Early detection of prostate cancer .2. Estimating the risks, benefits, and costs

Purpose: To evaluate the potential benefits, harms, and economic consequences of digital rectal examination and measurement of prostate-specific antigen (PSA) for the early detection of prostate cancer. Data Sources: Relevant studies were identified from a MEDLINE search (1966 to 1995), reviews, bibliographies of retrieved articles, author files, and abstracts. Study Selection: Probabilities for individual clinical outcomes were derived from various sources, including the largest screening study of community volunteers to date, analyses of Medicare claims, and recently published meta-analyses of the outcomes of alternative treatment strategies. Cost estimates were based on the 1992 Medicare fee schedule. Data Extraction: A cost-effectiveness model for onetime digital rectal examination and PSA measurement was constructed to examine the possible outcomes. Results: If a favorable set of assumptions is used, one-time digital rectal examination and PSA measurement may increase average life expectancy by approximately 2 weeks at a reasonable marginal cost for men who are between 50 and 69 years of age. Considerable iatrogenic illness would occur. If less favorable assumptions are used, the estimated net benefit would decrease and cost-effectiveness ratios would dramatically increase. Even if favorable assumptions are used, the model suggests that screening adds only a few days to the average life expectancy of men who are older than 69 years of age. If the assumptions are less favorable, older men are harmed. Conclusions: The model suggests that screening may be reasonable in younger men if optimistic assumptions consistent with existing observational data are made. The lack of direct evidence showing a net benefit of screening for prostate cancer seems to mandate more clinician-patient discussion for this procedure than for many other routine tests.