexual, and hormonal symptoms. Partners who underwent CST reported less caregiver strain, depression, and fatigue, and more vigor, with moderate effect sizes observed that approached conventional levels of statistical significance. These preliminary findings suggest that telephone-based CST is a feasible approach that can successfully enhance coping inAfrican American prostate cancer survivors and their intimate partners. Cancer 2007. © 2006 American Cancer Society.

Prostate cancer is the most frequently diagnosed cancer in men.1 It has been estimated that 232,090 men will be diagnosed with prostate cancer in 2005, with African American men being diagnosed at 1.6 times the rate of Caucasian men.1, 2 Prostate cancer is highly treatable, with 5-year survival rates ranging from 96% to almost 100% for men who are diagnosed in earlier stages. Survival rates drop precipitously in men diagnosed after prostate cancer has spread beyond local and regional disease.1 However, regardless of stage at diagnosis, survivors must often cope with physical and emotional symptoms that persist long after acute recovery from treatment.3, 4

There is a growing consensus that there is a significant gap between the ongoing informational and emotional needs of prostate cancer survivors and the psychosocial interventions available to them. Prostate cancer patients often avoid disclosure about their illness (diagnosis, ongoing symptoms, and treatment) for fear of burdening loved ones and/or being stigmatized.5, 6 In addition, compared with female cancer survivors, male prostate cancer survivors are much less likely to attend support groups,7 because of the embarrassing and personal nature of their symptoms. Finally, many prostate cancer patients report that during their medical treatment they are given little emotional support or guidance on how to communicate their feelings and symptoms associated with their disease to their family and friends.8–10

Unfortunately, African American men have been significantly under-represented in the cancer survivorship literature and in cancer research in general. African American participation rates in cancer clinical trials (5%–10%) are comparable to that of Hispanics (5%), higher than that of Asians (2%), and far below that of Caucasian participants, which is typically 80% or higher.11, 12 This is particularly troubling in light of the well-documented prostate cancer disparities between African American and Caucasian men, including a 60% higher incidence rate, slower recovery from treatment, lower QOL related to physical and psychological symptoms (e.g., increased pain, lower self-esteem), and a mortality rate that is more than twice of that of Caucasian men.1, 13–16

There is also a dearth of research on spouses and intimate partners of African American prostate cancer survivors. Spouses/partners of prostate cancer patients face multiple demands, including helping the patient deal with difficult and embarrassing treatment side effects (e.g., erectile dysfunction and urinary incontinence) and providing emotional support,5, 6, 8 while at the same time coping with increases in their own emotional and physical strain.17–22 Like studies with survivors, however, studies of spouses/partners have been conducted primarily with Caucasian samples. Therefore, little is known about how African American men and their partners cope after prostate cancer treatment or about the efficacy of psychosocial interventions to enhance coping among this group of survivors.

To identify and adequately respond to the needs of African Americans coping with prostate cancer, clinical researchers must address a number of barriers to recruitment and research participation. These barriers include long-standing suspicion and mistrust of medically-related research related to past abuses by clinical researchers;23–25 the cost associated with travel to the major tertiary cancer centers, where much of this research is conducted;26 and physical symptoms (e.g., urinary incontinence) that can interfere with willingness to travel to participate in research.27 For research studies evaluating psychosocial interventions, additional barriers may exist for African American men related to social stigma associated with psychological services, including concerns that seeking such services may be perceived as a sign of weakness.28, 29

In summary, the barriers described above illustrate many of the obstacles to conducting prostate cancer disparities research. To overcome these obstacles, there has been a call for increased cultural competency in health disparities research. Conceptual models and guidelines related to both treatment and clinical research have been put forward to help steer this work.30, 31 Collectively, these efforts to guide disparities research may be best described as contextual approaches that highlight the importance of considering the multiple contexts in which the participants are situated (e.g., economic, disease-related, family, ethnicity) at every stage of the research process, from study design and recruitment to data interpretation and dissemination.30–32

Guided by this contextual perspective, the current study includes several features that hold promise for facilitating research participation and addressing the psychosocial needs of African American prostate cancer survivors. First, the study evaluated coping skills training (CST), a comprehensive intervention that addresses the broad scope of survivorship issues, including physical, emotional, and social functioning. CST is based on cognitive behavioral theoretical approaches to pain and other symptom management.33, 34 CST involves teaching cognitive and behavioral techniques that increase self-efficacy for symptom management, which can reduce symptom severity and emotional distress.35 In controlled research studies, cancer survivors who have undergone CST report improvement in cancer symptoms, self-efficacy, and mood.26, 36, 37 Prostate cancer survivors, however, especially African American survivors, remain largely unstudied in the research evaluating CST.

