Helping patients with localized prostate carcinoma manage uncertainty and treatment side effects
Nurse-delivered psychoeducational intervention over the telephone
The objective of this study was to test the efficacy of an individualized uncertainty management intervention delivered by telephone to Caucasian and African-American men with localized prostate carcinoma and directed at managing the uncertainties of their disease and treatment.
The authors delivered a psychoeducational intervention by phone to men with prostate carcinoma, with or without supplemented delivery to a close family member, that was directed at managing uncertainty and improving symptom control. One hundred thirty-four Caucasian men and 105 African-American men were assigned randomly to one of two approaches to delivering the intervention or to the control condition. Men entered the study immediately after surgical treatment or in the first 3 weeks of radiation therapy. Trained nurses delivered the intervention through weekly phone calls for 8 weeks.
The authors found that the majority of intervention effects were from baseline to 4 months postbaseline, when treatment side effects are most intense. Both Caucasian men and African-American men who received either one of the two approaches for delivering the intervention improved in the two uncertainty management methods of cognitive reframing and problem solving. Similarly, when the intervention groups were combined, men who received the intervention also improved significantly in control of incontinence by 4 months postbaseline. Decreases in the number of treatment side effects differed by time and treatment/ ethnic group interactions as did satisfaction with sexual functioning.
This is one of the first tests of a psychoeducational intervention among men with prostate carcinoma and was the first test that included a sufficient number of African-American men to test by ethnic group. Therefore, replication of these findings is advised. Cancer 2002;94:1854–66. © 2002 American Cancer Society.
Prostate carcinoma is the second most common type of malignant disease in American men and is increasing in frequency, particularly among African-American men.1 Since the advent of prostate specific antigen (PSA) testing for prostate carcinoma screening, the number of men diagnosed with prostate carcinoma has increased.2 However, most of the psychoeducational interventions tested and reported in the literature to help patients manage their disease and treatment have been conducted among patients with breast carcinoma or among mixed groups of patients.3–5 Among the few psychoeducational interventions targeted to patients with prostate carcinoma, the strongest studies have been devoted to encouraging screening.6 Other studies that focus on the decision-making or symptom management aspects of the illness trajectory appear to be marked by design flaws,7 one treatment modality,8 small sample sizes, and noninclusion of minority men.9, 10
The lack of psychoeducational interventions for men with prostate carcinoma is surprising considering the life-altering nature of treatment. After undergoing either surgery or radiation therapy for prostate carcinoma, patients experience a variety of symptoms, including urinary incontinence, fatigue, bowel dysfunction, and impotence. All of these treatment-induced side effects have been associated significantly with depression, emotional distress, and changes in the person's lifestyle and relationships.11 For men with prostate carcinoma, the incontinence and impotence that follow most treatment modalities cause the most distress and significantly compromise quality of life.12–17 Whereas urinary control and sexual function are worse after surgery, bowel dysfunction is common after radiation therapy.18, 19 Patients report high rates of dysfunction after receiving treatment for early-stage prostate carcinoma and are surprised by the intensity and duration of treatment side effects, which play a major role in limiting physical functioning.20, 21 Changes in lifestyle may be necessary, and coping with the changes may require new types of information and the development of new skills.22
Men with prostate carcinoma must make decisions about treatment and management after treatment in the face of a great deal of uncertainty about the course of their disease, the effectiveness of treatment, and the impact of the disease on their future.23–25 Among men with early-stage prostate carcinoma, uncertainties about how to manage incontinence and how to gain control over urine flow are the immediate major concerns.26 These are followed closely by concerns about compromised potency. In patients with early-stage prostate carcinoma, treatment side effects persist as well as concerns about treatment efficacy, and patients may interpret any new symptom as a sign that the disease is returning.27
Like patients with other types of malignant disease, we have found that uncertainty is a significant predictor of emotional distress in men with prostate carcinoma.28, 29 A number of studies have found that continual uncertainties reduce quality of life, impact negatively on psychosocial adjustment, increase levels of depression, and increase emotional distress.30–34
Although uncertainty can be alleviated by information from health care providers, patients with prostate carcinoma have reported that they do not receive adequate help for the problems that result from the disease or from medical intervention.22 The intimate and embarrassing nature of the treatment side effects may make it difficult for older men to address these issues with medical staff.35, 36 Despite the predominance of this experience, little is known about whether psychoeducational interventions to promote skills in managing uncertainty among patients with prostate carcinoma can result in better informed patients with greater skill in managing treatment side effects and improved quality of life.
The purpose of this study was to test the efficacy of an individualized uncertainty management intervention delivered by telephone to Caucasian and African-American men with localized prostate carcinoma and directed at managing the uncertainties of their disease and treatment. We compared two approaches for delivering the intervention, either directly to the patient or supplemented by delivery to the patient and family member, with a control condition in which men received only usual care.
