Aug 3, 2005 - South Florida, Tampa, Florida. 3 Genitourinary Oncology Program, H. Lee Moffitt. Cancer Center and Research Institute, University of.
1381
Treatment Decision-Making Strategies and Influences in Patients with Localized Prostate Carcinoma Clement K. Gwede, Ph.D., M.P.H., R.N.1–3 Julio Pow-Sang, M.D.3 John Seigne, M.D.3 Randy Heysek, M.D.2,3 Mohamed Helal, M.D.2,3 Kristin Shade, B.S.4 Alan Cantor, Ph.D.5 Paul B. Jacobsen, Ph.D.1,6 1
Health Outcomes and Behavior Program, H. Lee Moffitt Cancer Center and Research Institute, University of South Florida, Tampa, Florida.
2
Radiation Oncology Program, H. Lee Moffitt Cancer Center and Research Institute, University of South Florida, Tampa, Florida.
3
Genitourinary Oncology Program, H. Lee Moffitt Cancer Center and Research Institute, University of South Florida, Tampa, Florida.
4
Clinical Research Operations, H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida.
5
Biostatistical Core, H. Lee Moffitt Cancer Center and Research Institute, University of South Florida, Tampa, Florida.
6
Psychosocial and Palliative Care Program, H. Lee Moffitt Cancer Center and Research Institute, University of South Florida, Tampa, Florida.
BACKGROUND. Patients diagnosed with localized prostate carcinoma need to interpret complicated medical information to make an informed treatment selection from among treatments that have comparable efficacy but differing side effects. The authors reported initial results for treatment decision-making strategies among men receiving definitive treatment for localized prostate carcinoma. METHODS. One hundred nineteen men treated with radical prostatectomy (44%) or brachytherapy (56%) consented to participate. Guided by a cognitive-affective theoretic framework, the authors assessed differences in decision-making strategies, and treatment and disease-relevant beliefs and affects, in addition to demographic and clinical variables.
RESULTS. Approximately half of patients reported difficulty (49%) and distress (45%) while making treatment decisions, but no regrets (74%) regarding the treatment choice they made. Patients who underwent prostatectomy were younger, were more likely to be employed, had worse tumor grade, and had a shorter time since diagnosis (P ⬍ 0.01) compared with patients who did not undergo prostatectomy. In multivariate analyses, compared with patients who received radical prostatectomy, patients who received brachytherapy were more likely to say that they chose this treatment because it was “the least invasive” and they “wanted to avoid surgery” (P ⬍ 0.0001). CONCLUSIONS. In general, patients who received brachytherapy chose this treatment because of quality of life considerations, whereas “cure” and complete removal of the tumor were the main motivations for patients selecting radical prostatectomy. Long-term data are needed to evaluate distress and decisional regret as patients experience treatment-related chronic side effects and efficacy outcomes. Decision-making aids or other interventions to reduce decisional difficulty and emotional distress during decision making were indicated. Cancer 2005; 104:1381–90. © 2005 American Cancer Society.
KEYWORDS: prostate carcinoma, treatment decision making, informed decision making, shared decision making, medical decision making.
P
Supported in part by a research grant from the American Cancer Society Institutional Research Grant (ACS-IRG 93-032-07). Address for reprints: Clement K. Gwede, H. Lee Moffitt Cancer Center and Research Institute, 12902 Magnolia Drive, MOD 9-RT, Tampa, FL 33612; Fax: (813) 979-7231; E-mail: gwede@moffitt.usf.edu Received December 17, 2004; revision received April 4, 2005; accepted May 11, 2005.
rostate carcinoma is the most common form of solid tumor (excluding skin carcinoma) in American men, with an estimated 232,090 new cases and approximately 30,350 deaths expected in 2005.1 These statistics constitute 33% of new cancer cases and 10% of all cancer deaths in men. The disease is generally diagnosed early (75% of all new cases), when disease is locally confined to the prostate gland, without regional involvement of pelvic lymph nodes or distant metastases to abdominal lymph nodes or other organs beyond the prostate gland. Patients with early-stage disease can choose from several treatment options. These include surgery (radical prostatectomy), radioactive permanent prostate seed implants alone (brachytherapy), external-beam pelvic radiotherapy (EBRT) alone or in com-
© 2005 American Cancer Society DOI 10.1002/cncr.21330 Published online 3 August 2005 in Wiley InterScience (www.interscience.wiley.com).