Second, the study uses a telephone-based protocol. Telephone-based educational interventions have appeared in the research literature with increasing frequency in recent years.38–41 These interventions can minimize transportation costs and avoid symptom exacerbation associated with travel.27 They may also be a more acceptable way for African American men to receive supportive care involving psychological services without going to a “therapist's” office. Telephone-based interventions with prostate cancer survivors vary from brief interventions involving a single educational session38, 41 to several sessions conducted over weeks40 or months.38 To date, these studies are few in number, and African American representation has been minimal in most studies,38, 39, 41 suggesting that more work is needed to realize the promise of telephone-based protocols for reaching understudied groups.

Another important feature of this study is inclusion of intimate partners. Some efforts to include support persons in educational interventions with prostate cancer survivors have occurred in recent years, either by including family members40, 42 or by matching newly diagnosed men with long-term survivors based on ethnicity.27 Intimate partners, however, remain noticeably absent from most intervention studies.38, 39, 41–43 For African American men, inclusion of intimate partners in a comprehensive psychosocial intervention integrates an important component of the social context in which survivors are situated and may increase their comfort level with a psychological intervention, despite social stigma surrounding psychological services.

This preliminary investigation addresses many of the challenges outlined earlier by examining the feasibility and efficacy of a low-cost, telephone-based CST protocol designed specifically to 1) facilitate research participation and 2) enhance QOL among African American prostate cancer survivors and their intimate partners.

MATERIALS AND METHODS

Participants

This study was approved by the Institutional Review Board, and informed consent was obtained from all participants. Eligible participants were 1) African American men who were beyond the acute diagnosis and treatment phase for prostate cancer and with Karnofsky Performance Status scores44 ≥60 (reflecting an ability to care for self with only occasional assistance needed), and 2) their identified intimate partners (either spouses or individuals with whom the survivors have a committed relationship). Survivors were recruited from a urology clinic, a regional tumor registry, and from Durham, North Carolina and surrounding communities (e.g., Durham, Raleigh, and Chapel Hill).

Survivor Measures

Self-efficacy

Patients were assessed for self-efficacy using the Self-Efficacy for Symptom Control Inventory (SESCI). The SESCI is a modified version of the self-efficacy scale for chronic pain patients,45 and assessed the patients' confidence in their ability to manage prostate cancer-associated symptoms. The SESCI includes 3 subscales: self-efficacy for symptom management, which assesses confidence in the ability to manage symptom severity and interference; self-efficacy for physical function, which assesses confidence in the ability to engage in certain physical activities; and self-efficacy for coping, which assesses confidence in ability to cope with other symptoms related to prostate cancer, including frustration, fatigue, and sadness. These items combine to form a total self-efficacy score, which ranges from 30 to 300. Higher self-efficacy scores indicate greater confidence in the ability to control symptoms. Strong internal consistency for this measure with African American prostate cancer survivors was demonstrated, with Cronbach's alpha of 0.97 for the total self-efficacy scale and 0.94 for each of the self-efficacy subscales.

Quality of life

Disease-specific QOL was assessed using the Expanded Prostate Cancer Index Composite (EPIC).46 The EPIC is a 50-item scale that assesses QOL across 4 major symptom domains related to prostate cancer, including urinary, bowel, hormonal, and sexual. The symptom domain scores are standardized on a 0–100 scale, with higher scores reflecting better QOL. Each of the 4 symptom domains demonstrated good internal consistency with Cronbach's alphas ranging from 0.80 to 0.93.