MATERIALS AND METHODS
The intervention tested here was based on the Theory of Uncertainty in Illness,37 which provides a framework for selecting interventions to improve psychological and behavioral outcomes. Uncertainty is defined as the inability to determine the meaning of illness-related events.37 It is assumed that uncertainty occurs when patients lack the information or knowledge needed to fully understand their illness, treatments, and treatment residuals. Uncertainty increases when the patient cannot use his education, social support, or relationship with health care providers to gain the needed information and knowledge. Uncertainty also is increased when a lack of information or understanding is heightened by events such as pain, illness, or medications.37 When uncertainty occurs, the patient cannot form a cognitive structure for illness-related events. The lack of cognitive structure can hamper decision-making and thereby inhibit performance. The end result is less problem-oriented coping, more emotion-oriented coping, and poorer psychosocial adjustment. Emotion-oriented coping that is not followed by problem-oriented coping is less effective in dealing with uncertainty, because it only addresses the emotional response to the uncertainty and does not provide the individual with the tools to manage it.
Uncertainty is a fluctuating experience and one that never totally disappears in illness. The goal is to manage rather than eliminate uncertainty. To do so, strategies that promote cognitive reframing and problem solving, provide skill training to improve patient-provider communication, and provide information on managing treatment side effects should supply the patient who has prostate carcinoma with skills to manage continual uncertainty.
Strategies for Managing Uncertainty in Patients with Prostate Carcinoma
At the time of diagnosis and treatment, patients' cognitive appraisals of their disease play an important part in how well they manage and adjust to the diagnosis.38 Under conditions of uncertainty that render situations vague, ill defined, and unclear, there are many opportunities for different types of appraisals.37 Without clear guidelines regarding the causes of prostate carcinoma or clarity about what the future may bring, men may experience self-blame for the disease and feelings of loss of control.37, 39 It has been found that patients with cancer distance themselves behaviorally from the disease due to the ambiguity of the outcome and uncontrollability of the disease.39 However, cognitive processing of illness-related events by talking about them with significant others, including health care professionals, can assist the patient with prostate carcinoma to reframe events into something manageable and less threatening.40 Also, patients can be taught how to form a cognitive schema from uncertain events and, within this schema, to change their view from loss of control to controllable.
Cognitive reframing occurs within the context of problem solving. When the patient with prostate carcinoma learns how to solve problems successfully through ambiguous events, these events are appraised as less threatening. Problem solving is one coping ability that may be made ineffective when uncertainty is present.41 It has been found that lower levels of problem solving may undermine quality of life.41 Problem solving can be strengthened through interventions aimed at teaching a problem-solving process. The steps of the problem-solving process have been identified as orienting the patient to the problem, defining the problem, generating alternatives, choosing among alternatives, and implementation.41 The steps in this process are likely to help dilute the uncertainty, because health professionals working with patients can bring in information to define the problem and add to possible alternatives. With each of these steps, the uncertainty is lessened, and the problems and solutions are likely to become clearer.
One major resource for uncertainty management is the information gained from health care providers. However, a common source of illness uncertainty is a lack of information about the nature, course, and prognosis of the disease.42 A number of investigations of communication between physicians and patients revealed communication problems that generated informational uncertainty. Uncertainty about disease status and disease containment appears to increase when the benefits from treatment are discussed in terms of statistics and probabilities, an approach commonly used by physicians when talking with newly diagnosed patients. However, patients prefer a fuller explanation.43, 44 Providing patients with information in a way that was meaningful to them reportedly resulted in significant benefits, including an enhanced sense of control, reduced anxiety and depression, improved communication with health care providers, and greater satisfaction with care.9, 45, 46
Becoming more knowledgeable about prostate carcinoma is another uncertainty management strategy. Gaining information on managing treatment side effects can help patients with incontinence. Specifically, patients who were given information about Kegel exercises were more capable of reducing incontinence after surgery,47, 48 therefore improving their quality of life.18, 19, 49 Psychoeducational interventions that have focused on providing information have been effective in improving knowledge of cancer among patients with a variety of malignant diseases. However, none of these studies addressed patients with prostate carcinoma.50
We evaluated an individualized psychoeducational intervention for men with prostate carcinoma that focused on the management of illness uncertainty. It consisted of a telephone-delivered uncertainty management intervention that was tested previously and was demonstrably effective in women with breast carcinoma.51 Several considerations went into the modification of the intervention. First, education about prostate carcinoma, managing side effects, and coping were emphasized, because it has been found that lack of knowledge and misinformation about prostate carcinoma are common among African-American men.52 Second, preference for an educational formant rather than a primarily affective format was selected, because it has been found that male patients with malignant disease prefer problem solving for managing problems.3, 8, 53, 54 Third, delivery of the intervention was modified by matching intervention nurses by ethnicity and gender to the patients. Gender matching was designed to assure that the sensitivity of the treatment side effects would be addressed by having patients talk with men. Matching by ethnicity was intended to address the lingering suspicions of African Americans about research participation based on memories of the Tuskegee trials and other trials that resulted in mistrust of the health care establishment.52, 55, 56
Fourth, the design of the study included two experimental arms. In one arm, the only the patient received the intervention (treatment direct group); in the other arm, both the patient and a family member received the intervention (treatment supplemented group). We asked patients to select a family member who was involved in helping them with their prostate carcinoma to join the study. Most of the family members selected were spouses. Family members were included because studies have shown that older men are reticent to discuss sexual concerns, and we hypothesized that enhancing the family member's ability to be a resource for the patient may enable men to disclose in sensitive areas.57, 58 Also, African-American women reportedly are more available and open in taking their problems to professional helpers, whereas African-American men appear to be more isolated and less available for professional help.59
The hypothesis derived from the uncertainty theory and tested in this study was that patients who received the uncertainty management intervention would demonstrate less uncertainty, improved management of uncertainty (through problem solving, cognitive reframing, increased knowledge, and improved patient-provider communication), and better control of treatment side effects (through decreased number of symptoms and intensity of symptoms, improved control over urine flow, satisfaction with sexual function, and ability to have an erection) compared with patients who received currently available care. We conducted further exploration to determine the relative effectiveness of treatment direct versus treatment supplemented intervention and possible interactions of treatment method with ethnicity for all major outcome variables.