1382
CANCER October 1, 2005 / Volume 104 / Number 7
bination with brachytherapy, or cryosurgery (cryoablation or freezing of the prostate gland).2– 4 Watchful waiting (deferring treatment) is also a viable management approach for some patients. This approach involves regular monitoring with prostate-specific antigen (PSA) tests and digital rectal examinations, with treatment initiated when disease or symptoms progress. Patients with prostate carcinoma and their families not only have to cope with the distress caused by the diagnosis of cancer,5 but they also have to interpret large volumes of complex and often conflicting or incomplete information regarding treatment options and associated outcomes. The challenge of choosing a treatment is compounded by treatment options that differ in their side effects but generally offer comparable survival.2– 4 The most common side effects for radical prostatectomy include urinary incontinence and erectile dysfunction. The side effects of radiotherapy (RT) may include urinary symptoms, sexual function impairment, and rectal bleeding. In addition, the patient’s decision-making process and final treatment selection are further complicated by the many treatment options recommended by physicians of different specialty training, background, and practice settings.6 Recognizing that patients with prostate carcinoma face various challenges, ranging from coping with distress associated with diagnosis to deciding on a specific treatment option, we designed the current study to determine the reasons why patients selected the treatment they did. The current article reports initial treatment decision-making data from an ongoing longitudinal study evaluating treatment decision making and quality of life (QOL) outcomes in patients with localized prostate carcinoma. The research was guided, in part, by the Cognitive-Social Health Information Processing (C-SHIP) framework.5,7,8 The C-SHIP framework postulates that health-related behaviors are determined by the way an individual processes health information at the cognitive (e.g., after considering the benefits and harms of a specific treatment option) and affective levels (e.g., considering fear of cancer and death). The purpose of the current study was twofold: 1) to describe information sources, treatment decisionmaking preferences, perceptions of treatment outcomes (benefits and risks) associated with various treatment options, perceived difficulty, and emotional distress during treatment decision making in patients with early-stage prostate carcinoma and 2) to examine whether demographic, disease, cognitive, and affective factors were independently associated with treatment selection in patients treated at a comprehensive cancer center.
MATERIALS AND METHODS The current study was reviewed and approved by the institutional review board of the University of South Florida (Tampa, FL) and by the protocol review and monitoring committee of the H. Lee Moffitt Cancer Center. All the personnel involved in the conduct of the study completed the National Institutes of Health Human Subjects Protection training and met the appropriate Health Information Portability and Accountability Act (HIPAA) education requirements before engaging in this research. Patients were recruited and informed consent was obtained; patients were enrolled between February 2002 and May 2003.
Study Population and Recruitment Procedures Men diagnosed with clinically staged localized prostate carcinoma (T1–2Nx– 0M0 by the American Joint Committee on Cancer [AJCC]/International Union Against Cancer [UICC] TNM classification) were recruited in the multidisciplinary Genitourinary Oncology Clinics at the H. Lee Moffitt Cancer Center. All patients were approached for participation after their treatment decision was determined. There was no study-related intervention to influence the patients’ treatment decision making. The treatment decision was made under usual care settings after consultation with a urologist, a radiation oncologist, or both, depending on consultation recommendations and patient preference. The study was conducted as part of a larger prospective QOL comparison study. After the patient’s treatment decision was made, the collaborating clinician or study personnel introduced the patient to the study and the study personnel discussed the study in detail face-to-face with the patient in clinic. In a few cases, study personnel contacted patients by telephone to introduce the study and arrange a face-toface visit in the clinic. Patients were eligible for enrollment in the larger study if they met the following criteria: 1) scheduled to receive radical prostatectomy or brachytherapy for clinically localized prostate carcinoma (T1–2Nx– 0M0 by the AJCC/UICC TNM classification) with or without neoadjuvant hormonal therapy before local therapy; 2) Karnofsky performance status ⬎ 60; 3) able to read and understand English; 4) planning to receive follow-up care at H. Lee Moffitt Cancer Center after treatment or willing to provide release of information so follow-up clinical data may be obtained; 5) no documented physical or mental impairment that would interfere with successful completion of study questionnaires; and 6) provided written informed consent and HIPAA authorization. The decision-making assessment was completed
Treatment Decision Making in Prostate Ca/Gwede et al.
once, either pretreatment or 3–12 months after treatment. Of 173 patients who were approached to participate in the larger QOL study, 12 were ineligible, 15 refused to participate, and 146 agreed to participate (an 84% participation rate). Because the decisionmaking component of the study was initiated after accrual had begun on the ongoing QOL outcomes comparison study, 27 patients were not assessed for treatment decision making. Of the 119 patients who participated in decision-making assessments, 50 completed their decision-making questionnaires before treatment and 69 completed the decision-making questionnaires an average of 5.6 months after treatment (median, 5.9 mos). Decision-making questionnaires were completed in the clinic during a regularly scheduled pretreatment or posttreatment visit. Patients returned the completed questionnaires to the study personnel in person or in postage-paid envelopes, and each questionnaire was checked for completeness. Patients were not paid to participate.
