BACKGROUND/RATIONALE:
Newly diagnosed localized prostate cancer patients face difficult decisions about treatment and management; including radiation therapy, radical prostatectomy, brachytherapy, and observation. Previous studies have cited patient preference and physician recommendation for treatment as major roles in treatment decisions.
OBJECTIVE(S):
This grant will be used to develop and evaluate a decision analysis based intervention to improve decision-making among veterans with newly diagnosed localized prostate cancer. While the ultimate goal of this study is to improve decision-making among these patients, factors influencing physician recommendations for treatment will be sought. The former will be achieved by evaluating the efficacy of providing physicians with information incorporating patient preference for alternative health states with probabilities of treatment outcomes. Factors influencing physician recommendations for treatment will be determined through qualitative interviews with physicians.
METHODS:
Patients were accrued at the VA Chicago Health Care System Lakeside Urology Clinic. Upon consent for participation in the study, comorbidities, histologic grade of the biopsy, and age were obtained for 13 patients with newly diagnosed localized prostate cancer. Using a standard gamble technique, interviewers obtained patient utilities for 5 distinct health states related to prostate cancer treatment. Clinical and pathologic characteristics were incorporated into the decision analytic model, and the derived quality-adjusted life expectancies (QALEs) were shared with the treating urologist before patient-physician discussion of treatment options. The feasibility of the decision analytic intervention and unbiased estimates of the impact of the intervention was evaluated. The long-range objective of this proposal is to design an intervention for veterans that increases physician understanding of patient treatment preferences and patient understanding of choices in prostate cancer treatment.
FINDINGS/RESULTS:
From April 30 to October 29, 2001, 17 of the 23 newly diagnosed localized prostate cancer patients treated at the VA Chicago-Lakeside Division participated in the decision analytic computer model intervention. Reasons for non-participation included: four patients did attend the clinic, 1 patient was legally blind, and 1 had advanced cancer. Four of the 17 enrolled patients were ineligible because they were accrued into a discontinued control arm or difficulty was experienced with the computer model. Four of the eligible patients were African-American and 9 were White. The mean age was 68.9 years and 46.1 percent were married. Four patients were unwilling to take any risk for any possible health state. One patient did not take any risk for current health, impotence, incontinence, and combined impotence and incontinence, but was willing to take risks for the watchful waiting health state. Seven patients had inconsistent utilities, rating the development of combined sexual and bladder dysfunction as better than the development of either dysfunction alone. No patients rated the health state of combined incontinence and impotence as worse than either health state alone. The model and physician recommendation were concordant for 7 patients. In one discrepancy, the doctor recommended surgery over the model’s recommendation of watchful waiting due to the patient’s younger age and good state of health. In a second, the doctor recommended radiation therapy, while the model recommended surgery due to significant patient comorbidities. In the 4 discordant cases, the model selected radiation therapy versus the physician’s choice of watchful waiting. The physician’s decision appeared to be influenced by the patient’s age and psychosocial concerns. For two patients that the urologist was asked about the appropriateness and usefulness of the model recommendation, the physician found the recommendation reasonable in the case of one but not helpful due to younger age in the other.
IMPACT:
Future refinements of both the utility assessment exercise and decision analytic model may be needed before the feasibility of the model in the clinical setting can be determined.
PUBLICATIONS:
Journal Articles
- Knight SJ, Nathan DP, Siston AK, Kattan MW, Elstein AS, Collela KM, Wolf MS, Slimack NS, Bennett CL, Golub RM. Pilot study of a utilities-based treatment decision intervention for prostate cancer patients. Clinical Prostate Cancer. 2002; 1(2): 105-14.