General health QOL was measured using the Physical Function and Mental Health Scales of the Short Form-36 Health Survey (SF-36).47 The SF-36 was developed as an instrument to evaluate general health in various medical populations, and has been validated worldwide. Higher scores on the subscales reflect better functioning. The internal consistency of both the Physical Function and Mental Health scales were also strong in this sample. Alphas were 0.91 and 0.79, respectively.

Partner Measures

Self-efficacy

Partners were assessed for self-efficacy using the partner version of the SESCI. The partner version of the SESCI was identical to the patient version except that the partners were asked to rate their confidence in helping the patient cope with the various symptoms associated with prostate cancer. Cronbach's alpha for the total self-efficacy score was 0.95. Self-efficacy subscales for partners ranged from 0.87 to 0.91.

Quality of life

Partner QOL was assessed using the Profile of Mood States-Short Form (POMS-SF) and the Caregiver Strain Index (CSI). The POMS-SF measures multiple mood states, specifically for the previous 7 days, including depression, anxiety, anger, and fatigue. Higher scores on the POMS-SF indicated more distress. The POMS-SF has been shown to have high internal consistency, with Cronbach's alphas for the current study ranging from 0.82 to 0.90.48 The CSI was used to assess the amount of strain experienced by partners.49 The modified CSI used in this study included 10 items that assessed different types of strain ranging from emotional adjustment to time demands and physical strain. Higher scores on the CSI reflect more caregiver strain. The CSI demonstrated good internal consistency with Cronbach's alpha of 0.79.

CST Intervention

The partner-assisted CST protocol was designed to systematically train survivors and their partners in skills for managing symptoms experienced after treatment for prostate cancer. The protocol focused on 3 goals: 1) providing information about prostate cancer and possible long-term side effects, 2) teaching problem-solving skills, and 3) training in specific cognitive and behavioral coping skills (e.g., communication skills, relaxation training, activity pacing).

Survivors and their partners in the partner-assisted CST condition received 6 1-hour telephone sessions for training in coping skills. The dosage of the intervention (i.e., 6 sessions) was based on our effort to strike a balance between delivering comprehensive treatment and minimizing participant burden. Sessions took place approximately once a week. To facilitate communication between each couple and their interventionist, couples were given a speakerphone for use during the session phone calls.

The sessions were conducted by African American, doctoral level medical psychologists knowledgeable about prostate cancer and skilled in CST interventions. The interventionists underwent systematic training on the 6-session protocol, including role-playing to master the delivery of each skill via telephone without nonverbal cues. To ensure consistency in delivering the intervention, the psychologists followed a detailed written outline for each session, and met weekly for supervision. Sessions were audiotaped and reviewed by a senior-level psychologist who evaluated sessions for adherence to the protocol and provided ongoing supervision to maintain treatment integrity.

Session 1 began with a rationale for CST which emphasized the following: 1) prostate cancer affects the patient, partner, and their relationship, and therefore involving the partner in training can be helpful, and 2) survivors and their partners can enhance their skills for managing cancer symptoms and emotional distress. Each couple was asked to identify specific challenges and problems they are coping with and to describe their coping style in their own words. The coping skills taught in later sessions focused on the challenges identified in Session 1. These skills could be modified by the therapist, as needed, to fit the unique coping style and circumstances of each couple. Toward the end of Session 1, therapists assessed the couple's need for basic education about prostate cancer and began CST by introducing problem solving as the first coping skill. Each dyad was taught a three-step problem solving method, including 1) identifying the problem, 2) generating possible solutions to the problem, and 3) selecting the best solution and implementing it.

During Session 2, survivors and partners underwent progressive muscle relaxation (PMR) training, which involved increasing participants' awareness of muscle tension signals and teaching them to use these signals as cues to relax.50 In Session 3, survivors and their partners began training in communication skills, focusing on effective speaking and listening skills,51 and application of these skills to interpersonal situations. During Session 4, activity-rest cycling was used to teach survivors to pace their activities to manage fatigue, pain, or other symptoms.52 Survivors and partners identified and set goals for using activity-rest cycling to increase involvement in pleasant activities. In Session 5, cognitive-restructuring53 was used to teach the patient and partner to identify irrational, maladaptive thoughts and to replace these with alternative, rational coping thoughts. Participants were also taught pleasant imagery54, 55 as an enhancement to PMR. Finally, in Session 6, a coping maintenance plan for continued skills practice was developed with each couple.51

At the conclusion of the final session, the interventionist briefly interviewed each couple about their experience with CST. The following 3 questions were asked: 1) What was it like for you to participate in the study? 2) What was it like to undergo coping skills training by telephone? and 3) Is there anything you would add or change about the CST program? Interview responses were audiotaped and transcribed for qualitative analysis as described below.