Instruments and Measures Standard demographic information including age, race, education, marital status, and household income was collected by self-report. Clinical information, including disease stage (TNM classification), Gleason score (a measure of tumor grade/aggressiveness), pretreatment PSA test value, and time since diagnosis to study enrollment, was collected by medical chart review. The preferred role in prostate carcinoma treatment decision making was assessed using the Control Preference Scale (CPS), a self-report scale designed to elicit patients’ preferences for control over medical decision making.9,10 The CPS consists of five statements about different roles individuals can assume in treatment-related decisions. Patients were asked to review all five statements and check the one statement that best describes their preferred approach to medical decision making.9,10 Two statements describe a passive role (the physician makes the treatment decision for the patient, or the physician makes the decision after considering the patient’s opinion), one statement describes a collaborative role (patient and physician make the treatment decision together), and two statements describe an active role (the patient makes the final decision alone, or the patient makes the final decision after considering the physician’s opinion). For analysis purposes, responses were categorized into one of three groups: active, collaborative, or passive decision-making preference.9,10 Patients were also asked to indicate whether they had consulted with a urologist or radiation oncologist or other physician before making a treatment deci-
1383
sion. We also assessed other information sources patients used including other physicians, nurses, family and friends, books, magazines and television, use of the National Cancer Institute’s Cancer Information Service, and the Internet. In addition, we asked patients to indicate their perception of the appropriateness of each treatment option as a treatment for their own prostate carcinoma. For example, patients rated a statement, “Radical prostatectomy is a good way to treat my cancer” using the following response scale: 1, not at all; 2, a little bit; 3, somewhat; 4, quite a bit; 5, very much. Cognitive and affective variables related to treatment decision making were assessed using questions derived from Diefenbach et al.5 With regard to cognitive variables, we assessed perceived seriousness of prostate carcinoma, treatment-related beliefs, and reasons for choosing a specific treatment using items scored on a 5-point response scale: 1, not at all; 2, a little bit; 3, somewhat; 4, quite a bit; 5, very much. Items used to assess the reasons why men chose the specific treatment included beliefs that the chosen treatment offered the best chance for a cure; recommendations of physicians; and perceived the invasiveness, painfulness, and convenience of the chosen treatment. Patients were asked to indicate the three most important reasons (in rank order) for their treatment selection. Affective variables assessed included individual questions relating to worry about having prostate carcinoma; the perceived risk for spread, difficulty, and distress experienced during the decision process; the level of concern regarding whether the patient made the right decision; and whether the patient had regrets about their treatment choice (decisional regret). These questions were rated using a 6-point scale: 0, not applicable; 1, not at all; 2, a little bit; 3, somewhat; 4, quite a bit; 5, very much.
Data Analysis Data analyses were conducted using Statistical Analysis Software (SAS) (version 9.1.2; SAS Institute, Inc., Cary, NC), with the alpha value set at the 0.05 significance level (2-sided test). Univariate analyses were conducted using chi-square tests and Student t tests to compare treatment groups with regard to demographic information, disease, decision-making preference, and cognitive and affective variables. Multivariate analyses using only variables that were significant in univariate analyses (P ⬍ 0.05) were conducted to evaluate factors associated with treatment selection. Hierarchical logistic regression models were then used to determine the independent association of individual cognitive factors with treatment selection.
1384
CANCER October 1, 2005 / Volume 104 / Number 7 TABLE 1 Demographic, Disease, and Treatment Characteristics Variables Age in yrs Mean (SD) Median (range) Race White Marital status Married/living with partner Education High school/technical or less Employment status Employed full/part time Tumor (clinical) classification T1C T2A Lymph node staging N0 NX Gleason score ⱕ6 ⱖ7 PSA at diagnosis Mean (SD) Median (range) Time since diagnosis (mos) Mean (SD) Median (range)
Radical prostatectomy (n ⴝ 52) (%)
Brachytherapy (n ⴝ 67) (%)
57.7 (6.6) 58.5 (39.6–71.1)
65.2 (7.8) 65.8 (45.7–89.2)
P value
⬍ 0.0001
45 (86.5)
65 (97.0)
NSa
48 (92.3)
58 (86.6)
NS
13 (25.5)
15 (22.4)
NS
40 (76.9)
29 (43.3)
0.0002 NS
41 (78.8) 11 (21.2)
48 (71.6) 19 (28.4)
10 (19.2) 42 (80.8)
12 (17.9) 55 (82.1)
40 (76.9) 12 (23.1)
66 (98.5) 1 (1.5)
6.0 (3.3) 4.8 (0.8–17)
6.4 (3.9) 5.8 (0.6–28.1)
NS
3.4 (3.3) 2.6 (0.7–25.2)
5.1 (4.7) 4.2 (1.3–31.5)
0.02
NS
0.0002
SD: standard deviation; NS: not statistically significant; PSA: prostate-specific antigen. a Not statistically significant (P ⬎ 0.05) based on the -square test statistics for categoric variables and independent-sample Student t tests for continuous variables.
RESULTS Preliminary analyses were conducted to compare the 27 patients who did not complete the treatment decision assessment with the 119 patients who completed the decision-making assessment. No significant (P ⬍ 0.05) differences in demographic or disease variables were evident between these two groups. We then compared patients based on the timing of treatment decision-making assessment. We explored differences between the 50 patients who completed the decisionmaking packet pretreatment versus the 69 patients who completed the packet posttreatment. No significant (P ⬍ 0.05) differences were evident with regard to demographic information, disease, or decision-making variables (cognitive, affective, decision-making preference, and information sources) between these two groups.
Their mean age was 62 years (standard deviation ⫾ 8.2 yrs). Virtually all patients in the radical prostatectomy group underwent the retropubic radical prostatectomy procedure with bilateral (87%) or unilateral (5%) nerve sparing to preserve functioning. The iodine-125 isotope (97%) was the primary radioactive source in the brachytherapy group. Table 1 summarizes the demographic and disease characteristics by treatment group. Patients who chose radical prostatectomy were younger (P ⬍ 0.0001) than those selecting brachytherapy and were more likely to be employed full or part time (P ⬍ 0.0002). In addition, patients who chose prostatectomy had more aggressive tumors as measured by the Gleason score (P ⬍ 0.0002). Finally, the interval between diagnosis and treatment was longer for patients who received brachytherapy compared with patients who received prostatectomy (P ⫽ 0.02).
Demographic and Clinical Variables Overall, patients were predominantly white (92%), married or living with a partner (86%), employed (58%), and had at least some college education (60%).