African-American men and Caucasian men (n = 327 patients) who were diagnosed with localized prostate carcinoma and were within 2 weeks postcatheter removal after undergoing surgical treatment and/or were within 3 weeks into current radiation therapy were eligible for this study. Also, men who were eligible for the study had no major cognitive impairment, no concurrent treatment for another form of malignancy, had a telephone and an identifiable family member willing to participate in the study, and planned to reside in their current community for 12 months. Because access to a telephone was a criterion for participation, we documented that, among men at the poverty level in North Carolina, 97% of Caucasian men and 98% of African-American men age ≥ 60 years had telephones. In this study, only one patient was ineligible because he did not have a telephone. Of the 252 men enrolled, 95% completed measurements at all three times, with a drop-out rate of 5%. The final sample consisted of 239 men (134 Caucasian men and 105 African-American men). The 75 men who refused to participate in the study were similar to the study sample in age (mean, 66.7 years) and education level (mean, 12.5 years).
We recruited men from nine treatment facilities in central and eastern North Carolina. With permission from each patient's physician, recruiters introduced the study to patients during a clinic visit. We then mailed the patient a package that included a recruitment letter, study brochure, and recruitment video. Ten days after mailing the package, we made a follow-up call to determine the man's decision about participation and to set a time for the baseline data collection if they agreed to participate. For African-American men, we used a more elaborate recruitment process. Our community workers, two older African-American men who were well-known in the community, contacted African-American men to arrange a personal visit to further explain the study after the recruitment package had been sent. Recruitment and retention efforts for male African-American patients have been described elsewhere.60
Design and procedure
The design for the study was a 3 × 2 randomized block, repeated-measures design with three levels of the intervention (uncertainty management direct, uncertainty management supplemented, and control) crossed with two levels of ethnicity (Caucasian and African American). Patients were blocked on ethnicity and randomly assigned to one of the treatment groups or to the control group. Measurement occurred at three points in time: at entry into the study (baseline; T1), 4 months postbaseline (T2), and 7 months postbaseline (T3). The T2 measure was timed to occur about 8 weeks after the intervention was completed, allowing time for the patient to manage uncertainties without the aid of the nurse intervener, and served as a test of the efficiency of the intervention. The T3 data collection point was selected to provide information on the durability of the effect. All study variables were measured at each point except for demographic data, which were collected only at baseline. Data collectors who were not involved in the delivery of the intervention collected all data for patients and family members in their homes. Only patient data are reported in this article.
Uncertainty management intervention
Patients who received the uncertainty management intervention, either directly or supplemented, participated in a weekly telephone call for 8 consecutive weeks with a nurse who was trained in the intervention. We matched nurses with the patient and family member (in the treatment supplemented arm) by ethnicity and gender.
During each call, which was made at a time convenient for the patient and the nurse, the nurse first used a semistructured interview to assess the patient's concerns related to prostate carcinoma, uncertainty associated with the concerns, and the degree of threat posed by the uncertainty. Although the nurse followed a standardized list of possible problems (e.g., treatment side effects) in each encounter, he or she also encouraged the patient to express his individual specific concerns. Interventions for cognitive reframing included supporting, validating, and reinforcing the views and behaviors of the patient when he already had framed the problem as manageable. This helped to reinforce the existing cognitive orientation. However, when the patient was unable to reframe the situation and saw it as threatening, interventions included providing information about the nature of the problem, how to manage it, or what to expect. The nurse also informed the patient about resources that he could activate or helped him understand the likely trajectory of the problem. We identified markers or timelines that were meaningful for the individual to personalize the trajectory and to stress that improvement was ahead. Strategies for managing the problem were exemplified by such activities as teaching Kegel exercises to regain urinary continence,61 potency enhancement methods for impotence, or information on expressing intimacy in ways other than sexual intercourse. The nurses presented all of these strategies in the context that problems were manageable and that the patient had resources and skills for taking action to address problem.
Problem solving with the patient was used to generate options, look at alternatives, and consider resources. Nurses taught patients how to monitor a problem themselves so that they could find patterns. They encouraged patients to try out potential solutions and then report on the result at the next phone call. If the solution was successful, then the nurse encouraged the patient to think of other disease- related problems in which the same solution may be applied. If the outcome required revision, then this was explored in the next intervention call.
Strengthening patient-provider communication occurred through providing information so that the patient was knowledgeable about his problems. The nurse also trained the patient in how to ask questions and generate a list of questions for the next physician visit. To promote self-advocacy, the intervention nurse taught the patient assertive communication skills to enhance the patient's participation in planning his care.