Usual Care Condition

Survivors and partners in the usual care condition received the routine care provided through their medical outpatient program. Neither survivors nor their partners received training in the coping skills methods outlined above.

Procedures

Eligibility of survivors was confirmed by their physicians. All identified intimate partners or spouses were eligible for the study. Participants in the study were administered pretreatment measures before being randomly assigned to 1 of the 2 conditions: partner-assisted CST or usual care as described above. After completing the intervention period (6 weeks of CST for the treatment group or 6 weeks of usual care for the control group), participants in both conditions completed posttreatment measures. Both evaluations (pretreatment and posttreatment) and all CST sessions were completed by telephone. All participants were sent formal thank-you letters for their involvement in the study, and a summary of the findings of the investigation. Each couple was reimbursed $100 ($50 for each of the 2 evaluations).

Cultural Sensitivity

A number of additional cultural sensitivity strategies were employed to ensure that all research procedures, including recruitment, intervention, and dissemination of study results, were carried out in a culturally sensitive manner. First, study personnel made presentations in the community and sought continued input from health, mental health, and spiritual leaders in the African American community regarding recruitment strategies and project implementation. Second, the interventionists were African American clinicians knowledgeable about the ongoing legacy of distrust of medical institutions that exists among African Americans in all regions of the country.25 The interventionists were prepared to respond sensitively to concerns related to any aspect of research participation (e.g., the informed consent process, data storage, and confidentiality). Third, the intervention was delivered from a strengths-based perspective,56 which focuses on the strengths of each individual and the couple to maximize treatment effectiveness. Finally, at the conclusion of this study, the quantitative research findings were disseminated to research participants and community gatekeepers in the form of a brief summary, which included the study purpose, participants, results, and plans for future research.31

Data Analysis

Statistical analyses

Univariate analysis of covariance (ANCOVA) analyses were conducted to compare posttreatment differences between the treatment and control groups in self-efficacy (both survivors and partners), disease-specific QOL (survivors), general health QOL (survivors), caregiver strain (partners), and mood/fatigue (partners). For each ANCOVA, the individuals' pre-treatment scores were used as covariates in the model. Within-group treatment effects were also examined using paired samples t test. Cohen's d was calculated to estimate effect sizes for each outcome measure. Effect sizes of 0.2, 0.5, and 0.8 were interpreted as small, moderate, and large, respectively.57 Baseline analyses have been published elsewhere.15

Qualitative analyses

Transcripts of each posttreatment interview were analyzed using a qualitative data analysis program, (N658) to search and code the transcript data (i.e., group it by topic). Transcripts were analyzed at the level of the couple to capture a more complete description of the couple's experience with telephone-based CST. Data from each transcript was categorized into themes in 2 separate steps. In the first step, raw transcript data was reviewed, coded and organized into themes, utilizing the indexing and text search functions of the N6 software. The interventionist reviewed the raw data and, where needed, the data was annotated by adding words or phrases for clarity or context. No text was deleted, however, thus preserving the raw data. In the second step, data coding was reviewed by a psychologist who was not directly involved in the delivery of the CST protocol. This step provided an independent check on the validity of the data coding.

RESULTS

A total of 223 patients were screened for the study, and 157 patients were deemed eligible. Of the eligible patients, 45 patients expressed willingness to participate, and their partners also agreed to take part. Forty patients and their partners completed both the consent process and pretreatment evaluation and were randomly assigned to 1 of the 2 conditions: 1) partner-assisted CST or 2) usual care. Nine couples did not complete the intervention or did not provide posttreatment data. One couple reported extreme scores on most measures related to significant medical complications experienced by the survivor. Based on outlier analysis, scores from this couple were removed from further analyses. Results from the remaining 30 couples (12 treatment, 18 control) are presented here.