Information Sources Patients were asked about information sources they used in making their treatment selection. Pooled anal-
Treatment Decision Making in Prostate Ca/Gwede et al.
1385
TABLE 2 Information Sources Used to Make Choice of Treatment: Number of Patients Responding “Yes” Variables Urologist Radiation oncologist Other physician Nurse or nurse practitioner Spouse (wife or significant other) Your children Other family members Friends Books, magazines, TV Internet National Cancer Institute’s Cancer Information Service Library
RP (n ⴝ 52) (%)
BT (n ⴝ 67) (%)
P value
All (n ⴝ 119) (%)
52 (100.0) 10 (21.7) 44 (86.3) 11 (25.6) 33 (70.2) 12 (27.3) 14 (31.1) 28 (56.0) 31 (64.6) 24 (49.0)
63 (95.5) 55 (83.3) 43 (69.4) 19 (33.9) 38 (62.3) 11 (20.4) 18 (32.1) 38 (60.3) 35 (58.3) 39 (62.9)
NSa ⬍ 0.0001 0.03 NS NS NS NS NS NS NS
115 (97.5) 65 (58.0) 87 (77.0) 30 (30.3) 71 (65.7) 23 (23.5) 32 (31.7) 66 (58.4) 66 (61.1) 63 (56.8)
17 (37.8) 7 (15.6)
33 (55.0) 16 (27.6)
NS NS
50 (47.6) 23 (22.3)
RP: radical prostatectomy; BT: brachytherapy; NS: not statistically significant; TV:television. a Not statistically significant (i.e., P ⬎ 0.05) based on -square test statistics.
ysis across treatment groups showed that patients used multiple sources of information including the urologist (97%); other physicians (77%); spouse (66%); books, magazines, and television (61%); the radiation oncologist (58%); and friends (58%). Table 2 provides a summary of the responses for the pooled analyses and by treatment group. Except for the radiation oncologist and other physicians, there were no significant (P ⬍ 0.05) differences noted between the two treatment groups. As would be expected, patients who received brachytherapy were more likely to report talking to a radiation oncologist (P ⬍ 0.0001) compared with patients who received prostatectomy. It is unclear why patients who received prostatectomy were more likely (P ⫽ 0.03) to report talking to other physicians.
Treatment Decision-Making Preferences The majority of patients preferred to have an active role in treatment decision making (63%). The minority preferred a collaborative role (29%) or passive role (8%). There were no significant (P ⬍ 0.05) differences by treatment group in medical decision-making preference as assessed by the CPS. No statistically significant association of cognitive and affective factors was found with treatment decision-making preference.
Affective Variables When data from both treatment groups were pooled, 49% of patients reported difficulty in making the treatment decision using a criterion of 3 (somewhat) or greater on the response scale, 35% were worried about whether they made the right decision (using the same
criterion), and 45% reported that they were distressed while making the treatment decision (using the same criterion). When asked if they had any regrets concerningtheir treatment decision, 74% of patients responded not at all, 20% responded a little bit or not applicable, and only 6% reported regrets regarding treatment decision using a criterion of 3 (somewhat) or greater on the response scale. There were no significant (P ⬍ 0.05) differences noted between the two treatment groups in all affective variables, suggesting that patients in both groups felt essentially the same way about the difficulty of decision making, worry about the decision made, distress while making the treatment decision, and decisional regret.
Cognitive Variables In pooled analyses, the average patient believed he received enough information about each treatment option (mean score, 4.4), and generally believed there are many ways to treat prostate carcinoma (mean score, 3.9). Brachytherapy and radical prostatectomy were rated similarly as good treatments for prostate carcinoma, as shown in Table 3 (mean scores, 3.6 and 3.4, respectively). Watchful waiting (mean score, 1.5) and cryotherapy (mean score, 1.7) were also rated similarly but less favorably than brachytherapy and prostatectomy. However, treatment group comparisons showed that patients differed in their perceptions of the efficacy of the two primary treatment options, as summarized in Table 3 (i.e., patients who chose prostatectomy more strongly endorsed the statement, radical prostatectomy is a good way to treat my cancer, than patients who chose brachytherapy [P
1386
CANCER October 1, 2005 / Volume 104 / Number 7
TABLE 3 Mean Scores and SD for Cognitive Beliefs Regarding Treatment Optionsa Characteristics There are several ways to treat early-stage prostate cancer successfully. Radiation seed implant (placement of radiation seeds into the prostate gland) is a good way to treat my cancer. Cryosurgery (freezing the prostate) is a good way to treat my cancer. Watchful waiting (observation or careful monitoring of untreated cancer) is a good way to manage my cancer. Radical prostatectomy (surgical removal of the prostate gland) is a good way to treat my cancer. The only chance for a cure is removal of the prostate. Given your understanding of your prostate cancer, how serious do you believe your prostate cancer is? I worry that my prostate cancer may not be eradicated by the treatment I chose.