These interventions were applied to a variety of concerns, including those about diagnosis, treatment, response to treatment, living with cancer, caring for oneself, and social/lifestyle issues. The intervention protocol designated the range of interventions, but the application of interventions was targeted to the specific concerns and needs of each patient at each phone call. We mailed printed materials, audiotapes, and videotapes for managing specific problems to patients after each weekly call. Any materials that were sent to the patient were reviewed with him at the next phone call.
Patients in each of the two treatment groups (direct or supplemented) received the same intervention. In the supplemented arm, the spouse or designated family support person also received a weekly phone call for 8 weeks from a nurse who was matched by ethnicity and gender and who conducted a similar assessment of the family members' concerns about the patient using a list similar to that used for the patients.
Men who were assigned randomly to the control group received their usual care. During data collection, printed general health information was given to patients to control for the nonspecific effects of the printed information materials sent to intervention patients by the nurse intervener. None of the health information was related to prostate carcinoma or the side effects from treatment to prevent the information from having the effect of alternative intervention. We offered men in the control group four intervention calls after completing the final data collection.
Demographic and Background Measures
The Mini-Mental State Examination (MMSE)62 was used to assess the cognitive status of potential patients during the recruitment process. The patient's graded response on the MMSE reflects parameters of orientation, registration, short-term memory, attention and calculation, constructional capacities, and the use of language. A MMSE score is produced by summing the points assigned to each successfully completed task for a total score of 0–30. Test-retest reliability for different periods and different patient populations range from 0.85 to 0.98.62, 63 Concurrent validity has been obtained by correlating scores on the MMSE with the Wechsler Adult Intelligence Scale; the Pearson correlation coefficient (r) was 0.77 for verbal intelligence quotient (IQ) and 0.66 for performance IQ.62 Whereas the specificity for the MMSE is lower for patients who are older or who have a low educational level,63, 64 other tests of cognitive function have similar limitations as well as other limitations.65
The following demographic and background information was obtained at baseline from patients: current age, number of years of formal education, marital status, monthly family income, and ethnicity. Medical information, including date of diagnosis (from pathology report), TNM staging, types and amounts of treatment, and names of treating physicians and institutions, was obtained with patient permission from the patient's medical record.
Comorbid health problems were measured by an investigator-developed list of 25 common, age-relevant health problems of men. Respondents answered yes or no to each item, and the positive responses were summed to obtain the total number of health problems.
Uncertainty was measured by a shortened, 26-item version of the Mishel Uncertainty in Illness Scale.66 The item response format was a Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). Items were scored in the direction of higher scores indicating more uncertainty (sample item: I do not know whether my treatment will work). The reliability of the shorter version in this sample was an α of 0.86 for the total sample (Caucasian men, 0.88; African-American men, 0.82). The validity of the shortened version was supported by positive correlation with emotional distress in African-American men with prostate carcinoma.29
Uncertainty management was measured by using four instruments: problem solving and cognitive reframing subscales of the Self-Control Scale67 and investigator-developed measures of prostate carcinoma knowledge and patient-provider communication. Problem solving is defined as the ability to identify and define concerns and generate solutions. The internal consistency of the 10-item problem solving subscale with this sample was an α of 0.90 (Caucasian men, 0.90; African-American men, 0.86). The scale items are in a visual analogue format ranging from 1 (not true about me) to 10 (true about me), with higher scores indicating more problem solving (sample item: when I have to do something that makes me worry, I try to think how I can handle my worry). Cognitive reframing is defined as the ability to view concerns as manageable. The nine-item cognitive reframing scale had an α of 0.90 with the total sample and 0.89 for each subsample. The format for the response choice was identical to the format for the problem solving scale (sample item: often by changing my way of thinking, I am able to change my feelings about almost anything). The validity of both subscales was supported by their correlation with uncertainty in the predicted direction.68 Scale scores were the mean of the items.
The Cancer Knowledge Scale is an investigator-developed list of 21 statements about prostate carcinoma and its treatment. Respondents used a true/false/don't know response format. A patient's score was the sum of the number of correct items (sample item: untreated or undetected prostate carcinoma can spread to the bones). The KR20 reliability of the Cancer Knowledge Scale for all prostate carcinoma patients was 0.67 (Caucasian men, 0.64; African-American men, 0.73).
Patient-provider communication was measured by an investigator-developed scale containing five items that asked the patient about the degree to which he had communicated with his health care provider. The five questions were as follows: How much does the patient participate in planning his care? How much does the patient tell the doctor? How much does the patient tell the nurse? How much information does the doctor tell the patient about his illness and how it will be treated? How much information does the nurse tell the patient about his illness and how it will be treated? All five items were reported on a 1 to 5 scale ranging from a low of nothing at all to a high of a great deal, with higher scores reflecting a greater amount of communication. Factor analysis of the patient-provider communication items resulted in one factor with an Eigen value > 1 and all items loading at 0.50 or above. Thus. a mean total score was used. The α for the total sample was 0.75 (Caucasian men, 0.74; African-American men, 0.75).