Survivors had a mean age of 62.1 years (SD 8.9) and partners had a mean age of 58.7 years (SD 9.78). Eighty-three percent of the couples were married. Eighty-six percent of the survivors and 93% of partners in the study had completed high school. Fifty-three percent of survivors were working either full- or part-time, and 37% were retired. Among partners, 46% were working full time and 46% were retired. Sixty-seven percent of couples reported an annual income of $30,000 or greater. The survivors varied with regard to stage of recovery, with time since treatment ranging from 1 week to 4 years (mean of 18 months; SD 14 months). The groups did not differ significantly with regard to age, marital status, education, occupational status, income level, treatment modality, or time since treatment. The demographic and clinical characteristics of participants in each of the study conditions are described in Table 1.

Table 1. Demographic and Clinical Characteristics of Sample

Patient variables	Coping skills training no. (%)	Usual care no. (%)
SD indicates standard deviation.
Mean age, y [SD]	60.9 [8.0]	62.9 [9.7]
Education
<High school graduate	11 (92)	3 (17)
≥High school graduate	1 (8)	15 (83)
Occupational status
Working	5 (42)	9 (50)
Not working	7 (58)	9 (50)
Income
<$10,000	1 (8)	0 (0)
$10,000 to $30,000	3 (25)	6 (33)
$30,001 to $100,000	5 (42)	9 (50)
>$100,000	1 (8)	4 (17)
Prostate treatment
Radical prostatectomy	11 (92)	17 (94)
Hormone injections	1 (8)	4 (17)
Hormone pills	1 (8)	1 (6)
Watchful waiting	11 (8)	1 (6)
External beam radiation	0 (0)	4 (17)
Time since treatment, m	17.8 [SD 15.7]	16.5 [SD 13.9]
Partner
Mean age, y [SD])	58.3 [8.8]	58.9 [10.7]
Education
<High school graduate	1 (8)	1 (6)
≥High school graduate	11 (92)	17 (94)
Occupational status
Working	6 (50)	8 (44)
Not working	6 (50)	8 (44)
Unknown	0 (0)	2 (12)

Feasibility Issues

Recruitment and retention

Participants were recruited from 3 sources: physician referral from a urology specialty clinic (22 patients), a regional tumor registry (11 patients), and self-referral from the community (7 patients). Twenty couples were randomized to the treatment group, and 15 of these couples completed treatment. Of the 15 couples that completed treatment, 2 provided only pretreatment data and 1 couple's data were excluded based on outlier analysis as noted above, resulting in pre- and posttreatment data from 12 couples. Examination of session attendance data indicated that, on average, the 6-session intervention was delivered within a 7-week period, indicating good adherence by study participants to the weekly treatment format. Of the 20 couples randomized to usual care, 2 couples provided only pretreatment data, resulting in pre- and posttreatment data from 18 couples.

While the number of couples who completed only the pre-treatment evaluation was identical across study conditions (2 couples in each condition), 5 additional couples in the treatment condition withdrew from the study after beginning treatment; 2 couples completed treatment, and 3 couples completed only the initial CST sessions. This differential attrition was of concern and required further exploration to determine whether these dropouts were random.59 Examination of demographic data and baseline measures did not reveal a pattern of systematic attrition. The couples varied with regard to socioeconomic status and occupational status. Survivors also varied in some aspects of treatment. All but one had undergone prostatectomy surgery, which reflects the high prevalence of that treatment modality in this sample (93%). They differed, however, with regard to additional therapies (e.g., external beam radiation, orchiectomy, hormone therapy). Overall, examination of baseline self-efficacy and QOL measures indicated variation across couples, and among both survivors and partners. Among couples that withdrew from the study, the most common reason was difficulty scheduling a time when both the survivor and partner were available.