RP (n ⴝ 52) Mean (SD)
BT (n ⴝ 67) Mean (SD)
P value
All (n ⴝ 119) Mean (SD)
3.6 (1.29)
4.1 (0.96)
0.008
3.9 (1.13)
2.1 (1.22)
4.7 (0.49)
⬍ 0.0001
3.6 (1.56)
1.6 (0.83)
1.8 (0.98)
NSb
1.7 (0.92)
1.3 (0.77)
1.8 (1.07)
NS
1.5 (0.96)
4.7 (0.87)
2.4 (1.44)
0.0001
3.4 (1.65)
3.7 (1.40)
1.6 (1.15)
0.0001
2.5 (1.64)
3.9 (1.14)
3.2 (1.02)
0.0009
3.5 (1.12)
2.1 (1.15)
2.2 (1.01)
NS
2.1 (1.07)
SD: standard deviation; RP: radical prostatectomy; BT: brachytherapy; NS: not statistically significant. a Response scale: 1: not at all; 2: a little bit; 3: somewhat; 4: quite a bit; 5: very much. b Not statisticaaly significant (i.e., P ⬎ 0.05) based on independent-sample Student t tests.
⬍0.0001]). Similarly, patients who chose brachytherapy more strongly endorsed the statement, “radiation seed implants is a good way to treat my cancer” than patients who selected prostatectomy (P ⬍ 0.0001). In addition, patients who selected radical prostatectomy more strongly endorsed the statement, “the only chance for a cure is removal of the prostate” than patients who chose brachytherapy (P ⬍ 0.0001). Comparisons by treatment group indicated that the average patient who underwent prostatectomy considered his disease more serious than the average patient who underwent brachytherapy (P ⫽ 0.0009). With regard to the reasons for treatment selection, there were statistically significant differences noted between the two treatment groups as summarized in Table 4. The prostatectomy group believed more strongly that the treatment they chose offered the best chance for cure (P ⬍ 0.0001). The brachytherapy group believed more strongly that the treatment they chose was less invasive (P ⬍ 0.0001), promised the fewest side effects (P ⬍ 0.0001), was the least painful (P ⬍ 0.0001), and was convenient (P ⬍ 0.0001). Not surprisingly, patients who received brachytherapy believed more strongly that they chose the treatment they did because they “wanted to avoid surgery” (P ⬍0.0001). We also asked patients to indicate which of the
reasons above was the single most important reason for selecting the treatment they chose. For the pooled sample, the single most important reason for selecting the treatment was because it offered the best chance for cure (44%), followed by because it promised the fewest side effects (14%), and it was the least invasive (13%). For patients in the prostatectomy group, the single most important reason for selecting this treatment was because it offered the best chance for cure (74%). In contrast, most patients in the brachytherapy group said the single most important reason they chose this treatment was because it was least invasive (23%) or it promised the fewest side effects (21%), or it offered the best chance for cure (20%).
Factors Associated with Treatment Choice Univariate logistic regression We assessed whether the individual cognitive and affective factors accounted for the variance in treatment selection over and above the effects of demographic and disease variables. First, we developed a fitted model of control variables by regressing treatment selection against demographic (age, employment status) and disease (Gleason score, time since diagnosis) factors that demonstrated significant associations (P ⬍ 0.05) in bivariate analyses. No association was found for employment status and time since diagno-
Treatment Decision Making in Prostate Ca/Gwede et al.
1387
TABLE 4 Mean Scores and (SD) for Cognitive Factors (Reasons for Treatment Selection)a Overall, I chose the treatment I did because:
RP (n ⴝ 52) (%)
BT (n ⴝ 67) (%)
P value
It offers the best chance of cure It was the least painful It was less invasive Different doctors recommended it It promised to have the fewest side-effects I had confidence in my doctor I wanted to avoid surgery Because it was convenient I know of people who had this treatment and were satisfied
4.9 (0.5) 1.5 (1.0) 1.5 (1.0) 3.7 (1.4) 2.5 (1.4) 4.7 (0.6) 1.5 (1.3) 1.2 (0.6)
4.2 (0.8) 3.6 (1.3) 4.3 (1.0) 3.7 (1.3) 4.4 (0.9) 4.5 (0.7) 4.3 (1.0) 3.1 (1.5)
⬍ 0.0001 ⬍ 0.0001 ⬍ 0.0001 NSb ⬍ 0.0001 NS ⬍ 0.0001 ⬍ 0.0001
3.5 (1.4)
3.8 (1.4)
NS
RP: radical prostatectomy; BT: brachytherapy; NS: not statistically significant. a Response scale: 1: not at all; 2: a little bit; 3: somewhat; 4: quite a bit; 5: very much. b Not statistically significant (i.e., P ⬎ 0.05) based on independent-sample Student t tests.
TABLE 5 Univariate Logistic Regression Analyses: Modeling the Likelihood of Choosing Radical Prostatectomy Versus Brachytherapya Wald -square test Characteristics Control variables Age (continuous) Gleason score (ⱕ 6 vs. ⱖ 7) Cognitive variablesc It offers the best chance of cure It was least painful It was less invasive It promised to have the fewest side effects I wanted to avoid surgery Because it was convenient
OR
Chi-square test
P value
OR
95% CIb
18.8 8.4
⬍ 0.0001 0.004
0.85 25.13
0.79–0.92 2.84–222.70
14.8 25.8 21.5 24.1 32.1 16.6
0.0001 ⬍ 0.0001 ⬍ 0.0001 ⬍ 0.0001 ⬍ 0.0001 ⬍ 0.0001
8.33 0.25 0.13 0.28 0.32 0.23
2.83–24.56 0.14–0.42 0.06–0.31 0.17–0.47 0.21–0.47 0.11–0.47
OR: odds ratio; 95% CI: 95% confidence interval. a For all predictors, a 1-unit change in the response scale was associated with the reported odds ratio. An odds ratio ⬎ 1 reflected a positive association, whereas an odds ratio ⬍ 1 depicted a negative relation. Radical prostatectomy was the reference category. b 95% confidence intervals for odds ratios (lower, upper). c Ordinal 1–5.