Finally, the management of treatment side effects was assessed with the Symptom Distress Scale,69 as modified for incorporation in the Southwest Oncology Group Quality of Life Questionnaire. For the current study, the 15-item Symptom Distress Scale was used to index the degree of problems the patient had with treatment side effects.70, 71 Symptom distress yielded two scores: the total number of symptoms reported and the average intensity of the symptoms reported on a 1 to 5 scale. Scores were reversed so that higher scores reflected more improvement in the symptom. The internal consistency of the intensity score for this sample was 0.70 (Caucasian men, 0.68; African-American men, 0.74). Three individual symptoms also were examined separately: control over urine flow, which refers to incontinence; ability to have a normal erection; and satisfaction with sexual functioning since surgery or radiation. All three items were reported on a 1 to 5 scale, with higher scores reflecting more success in dealing with the symptom or more satisfaction with the outcome.
We conducted repeated-measures multivariate analyses of variance (MANOVAs) to determine uncertainty management scores (cognitive reframing, problem solving, cancer knowledge, and patient-provider communication) and symptom management scores (number of symptoms, symptom intensity, control over urine flow, ability to have a normal erection, and satisfaction with sexual functioning) as a function of treatment group, ethnic group, and treatment by ethnic interaction effects. When the MANOVA revealed a significant difference, follow-up repeated-measures analyses with planned contrasts to test the hypotheses were performed for the individual variables. To look at patterns of change over time, planned contrasts between T1 and T2 and between T2 and T3 were constructed to test for both initial impact and delayed impact of the intervention. Contrasts for these time differences also were constructed and tested for within ethnic groups to determine whether differences existed for only one ethnic group.
Given the few studies of patients with this type of malignancy, particularly African-American patients, any noticeable statistical trend in the desirable direction from the intervention is valuable; therefore, the significance level used in the analysis was < 0.10. These findings are useful for an initial understanding of treatment efficacy and for providing direction for further studies. Although it is understood that these findings will require further investigation, the risk of rejecting important research hypotheses was judged more important than the risk of a Type I error. Probability values (P) > 0.05 but < 0.10 are reported here as trends.
Characteristics of the Study Population
The average age for the study sample was 64.0 ± 6.9 years (± standard deviation). Thirteen percent of patients had < 8 years of formal education, 30% of patients had 9–12 years of education, and 57% of patients had > 12 years of education. Because many of the men in the study grew up in rural towns where time was lost from school due to harvesting crops, the calculated of years of education may not be accurate. The monthly income of the sample was diverse, with 17% of patients reporting a monthly income < $1000, 20% of patients reporting a monthly income of $1000–2000, 18% of patients reporting a monthly income of $2000–3000, and 45% of patients (the largest group) reporting a monthly income > $3000. The study participants were divided almost evenly by ethnicity, with 56% Caucasian men and 44% African-American men. Most of the patients were married (84%). Utilizing the tumor-lymph node-metastases (TNM) staging classification system, 8% of patients had stage T1 tumors, 61% of patients had stage T2 tumors, 27% of patients had stage T3 tumors, and 4% of patients did not have staging information available. Most patients underwent surgery, with 56% of patients undergoing radical prostatectomy, 23% of patients undergoing nerve-sparing surgery, and 21% of patients receiving radiation therapy.
The intervention and control groups were compared on demographic and medical data collected at baseline to check the randomization procedure. They differed significantly on marital status (chi-square test, 18.26; P = 0.001), tumor stage (Fisher exact test; P = 0.007), and the number of health problems (F2, 236 = 3.76; P = 0.02). The treatment supplemented group had a greater proportion of married patients compared with the treatment direct group or the control group. However, because eligible participants were required to have a family member, differences in marital status should not have biased the intervention effect. With regard to tumor stage, the control group had slightly more patients with lower stage tumors, which should have made it harder to find a significant intervention effect. The one difference that may have biased the intervention effect was the number of health problems. Therefore, we used health problems as a covariate in all analyses.
Using the data from the intervention data base, we identified the most frequent problems expressed by the men. Four problems occurred most often for both Caucasian and African-American men. The four problems, in order of descending frequency, were leaking urine, erection problems, communication with health care providers, and general treatment side effects.
Uncertainty and Uncertainty Management
The least-square mean values of outcome measures (adjusting for health problems) for the treatment groups at baseline, at 4 months, and at 7 months are displayed in Table 1. There was a significant decline over time in uncertainty, but this change did not differ by treatment group, ethnicity, or treatment group by ethnicity. However, the MANOVA for uncertainty management variables indicated that there were significant differences over time among the groups (Wilks λ F16, 438 = 1.96; P = 0.01). Follow-up analyses revealed that scores for cognitive reframing and problem solving were significantly different between the groups. There was a significant difference over time by treatment group for cognitive reframing (Wilks λ F4, 456 = 3.81; P = 0.005). Preplanned contrasts showed that there was a significant change for the combined treatment groups from baseline to 4 months (F1, 229 = 6.84; P = 0.009). The combined treatment group had an increase in cognitive reframing, whereas the control group declined from baseline to 4 months; however, the effect did not hold at the 7-month follow-up. Seven months from baseline, the control group had improved significantly in cognitive reframing by regaining their initial loss and returning to their baseline level (F1, 229 =10.13; P = 0.002). With some decline in the intervention groups, all patients were approximately equal by 7 months postbaseline.