Participant feedback on CST

A total of 15 couples provided interview data in the final CST session. As noted above, 2 of these couples did not provide posttreatment quantitative data, but their interview transcripts were included in the qualitative analyses. Qualitative analysis of interview data identified themes related to overall perceived benefit, usefulness of specific skills, telephone-based format, initial reservations about the CST intervention, and suggested changes to the CST intervention. Thirteen of 15 couples (87%) described the CST intervention as beneficial, noting both individual and couples benefits. Regarding the usefulness of specific skills, 9 of 15 couples (60%) found communication skills training to be especially valuable. With regard to telephone-based participation, 4 couples (27%) remarked upon the convenience of participating from their own homes, while 2 couples observed that the telephone format made it easier for survivors to be more vocal than usual, even about sensitive topics. Five couples (33%) expressed a preference for some degree of face-to-face contact in spite of the benefits of telephone-based participation. Interestingly, 4 couples (27%) openly shared their initial skepticism or reservations about the study. Finally, 4 couples (27%) recommended specific changes to the protocol, including adding videos to demonstrate skills such as PMR, and more time spent on communication and problem solving skills.

Group Differences: Survivors

Disease-specific QOL

The strongest treatment effect was observed in the bowel domain of the EPIC. Men who underwent CST reported significantly higher QOL in relation to bowel bother (i.e., less bowel bother; F (1,30) = 4.554, P = .042, ES = 0.471). The bowel function and bowel total scales did not differ significantly between the groups (Table 2).

Table 2. Patient Results

Variable	Coping skills training post, n = 12 Mean (SE) adjusted	Usual care post, n = 18 Mean (SE) adjusted	ANCOVA P	Effect size cohen d
SE indicates standard error, EPIC, Expanded Prostate Cancer Index Composite, SF-36, Short Form 36 Health Survey. * P < .05.
EPIC
Urinary function	72.39 (4.34)	75.57 (3.53)	.577	0.144
Urinary bother	82.09 (3.70)	74.24 (3.02)	.114	0.326
Urinary irritation	87.09 (2.69)	83.41 (2.19)	.300	0.254
Urinary incontinence	61.94 (6.38)	62.11 (5.20)	.984	0.000
Urinary total	77.98 (3.47)	74.84 (2.82)	.491	0.141
Bowel function	76.85 (3.12)	76.28 (2.53)	.891	0.004
Bowel bother	95.57 (2.46)	88.67 (1.99)	.042*	0.471
Bowel total	86.34 (2.48)	82.39 (2.00)	.236	0.311
Sexual function	30.35 (5.33)	23.44 (4.33)	.328	0.252
Sexual bother	42.92 (7.08)	28.68 (5.76)	.133	0.448
Sexual total	34.36 (5.22)	24.96 (4.25)	.177	0.344
Hormonal function	87.57 (3.20)	81.07 (2.60)	.131	0.380
Hormonal bother	89.79 (2.44)	86.67 (1.99)	.334	0.190
Hormonal total	88.79 (2.26)	84.17 (1.84)	.123	0.302
SF-36
Physical function	27.52 (.814)	26.10 (.664)	.192	0.341
Mental health	24.90 (.664)	25.24 (.540)	.699	0.008
Self-efficacy
Activity efficacy	91.39 (3.46)	86.05 (2.82)	.245	0.314
Coping efficacy	80.24 (3.18)	76.92 (2.59)	.428	0.134
Symptom management efficacy	80.81 (4.56)	79.80 (3.71)	.865	0.005
Total self-efficacy	252.23 (9.30)	242.91 (7.57)	.446	0.158

In some component of the urinary, sexual, and hormonal domains (subscale scores or total scores) of the EPIC, marginally significant treatment effects were found between the treatment and control groups. Effect sizes in or approaching the moderate range were observed for in subscale scores in these symptom domains. In the urinary domain, an effect size of 0.33 was observed for urinary bother. In the sexual domain, an effect size of 0.45 was observed for sexual bother. Finally, in the hormonal domain, an effect sizes of 0.39 observed for hormonal function (Table 2).

General health QOL and self-efficacy

No significant differences were found between the treatment and control groups on the physical functioning and mental health subscales of the SF-36 (effect sizes of 0.33 and 0.03, respectively) or on the self-efficacy subscales or total self-efficacy scores (effect size range 0.01–0.26; Table 2).

