BACKGROUND: Whether to perform periodic rectal examinations in asymptomatic men as a screening test for prostatic cancer remains controversial. A randomized clinical trial that tests the efficacy of further evaluation and treatment of men who have been found to have asymptomatic prostate nodules may never be carried out. Decision analysis was therefore used to further investigate this clinical issue. METHODS: A decision tree was developed to model the decision of whether to biopsy an asymptomatic prostate nodule found by digital rectal examination in a 65-year-old man by his primary care physician. Test operating characteristics, probabilities of disease at different stages, probabilities of side effects from various treatments, and average life expectancies were obtained from the medical literature. Utilities for the various possible health outcome states were obtained from ratings by two experienced primary care physicians using the Kaplan-Anderson Quality of Well-Being Scale. These were used to adjust the quality-of-life expectancies for each outcome state. Multiple sensitivity analyses were performed to assess the robustness of the conclusions. RESULTS: Disregarding patient utilities, the average survival benefit of evaluation and treatment is 1.1 months. When quality-of-life adjustments are included in the analysis, evaluation and treatment results in an average loss of 3.5 quality-adjusted months of life. Factors that shift the decision toward evaluation and treatment include a positive predictive value of a prostate nodule for cancer of 49% or greater, specificity of prostate biopsy of 98.3% or greater, and the availability of much more effective treatment for stage D cancers. Factors that do not substantially affect the decision are cancer-free life expectancy, the percentage of cancers that are stage B at time of discovery, the sensitivity of prostate biopsy, and more effective treatment for stage C cancer, assuming the same rate of adverse consequences from treatment. CONCLUSIONS: The evaluation and treatment of prostatic nodules found by digital rectal examination in asymptomatic men in the primary care setting does not lead to significant improvement in life expectancy and adversely affects quality of life. Digital rectal examination should not be performed by primary care physicians as a screening test for prostate cancer.
BACKGROUND: Whether to perform periodic rectal examinations in asymptomatic men as a screening test for prostatic cancer remains controversial. A randomized clinical trial that tests the efficacy of further evaluation and treatment of men who have been found to have asymptomatic prostate nodules may never be carried out. Decision analysis was therefore used to further investigate this clinical issue. METHODS: A decision tree was developed to model the decision of whether to biopsy an asymptomatic prostate nodule found by digital rectal examination in a 65-year-old man by his primary care physician. Test operating characteristics, probabilities of disease at different stages, probabilities of side effects from various treatments, and average life expectancies were obtained from the medical literature. Utilities for the various possible health outcome states were obtained from ratings by two experienced primary care physicians using the Kaplan-Anderson Quality of Well-Being Scale. These were used to adjust the quality-of-life expectancies for each outcome state. Multiple sensitivity analyses were performed to assess the robustness of the conclusions. RESULTS: Disregarding patient utilities, the average survival benefit of evaluation and treatment is 1.1 months. When quality-of-life adjustments are included in the analysis, evaluation and treatment results in an average loss of 3.5 quality-adjusted months of life. Factors that shift the decision toward evaluation and treatment include a positive predictive value of a prostate nodule for cancer of 49% or greater, specificity of prostate biopsy of 98.3% or greater, and the availability of much more effective treatment for stage D cancers. Factors that do not substantially affect the decision are cancer-free life expectancy, the percentage of cancers that are stage B at time of discovery, the sensitivity of prostate biopsy, and more effective treatment for stage C cancer, assuming the same rate of adverse consequences from treatment. CONCLUSIONS: The evaluation and treatment of prostatic nodules found by digital rectal examination in asymptomatic men in the primary care setting does not lead to significant improvement in life expectancy and adversely affects quality of life. Digital rectal examination should not be performed by primary care physicians as a screening test for prostate cancer.