Table 1. Adjusted Means and Standard Deviations for Dependent Variables by Treatment Groups
|Uncertainty||2.34 (0.45)||2.22 (0.45)||2.20 (0.45)||2.25 (0.48)||2.15 (0.48)||2.16 (0.48)||2.37 (0.47)||2.27 (0.47)||2.21 (0.37)|
|Uncertainty management strategies|
| Cognitive reframing||7.70 (1.37)||8.20 (1.37)||8.04 (1.37)||7.91 (1.45)||8.00 (1.45)||8.00 (1.45)||7.83 (1.40)||7.60 (1.40)||8.02 (1.31)|
| Problem solving||8.02 (1.28)||8.32 (1.28)||8.18 (1.28)||8.00 (1.37)||8.22 (1.37)||8.10 (1.37)||8.11 (1.31)||7.80 (1.31)||8.07 (1.31)|
| Cancer knowledge||14.53 (3.12)||15.53 (2.84)||16.27 (2.75)||14.37 (3.39)||5.45 (3.06)||16.33 (2.90)||14.40 (3.17)||14.76 (2.89)||15.66 (2.70)|
| Patient-provider communication||3.46 (0.90)||3.62 (1.00)||3.60 (0.81)||3.76 (1.03)||3.62 (1.11)||3.57 (0.87)||3.55 (0.93)||3.71 (1.03)||3.61 (0.84)|
| No. of symptoms||7.72 (2.80)||6.15 (2.80)||5.78 (2.38)||7.42 (2.98)||5.69 (2.90)||5.43 (2.50)||7.57 (2.70)||6.45 (2.70)||5.99 (2.33)|
| Symptom intensity||1.86 (0.37)||1.65 (0.27)||1.58 (0.27)||1.82 (0.40)||1.54 (0.32)||1.50 (0.24)||1.84 (0.37)||1.68 (0.28)||1.61 (0.28)|
| Control over urine flow||3.64 (1.16)||4.52 (0.71)||4.56 (0.71)||3.59 (1.19)||4.59 (0.79)||4.73 (0.79)||3.88 (1.15)||4.41 (0.71)||4.51 (0.71)|
| Ability to have an erection||1.87 (1.19)||2.41 (1.01)||2.48 (1.01)||1.67 (1.26)||2.41 (1.18)||2.65 (1.10)||1.85 (1.18)||2.39 (1.09)||2.44 (1.00)|
| Satisfaction with sexual function||2.15 (1.35)||2.22 (1.26)||2.31 (1.26)||1.62 (1.45)||2.43 (1.39)||2.17 (1.32)||2.18 (1.38)||2.19 (1.29)||2.55 (1.29)|
Problem solving showed similar results, with a significant difference over time between the treatment groups and the control group (Wilks λ F4, 456 = 2.40; P = 0.049) and with significant differences between values at baseline and at 4 months (F1, 229 = 8.04; P = 0.005). The treatment groups showed an increase in problem solving from baseline to 4 months, whereas the control group lost ground in their ability to problem solve over the same interval. By 7 months, the treatment groups began to decline in problem solving but stayed above their baseline level, whereas the control group regained substantially, approaching their initial level of problem solving by 7 months postmedical intervention (F1, 229 = 6.38; P = 0.012). We found no significant differences between the treatment groups and the control group for the patient-provider communication scale or for cancer knowledge.
The MANOVA for symptom management indicated that there were significant differences over time between the control group and the intervention groups (Wilks λ F10, 165 = 2.25; P = 0.02). Follow-up analyses revealed that there was a significant decrease in the number of symptoms over time for all groups (Wilks λ F2, 228 = 26.82; P = 0.001). From the preplanned contrast, there was a significant decrease from baseline to 4 months in the mean number of symptoms reported for Caucasian men in the treatment supplemented group compared with Caucasian men in the control group (change scores of 2.21 vs. 1.15, respectively; F1, 228 = 5.53; P = 0.02). There were no significant differences between these groups from 4 months to 7 months. However, there was a significant decrease in the mean number of symptoms reported for African-American men in the treatment direct group from 4 months to 7 months compared with the control group (change scores of 0.96 vs. 0.02, respectively; F1, 229 = 3.96; P = 0.047). The reported intensity of the symptoms decreased over time for all groups (F2, 228 = 41.12; P = 0.001) but did not differ by intervention group.
According to specific items related to prostate carcinoma in the Symptom Distress Scale, men in the intervention groups showed a trend toward improvement in control over urine flow (Wilks λ F4, 424 = 2.19; P = 0.069). Because of the importance of this variable in patients with prostate carcinoma, we proceeded with the preplanned contrast. Significant improvement in control over urine flow was reported for patients in the combined treatment direct and treatment supplemented groups compared with the control group (Wilks λ F2, 212 = 3.71; P = 0.03). This improvement occurred between baseline and 4 months (F1, 213 = 7.05; P = 0.01). Improvement continued between 4 months and 7 months, but the difference from the control group was not significant.