sis. As shown in Table 5, older patients were less likely to choose radical prostatectomy (odds ratio [OR] ⫽ 0.85; 95% confidence interval [95% CI], 0.79 – 0.92) and patients with Gleason Grade ⱖ 7 were more likely to choose radical prostatectomy (OR ⫽ 25.13; 95% CI, 2.84 –222.70). These two significant variables were retained as control variables in all subsequent analyses. We then we assessed the association of individual cognitive and affective factors on treatment selection in separate logistic regression models containing control variables (age and Gleason score). Several cognitive factors demonstrated statistically significant associations with treatment selection. Results (Table 5) showed that after controlling for age and Gleason score, patients who underwent prostatectomy were
more likely to say they chose their treatment because it offers the best chance for cure (OR ⫽ 8.33; 95% CI, 2.83–24.56). In contrast, patients who underwent brachytherapy said they chose their treatment because it was least painful (OR ⫽ 0.25; 95% CI, 0.14 – 0.42), it was less invasive (OR ⫽ 0.13; 95% CI, 0.06 – 0.31), it promised to have the fewest side effects (OR ⫽ 0.28; 95% CI, 0.17– 0.47), I wanted to avoid surgery (OR ⫽ 0.32; 95% CI, 0.21– 0.47), and because it was convenient (OR ⫽ 0.23; 95% CI, 0.11– 0.42). No statistically significant associations were found for affective variables.
Multivariate logistic regression We then performed multivariate analyses using a hierachical approach (adding one predictor at a time) to
1388
CANCER October 1, 2005 / Volume 104 / Number 7
determine which cognitive variables were independently associated with treatment choice after adjusting for age and tumor grade. The six cognitive factors found to be significantly associated with treatment selection in univariate analyses above (Table 5) were included hierachically as potential predictors. In the final multivariate model, patients who underwent prostatectomy were less likely than patients who received brachytherapy to say they chose their treatment because it was less invasive (OR ⫽ 0.19; 95% CI, 0.08 – 0.46). Patients who received prostatectomy also were less likely than patients who received brachytherapy to report that they chose their treatment to avoid surgery (OR ⫽ 0.44; 95% CI, 0.23– 0.87). No interaction effects between cognitive factors and with age or Gleason score were found.
DISCUSSION The current study represents a practical method of assessing treatment decision making and reasons for treatment selection in patients with early-stage localized prostate carcinoma. Although the topic of treatment decision making in this population has been investigated,6,10 –19 to our knowledge only a handful of studies used prospective, quantitative, cognitive-affective approaches to assess reasons for treatment selection from the patients’ perspective.5 Furthermore, the current study adjusted for the effects of demographic and disease characteristics to determine the independent role of cognitive and affective factors in treatment selection and found that cognitive factors account for variance in treatment selection over and above the effects of age and Gleason score. Our findings suggest that patient-centered considerations are important in treatment decision making when treatment options have comparable efficacy and survival but differing side effects. With regard to theoretic perspectives, our findings partially support the relevance of the C-SHIP model in prostate carcinoma treatment decision making, and corroborate previous literature.5 Certain cognitive factors were statistically and significantly associated with treatment selection, whereas no associations were found for affective factors. Nevertheless, the moderate to high decisional difficulty and distress reported during decision making demonstrate the importance of measuring and addressing affective variables in this context. With regard to information sources used in decision making, the current study results support and extend the findings from previous literature.5 These findings provide a more comprehensive picture of the process and sources of information used in treatment decision making in patients with early-stage prostate carcinoma. Physicians, notably the urologist, were
commonly cited as the most important source of information. It is noteworthy that the influence of the urologist was prominent in both groups of patients given that previous research6 suggested that urologists and radiation oncologists each express bias toward the treatment they provide rather than accurately informing patients about the pros and cons of each treatment option. Although we did not assess the content and delivery of information, these findings suggest that urologists at a comprehensive cancer center most likely offered the patient a balanced presentation of the relevant management options. Patients also used other sources of information, suggesting that patients with localized prostate carcinoma seek to be informed about their treatment options and outcomes before making a final treatment decision. The variety of information sources, although encouraging, may also add to the challenge of deciphering often complex, conflicting, and incomplete information from multiple sources. Furthermore, the quality of these information sources (Internet, books, magazines, friends, family members, survivors) is unknown, and physician bias toward specific treatments6 remains a critical challenge as uncertainties regarding the optimal management option remain. The results of the current study demonstrated that the majority of patients (63%) want to be actively involved in making the final treatment selection and 29% prefer a collaborative approach of shared decision making with their physician, whereas only 8% deferred the decision to their physician. These findings are consistent with previous literature showing that 43% preferred an active role, 47% preferred a collaborative role, and 10% preferred a passive role in medical decision making.20 The current study found no statistically significant associations of treatment selection or cognitive-affective factors with decisionmaking preference, and the benefits of assuming an active role in treatment decision making are not well understood. However, Davison and Degner18 reported that men who received an empowerment intervention assumed a more active role in treatment decision making and had lower levels of state anxiety compared with men in the control condition. Clearly, the current findings and the literature5,21 highlight the importance of shared informed decision making, in which clinicians recognize the patient’s role in decision making and balance their discussion of all treatment options and consider the impact of the treatments on patient QOL and function. With regard to affective variables, our analyses suggest that the decision-making process is difficult and distressful for many patients irrespective of the treatment selected. Nearly 50% of patients expressed
Treatment Decision Making in Prostate Ca/Gwede et al.
