Coley and colleagues [1] analyzed the consequences of early detection and treatment of prostate cancer by using a pessimistic estimate of normal life expectancy. Because the general population includes many seriously ill persons, using the life expectancy derived from the National Center for Health Statistics data (the authors' Table 1) minimizes the potential benefits of any screening strategy.

Clinicians screen healthy men for prostate cancer, not all men in the population. A projection of potential years of life gained by screening that reflects the good general health of candidates for screening would reveal greater advantages for screening. The insurance industry provides data for a suitable projection of the consequences of screening a healthy population. Our underwriting process effectively selects a segment of the population that has considerably longer life expectancy than shown in Coley and colleagues' Table 1. We identify persons who lack significant illness, have acceptable results on common laboratory tests (such as urinalysis, serum chemistry panels, and electrocardiography), and abstain from the use of tobacco. For such a population, the Transactions of the Society of Actuaries 1991-1992 Reports presents the following life expectancy for men: 55 years of age, 27.4 years; 65 years of age, 20.3 years; and 75 years of age, 12.6 years. The comparable figures in the article by Coley and associates were 21.4, 14.45, and 8.95 years, respectively.

The personal physician knows all of this information about the candidate for screening and should consider the decision for screening in light of the true life expectancy of each individual patient. When this projection is used, early detection and early treatment of prostate cancer generate profoundly greater benefits. Would the authors recalculate their figures to show the effect of screening on men with this favorable health status?