There was a significant difference over time in satisfaction with sexual functioning among the groups (Wilks λ F4, 370 = 2.88; P = 0.02). Preplanned contrasts showed significant improvement for African-American men in the treatment supplemented group compared with the control group (1.14 vs. −0.03, respectively) from baseline to 4 months (F1, 186 = 6.57; P = 0.01). However, the mean rate of satisfaction with sexual functioning did not continue to improve from 4 months to 7 months. African-American men in the supplemented treatment group changed in their level of sexual satisfaction below their counterparts in the treatment direct group (−0.49 vs. 0.30, respectively; F1, 186 = 4.15; P = 0.04) and in the control group (−0.49 vs. 0.45, respectively; F1, 186 = 5.64; P = 0.02).
For the patient's ability to have an erection, there was no main effect from the intervention, but there was an overall trend toward an interaction effect for African-American men in the treatment supplemented group compared with the control group (Wilks λ F2, 220 =2.38; P = 0.09). However, we proceeded with the preplanned contrasts, because it was important to have some information on how minority men may respond to the intervention on this important outcome variable. Preplanned contrasts showed a trend toward improvement between 4 months and 7 months, when African-American men in the treatment supplemented group reported a higher level of erectile functioning compared with the men in the control group, although the difference was not significant (F1, 221 = 3.01; P = 0.08).
A key finding of this study was that Caucasian and African-American men who received the intervention either directly or supplemented by their family member had significant improvement in two uncertainty management methods, cognitive reframing and problem solving, at 4-month follow-up. These early months after medical treatment are the most distressful in terms of worsening of incontinence and severe distress about impotence.11, 72 Delivering a psychoeducational intervention at this time is most opportune, because patients with prostate carcinoma need to learn new skills for coping with the disruption in their lives and for managing the extreme treatment side effects. The ability of men receiving the intervention to explore their disease-related concerns with a receptive nurse and learn new skills appears to have facilitated cognitive processing of their experience during the most intense time for treatment side effects.40
Cognitive reframing reportedly is a common coping strategy among cancer patients.73 Lepore and Helegson40 identified cognitive processing as an important activity that helps men with prostate carcinoma reinterpret traumatic events into manageable events and integrate uncertain aspects of their experience into a coherent cognitive state during the time of most distress. Consistent with the uncertainty in illness theory37 and recent work by Pennebaker74 expressing one's experience to a trusted individual places some structure on the experience, thereby facilitating cognitive restructuring of the event. Through sharing over time, the men with prostate carcinoma began to organize and structure their treatment experience so that events became more manageable. The intervention strategies of validation and affirmation of men's attributions and explanation for their situation seem to have facilitated this process. This approach for facilitating cognitive processing of traumatic events has been supported by previous research.75
Problem solving also improved from baseline to 4 months postbaseline among men who received the intervention. Problem solving is an active coping strategy that contrasts with distancing, the cognitively oriented, passive strategy reportedly used by cancer patients to manage uncertainty and ambiguity.39 Problem solving skills were enhanced by the resources provided to patients in the intervention. We sent a list of community specific resources to every man who received the intervention along with a number of publications, web sites, and other mediated materials relevant to men's specific concerns. The final component was teaching men how to use these materials to address their concerns.
Because problem solving is based on the appraisal of a situation, it is likely to occur when cognitive reframing is operative in viewing the problem as manageable. These two strategies, one cognitive and one behavioral, both were necessary to manage uncertainty successfully.
The lack of intervention effect on cognitive reframing and problem solving between 4 months and 7 months is puzzling. It is possible that the effect is less durable in men after the intervention stops. Conversely, cognitive restructuring may not be as necessary at 7 months, when the intensity or distress from side effects begins to abate. However, problem solving is a skill that can be used as long as the side effects continue. The intervention seems to have a short-term benefit and may need to incorporate a booster session to reinforce skill learning over time.
Throughout the 7-month follow-up, no evidence was found to establish a benefit from the intervention for either patient-provider communication or cancer knowledge, although it was found that concerns about communicating with the physician and lack of information about cancer were two of the major problems. Providing information and access to resources and training for communication skills were frequent interventions; however, they did not seem to have an impact on these two variables. Although the intervention included specific methods for training the patient how to ask questions and prepare for the doctor visit, more attention to these strategies may strengthen the intervention.
We had hypothesized that patients who received the intervention would be more capable of managing their treatment side effects compared with patients in control group. When the treatment groups were combined, there was support for an intervention effect on self-reported control over urine flow from baseline to 4 months postbaseline. Due to the life-changing effect of surgically induced incontinence, we used the specific technique of teaching Kegel exercises to men over the telephone along with weekly reinforcement for progress and modification of the protocol to fit into the man's lifestyle and activity level.
Traditionally, patients are offered very little to handle this problem except for pharmacologic or surgical therapy.76 Behavioral methods for treatment have been overlooked until very recently. Mathewson-Chapman48 reported the use of biofeedback and pelvic muscle exercise as a successful approach to managing incontinence. Van Kampen et al.77 likewise found that improvement in pelvic floor reeducation in men after radical prostatectomy was most evident during the first 4 months after surgical intervention. After that time, the rate of improvement began to level out. The consistency of those findings and ours is noteworthy, in that our intervention strategy accomplished similar results by training over the telephone compared with laboratory-based treatment with exercise and biofeedback. Our approach helped the men build the practice of Kegel exercises into their life routine so that it could become habitual behavior.