Group Differences: Partners

Negative mood

No statistically significant differences in anger, confusion, depression, fatigue, anxiety, or vigor were observed between partners who underwent CST versus usual care (effect size range 0.01–0.46). With regard to effect sizes, moderate effects were observed for depression, fatigue, and vigor, with partners who received CST reporting less depression and fatigue and more vigor than partners in the usual care condition. Respective effect sizes were 0.46, 0.39, and 0.40 (Table 3).

Table 3. Partner Results

Variable	Coping skills training post, n = 12 Mean (SE) adjusted	Usual care post, n = 18 Mean (SE) adjusted	ANCOVA P	Effect size cohen d
SE indicates standard error; POMS-SF, Profile of Mood States Short Form.
Caregiver strain	2.91 (.485)	3.73 (.395)	.209	0.265
POMS
Anger	2.16 (1.15)	2.42 (.975)	.869	0.005
Confusion	3.87 (.879)	4.96 (.752)	.357	0.259
Depression	1.85 (.891)	3.78 (.760)	.117	0.460
Fatigue	4.07 (1.22)	6.62 (1.04)	.128	0.392
Anxiety	3.95 (1.16)	5.30 (.989)	.385	0.293
Vigor	10.90 (1.04)	8.87 (.889)	.153	0.396
Self-efficacy
Activity efficacy	86.61 (3.52)	87.02 (2.86)	.930	0.031
Coping efficacy	75.82 (4.46)	77.72 (3.62)	.745	0.085
Symptom management efficacy	79.22 (3.79)	72.75 (3.09)	.200	0.301
Total self-efficacy	242.88 (10.08)	236.65 (8.20)	.639	0.121

Caregiver strain

There were no significant differences in caregiver strain between intimate partners who participated in CST and intimate partners who were involved in usual care. Both groups of partners reported low levels of caregiver strain at posttreatment (2.9[SD 0.485] in treatment group; 3.7[SD 0.395] in control group) with minimal variability. A effect size (0.27) was observed in caregiver strain scores, with intimate partners who participated in CST reporting lower levels of caregiver strain than partners in the usual care condition (Table 3).

Self-efficacy

There were no significant differences in self-efficacy between partners in the treatment and control groups (effect size range 0.03–0.30). A small-to-moderate effect size (0.30) was observed for the symptom management efficacy subscale, with partners in the CST condition reporting higher self-efficacy than partners in the usual care condition (Table 3).

Within-Group Treatment Effects

Paired-samples t test was conducted to evaluate pre- to posttreatment change among survivors within each of the study conditions. The results indicated that the survivors in the treatment group reported significant overall improvements in QOL related to bowel symptoms (t(11) = −2.57, P < .05), overall improvements in QOL related to hormonal symptoms (t(11) = −2.26, P < .05), and specific improvements in bowel bother (t(11) = −3.25, P < .01), which were not found in the control group. The treatment group also reported improvements in activity self-efficacy (t(11) = −2.40, P < .05) from pretreatment to posttreatment, whereas the control group did not. In addition to statistically significant improvement in the treatment group there were several trends in toward improvement. In symptom-related QOL, improvements in hormonal functioning, (t(11) = −2.01, P = .07), urinary bother (t(11) = −1.70, P = .12), and sexual bother (t(11) = −1.63, P = .13) were observed. Effect sizes were 0.49, 0.46, and 0.58, respectively. Among partners who underwent CST, there was a trend toward decreased fatigue (t(10) = 1.25, P = .13), with an effect size of 0.28, while among partners in the usual care condition trends toward increased confusion (t(14) = −1.71, P = .11) and anxiety (t(11) = −2.12, P = .052) were observed. Respective effect sizes were 0.44 and 0.54.

DISCUSSION

This pilot study explored the feasibility and efficacy of a culturally sensitive, telephone-based cognitive-behavioral intervention, CST, designed for African American prostate cancer survivors and their intimate partners. With regard to feasibility, telephone-based CST required no travel time and was delivered successfully to the majority of participants over 6 consecutive weeks. Participant feedback indicated that CST was beneficial. Of all the skills taught, communication skills appeared to be especially valuable to the majority of couples. Participants also noted the convenience of the telephone format, including its positive impact on communicativeness and self-disclosure, while at the same time retaining an appreciation for face-to-face interactions. Finally, when asked to recommend changes or additions to the program, participants suggested adding videos to the study materials and spending more time on certain skills, including communication and problem solving.

