To the Editor.
—The article by Dr Krahn and colleagues1 on screening for prostate cancer is misleading and cannot be applied to the average prostate cancer patient. The study perpetuates erroneous assumptions, questionable utilities, and flawed data on rates of metastases, mortality, and treatment complications used in decision analytic studies of prostate cancer.2 Many of the data are outdated and precede the recent use of prostate-specific antigen (PSA) for screening, thus underestimating the sensitivity of this test. Recent studies also have shown a stage shift from 35% curable cancer in unscreened men to 70% to 85% curable cancer in screened populations.3By basing the analysis solely on the study by Johansson et al,4 Krahn et al substantially underestimate the lethal potential of prostate cancer. The population in the study by Johansson et al contained a disproportionate number of elderly men with less aggressive cancers. These men