Because incontinence is associated with skin breakdown and irritation, which can become chronic problems, there are medical benefits from early control.12, 14, 21, 49 The psychological effects of incontinence include loss of dignity and feelings of powerlessness and isolation due to the stigma attached to incontinence.47 Early management helps the patient regain some control of his life and avoid these negative psychological states. Our procedure is transferred easily to the clinic setting; however, it does require that weekly monitoring of patient progress, encouragement, and modification of practice to fit into life routines. Hopefully, instruction on pelvic muscle strategies will become a standard component of posttreatment care.
The impact of impotence was modified for some of the men who received the intervention. African-American men in the treatment supplemented group showed significant improvement in their satisfaction with sexual functioning and some improvement in their ability to have an erection. However, the time frame differed, with improvement in satisfaction with sexual functioning occurring from baseline to 4 months and a trend toward better erectile ability occurring from 4 months to 7 months. Because this study was conducted before the advent of Viagra™, men in the treatment supplemented group were introduced, along with their spouses, to a variety of mechanical devices to promote erection. Many Caucasian men and their spouses in the treatment supplemented group reported that they had tried these approaches, but spouses complained about the mechanical nature of the methods, so they were abandoned. However, many of the African-American couples were less aware of these aids until the nurse intervener suggested them, and they expressed a commitment to try these approaches. Active involvement in trying to improve sexual performance by using erectile aids reportedly improved patients' rating of satisfaction with sexual functioning.11, 78 However, over time, due to the artificiality of penile self-injection and vacuum devices, these men and their spouses reported that they found these methods distasteful. Nevertheless, the use of these methods can improve a man's ability to have an erection13 and may account for the trend toward an improvement in erectile function from 4 months to 7 months postbaseline.
All men had a decrease in the number of symptoms over time. In the first 4 months of the study, Caucasian men in the treatment supplemented group reported greater reductions compared with patients in the control group. African-American men in the treatment supplemented group also reported a decline in the number of symptoms from baseline to 4 months, but their decline was not significantly different compared with the decline in the control group. However, men who received the intervention commented that one of its most meaningful aspects was that it provided a means for patients to get questions answered about treatment side effects and information about how to manage those side effects. The early benefit can be attributed to the encouragement and support offered by family members who also received the intervention. The patient and family member could work collaboratively in using the suggestions from the nurse. At each telephone call, the nurse encouraged the spouse to discuss the patient's symptoms and then either offered information on how to help him manage them or encouraged the spouse to problem solve with the nurse to identify management methods.
From 4 months to 7 months, African-American men in the treatment direct group showed the greatest benefit of reduced numbers of symptoms. This pattern suggests that it takes longer for men to use their intervention skills to manage symptoms when their support person does not receive the intervention. Because the goal of reducing the number of symptoms is not as focused as managing incontinence or impotence, a greater number of strategies would be considered during the intervention. It is likely that the man without a family member involved in the intervention may need more time to learn and practice these strategies. Eventually, by 7 months postbaseline, the benefit from the intervention was evident.
Although this intervention was effective in reducing the number of problems, it did not affect the intensity of symptoms. This is consistent with the results of previous studies of treatment side effects that showed a lessening of severity over time.20, 21
Further research should improve on several limitations of this investigation. First, few intervention effects were found to hold over time, which may indicate that the intervention needs to be longer than 8 weeks or that a booster session is necessary. Researchers should try scheduling such an intervention with a booster session to determine whether the treatment effect will then hold over time. Also, our matching of study participants with intervening nurse by gender and race may limit generalizabilty of the findings. Future investigators should vary the gender and/or race of the intervener, especially for African-American men and women, to heighten generalizability and improve applicability to the current distribution of health care professionals.
Uncertainty is a common experience for cancer patients, and the current study found that a psychoeducational, telephone-delivered intervention directed at helping men with prostate carcinoma manage their uncertainties had identifiable benefits, which included learning new cognitive and behavioral skills during the time of highest symptom distress. Providing a resource to patients after initial treatment is particularly relevant, because follow-up care has decreased markedly in the current American health care system. Many intervention patients commented that they appreciated the information they received. The predominant comments were that they did not feel alone and that they knew a health professional was concerned about their welfare.
Because few African Americans have been included in prior studies of psychoeducational interventions, the current findings are particularly important, because they support both treatment benefits shared with the Caucasian sample and treatment benefits that are specific to minority men. Because this is one of the first tests of a psychoeducational intervention among patients with prostate carcinoma and was the first test that included a sufficient number of African-American men to test by ethnic group, replication of these preliminary findings with similarly diverse samples is strongly encouraged.
The authors acknowledge the following individuals: Dr. Lorna Harris for her contribution to minority recruitment; Drs. David Paulson, Paul Godley, and Mitchell Anscher for facilitating patient recruitment; Andrea Ware and Sheron Sumner, project managers; and Robert Ahmed, Kay Gruninger, Richard Kelsey, Joe Sanders, Lynn Scallion, Sheila Tedder, James Trogdon, and Howard Williams, nurse interventionists.