Regarding the efficacy of the intervention, these preliminary results indicate that CST shows promise as a way to improve posttreatment QOL. On the basis of both statistical analyses and effect size calculations, survivors reported improvement in QOL across several symptom domains, with larger effect sizes related to bowel and sexual symptoms and moderate effect sizes in urinary and hormonal symptom domains. The larger treatment effects observed in the bowel symptom domain is curious given that the majority of participants underwent surgical treatment for prostate cancer. Postsurgical bowel complications are believed to be less prevalent than bowel problems following radiation therapy;60 recent data, however, does suggest that prostate cancer survivors view bowel complications more negatively than other symptoms.61 The significance of these findings awaits further examination in future studies. Finally, consistent with previous literature noting the importance of using both disease-specific and more general health QOL measures when assessing QOL in prostate cancer survivors,46 the current study found larger treatment effect sizes on measures of disease-specific QOL, and much smaller effects on more global QOL measures.

CST also demonstrated potential for improving partner functioning, particularly in the areas of depression, fatigue, and vigor, where moderate effect sizes were observed. Caregiver strain was not only low at baseline, but remained low among partners in both study conditions at posttreatment, with little variability observed. Although this finding should not be interpreted as evidence that African American partners of cancer patients are immune from strain related to caregiving, this finding is consistent with previous research reporting less caregiver strain in African American caregivers compared with caregivers from other ethnic minority groups.62 This finding may also reflect the fact that survivors were capable of self-care per the inclusion criteria.44

Finally, although preliminary, this study makes a novel contribution to the literature by combining a number of important central features, including a focus on African American men, who have been understudied; the use of a telephone-based protocol; a comprehensive, theory-driven intervention; inclusion of intimate partners; and attention to cultural sensitivity at each stage of the research process, including dissemination of study findings.

Some study limitations should be noted. First, the sample size was quite small and some analyses were likely underpowered. By calculating effect sizes, we were able to capture improvements in both survivor and partner outcomes that were less evident when examining indices of conventional statistical significance. A larger sample, however, would improve statistical power allowing for more rigorous evaluation of the efficacy of CST intervention. Recruitment of larger samples will require a broader and more effective community-based recruitment network than was used in this study. Also, although we included a control group and randomized assignment to treatment and control conditions, we did not control for attention in this pilot study, which should be remedied in future studies. Attrition patterns also raised questions, although no clear pattern of systematic attrition emerged based on examination of demographic characteristics or baseline measures. Attrition related to features of the CST intervention (e.g., time commitment, conflicting schedules between survivors and partners) warrants further exploration. Study findings may have also been limited by the use of a CBT intervention that was briefer (6 sessions) than standard doses of cognitive-behavioral treatment (8–12 sessions).63, 64 On the basis of positive feedback from participants, it is quite possible that a slightly longer intervention consistent with standard CBT treatment would produce stronger treatment effects. Finally, given that the vast majority of survivors in this study underwent radical prostatectomy as their primary treatment for prostate cancer, these results may not generalize to men treated primarily using other modalities (e.g., radiation therapy).

Given our focus on facilitating research participation, some exclusion criteria deserve further scrutiny as possible limitations to study feasibility, specifically, excluding persons without a phone and excluding men without partners or whose partners were not willing to participate. Unfortunately, consistent phone service cannot be assumed for the most underserved individuals, who may be most in need of supportive care interventions. Likewise, single men or men without willing partners may be particularly in need of psychosocial interventions like CST as a source of support. These economic and social issues deserve further consideration to reach prostate cancer survivors who may be most in need.

In future research, we plan to replicate the positive features of the current study (e.g., randomized design, continued focus on African Americans survivors and partners, cultural sensitivity, and cost-effective telephone based protocol) while at the same time addressing the limitations noted above to reach a larger number of underserved African Americans coping with prostate cancer.

Acknowledgements

Supported by grant CA14236-2853 from the National Cancer Institute.