having experienced difficulty (49%) and distress (45%) while making their treatment decision. However, despite the difficulty and distress associated with treatment decision making, once the decision was made, patients had no regrets regarding the decision and would make the same choice if they had to make the decision again. Although this cognitive dissonance is consistent with the literature,5 the question remains would patients experience distress and regret in longterm follow-up as they experience the reality of chronic side effects with alteration in QOL and are faced with actual negative disease outcomes such as recurrence or metastases? A related question is would such long-term distress and decisional regret differ by type of initial treatment selected? Additional longterm data are needed to explore and understand these questions, and elucidate the importance of affective factors in this context. An intriguing cognitive dimension was the patients’ beliefs in the efficacy and appropriateness of the different treatment options (Table 3). Although the average patient strongly believed (mean score, 3.6) that there are several ways to treat early-stage localized prostate carcinoma successfully, there were notable treatment group differences with regard to the best way to treat each man’s own cancer. Generally, each group of patients believed strongly in the therapy they chose and rated the other treatments poorly. These findings offer practical and meaningful insight as patients strongly endorsed the treatment they selected, despite the prevailing belief among clinicians that the primary treatment options offer comparable survival. With regard to cognitive reasons for treatment selection, the univariate and multivariate analyses offer consistent findings and demonstrated treatment group differences. Univariate logistic regression analyses showed that six cognitive factors were associated with treatment selection over and above the effects of age and Gleason score. These findings suggest that cure is a paramount consideration among men who chose radical prostatectomy. For patients who received brachytherapy, pain, invasiveness, side effects, convenience, and the desire to avoid surgery were important considerations in treatment selection over and above age and Gleason score. Multivariate analyses also showed that patients chose brachytherapy because it was less invasive and they wanted to avoid surgery. Taken together, the current study findings suggest that the prostatectomy group prioritized cure over everything else whereas the brachytherapy group emphasized the preservation of QOL in addition to wanting cure. These findings reflect both the complexity and challenges of decision making because pa-
1389
tients are faced with treatment options with clearly disparate side effects and have to balance or prioritize the values they seek from treatment. This presents an opportunity to assist patients through educational or decision-making aids that can be used to complement the interactions with clinicians. Furthermore, these findings partially support the relevance of the C-SHIP model as a salient framework for understanding prostate carcinoma treatment decision making. During the past decade studies have investigated prostate carcinoma treatment preferences using a variety of descriptive methodologies, including retrospective11–13 and qualitative designs with patients alone, couples, and prostate carcinoma survivors.14,15 Mazur and Hickam13 showed that among patients who received prostatectomy, complete removal of the tumor was the main reason for their treatment choice, whereas fear of surgical complications was the main reasons patients chose watchful waiting. Findings from qualitative research also showed that patients have misconceptions concerning their treatment options14 and need more information to facilitate their decision making.15 A recent prospective study5 compared decision making among patients treated with prostatectomy, EBRT, brachytherapy, and watchful waiting. These investigators5 found that treatment decisions are influenced by cognitive beliefs and affective factors. Factors that favored radical prostatectomy were the perceived seriousness of prostate carcinoma diagnosis, curability of prostate carcinoma, and risk of prostate carcinoma spreading. Factors that favored RT options included perceived invasiveness, painfulness, and convenience of treatment. Furthermore, these authors also report that once the treatment decision was made and implemented, there was little cognitive dissonance or regret over the treatment option chosen. Despite the limited sample size and single-institution participation, the findings of the current study corroborate and extend the literature5 concerning treatment decision making as the optimal treatment option for early-stage prostate carcinoma remains debated. Patients with localized prostate carcinoma need to be afforded ample time and resources to facilitate what is sometimes a lengthy, difficult, and distressful decision-making process. Additional data are needed to assess whether patients experience distress and regret in long-term follow-up as they experience long-term disease outcomes and treatmentrelated chronic side effects. To our knowledge, only a few studies to date have assessed decisional regret within the first year19,22 or 2 years23 after treatment and found no regrets. Longer follow-up is needed. The current findings, in addition to the existing literature, support the need for the development of
1390
CANCER October 1, 2005 / Volume 104 / Number 7
decision aids and interventions to reduce the decisional difficulty and psychologic distress associated with treatment selection. A shared decision-making approach is important because patients do want to take an active or collaborative role in treatment decisions. Clinicians must recognize the need to provide each patient, irrespective of his decision-making preference (active, collaborative, or passive), with a balanced discussion of all relevant treatment options and allow adequate time for informed decision making. Decision aids that address both general information needs and patient-centered concerns and outcome priorities are still being developed and evaluated20,24 –27 for effectiveness and impact on patients’ decisions, knowledge, anxiety, depression, satisfaction, and acceptability. It is plausible that such interventions may reduce the psychologic distress associated with treatment decision making, but much research remains to be done.
REFERENCES 1.
Jemal A, Murray T, Ward E, et al. Cancer statistics, 2005. CA Cancer J Clin. 2005;55:10 –30. 2. Scher HI, Isaacs JT, Zelefsky MJ. Prostate cancer. In: Abeloff MD, Armittage JO, Lichter AS, Niederhuber JE, editors. Clinical oncology. 2nd ed. New York: Churchill Livingstone, 2000:1823–1884. 3. Ragde H, Korb LJ, Elgamal AA, Grado GL, Nadir BS. Modern prostate brachytherapy. Prostate specific antigen results in 219 patients with up to 12 years of observed follow-up. Cancer. 2000;89:135–141. 4. Cookson MM. Prostate cancer: screening and early detection. Cancer Control. 2001;8:133–140. 5. Diefenbach MA, Dorsey J, Uzzo RG, et al. Decision-making strategies for patients with localized prostate cancer. Semin Urol Oncol. 2002;20:55– 62. 6. Fowler FJ Jr., McNaughton Collins M, Albertsen PC, Zietman A, Elliott DB, Barry MJ. Comparison of recommendations by urologists and radiation oncologists for treatment of clinically localized prostate cancer. JAMA. 2000;283:3217–3222. 7. Miller SM, Shoda Y, Hurley K. Applying cognitive-social theory to health-protective behavior: breast self-examination in cancer screening. Psychol Bull. 1996;119:70 –94. 8. Miller S, Diefenbach M. The cognitive-social health information processing (C-SHIP) model: a theoretical framework for research in behavioral oncology. In: Krantz D, editor. Perspectives in behavioral medicine. Mahwah, NJ: Lawrence Erlbaum, 1998:219 –244. 9. Degner LF, Sloan JA, Venkatesh P. The control preferences scale. Can J Nurs Res. 1997;29:21– 43. 10. Davison BJ, Gleave ME, Goldenberg SL, Degner LF, Hoffart D, Berkowitz J. Assessing information and decision preferences of men with prostate cancer and their partners. Cancer Nurs. 2002;25:42– 49.
11. Mazur DJ, Merz JF. Older patients&apso; willingness to trade off urologic adverse outcomes for a better chance at five-year survival in the clinical setting of prostate cancer. J Am Geriatr Soc. 1995;43:979 –984. 12. Mazur DJ, Merz JF. How older patients’ treatment preferences are influenced by disclosures about therapeutic uncertainty: surgery versus expectant management for localized prostate cancer. J Am Geriatr Soc. 1996;44:934 –937. 13. Mazur DJ, Hickam DH. Patient preferences for management of localized prostate cancer. West J Med. 1996;165:26 –30. 14. O’Rourke ME, Germino BB. Prostate cancer treatment decisions: a focus group exploration. Oncol Nurs Forum. 1998; 25:97–104. 15. O’Rourke ME. Narrowing the options: the process of deciding on prostate cancer treatment. Cancer Invest. 1999;17: 349 –359. 16. O’Rourke ME. Urinary incontinence as a factor in prostate cancer treatment selection. J Wound Ostomy Continence Nurs. 2000;27:146 –154. 17. Davison BJ, Degner LF, Morgan TR. Information and decision-making preferences of men with prostate cancer. Oncol Nurs Forum. 1995;22:1401–1408. 18. Davison BJ, Degner LF. Empowerment of men newly diagnosed with prostate cancer. Cancer Nurs. 1997;20:187–196. 19. Davison BJ, Goldenberg SL. Decisional regret and quality of life after participating in medical decision-making for earlystage prostate cancer. BJU Int. 2003;91:14 –17. 20. Davison BJ, Parker PA, Goldenberg SL. Patients’ preferences for communicating a prostate cancer diagnosis and participating in medical decision-making. BJU Int. 2004;93:47–51. 21. Davison BJ, Keyes M, Elliott S, Berkowitz J, Goldenberg SL. Preferences for sexual information resources in patients treated for early-stage prostate cancer with either radical prostatectomy or brachytherapy. BJU Int. 2004;93:965–969. 22. Moore KN, Griffiths D, Hughton A. Urinary incontinence after radical prostatectomy: a randomized controlled trial comparing pelvic muscle exercises with or without electrical stimulation. BJU Int. 1999;83:57– 65. 23. Potosky AL, Legler J, Albertsen PC, et al. Health outcomes after prostatectomy or radiotherapy for prostate cancer: results from the Prostate Cancer Outcomes Study. J Natl Cancer Inst. 2000;92:1582–1592. 24. Davison BJ, Degner LF. Feasibility of using a computerassisted intervention to enhance the way women with breast cancer communicate with their physicians. Cancer Nurs. 2002;25:417– 424. 25. Davison BJ, Goldenberg SL, Gleave ME, Degner LF. Provision of individualized information to men and their partners to facilitate treatment decision making in prostate cancer. Oncol Nurs Forum. 2003;30:107–114. 26. Auvinen A, Hakama M, Ala-Opas M, et al. A randomized trial of choice of treatment in prostate cancer: the effect of intervention on the treatment chosen. BJU Int. 2004;93:52–56. 27. Whelan T, O’Brien MA, Villasis-Keever M, et al. Impact of cancer-related decision aids. Evid Rep Technol Assess (Summ). 2002;(46):1– 4.
