Both Dr. Pauker [1] and Dr. Schapira and colleagues [2] appear to have missed the point in their analysis of screening for preclinical ovarian cancer. I do not care for "average" patients; I care for individual patients. The authors' data indicate that the life expectancy of my patients who do have preclinical ovarian cancer will be increased, on average, by 11.6 years by a screening program using CA 125 and transvaginal sonography. The optimal benefit of such a program would be 26 additional years of life expectancy—more than a quarter of a century. That's better than winning the lottery!

Now, what's the risk? The data indicate that the life expectancy of the average subject will be shortened by 25 minutes due to "complication of laparotomy when results of both CA 125 level and the transvaginal sonogram are falsely positive" [1]. So, what is 25 minutes in an average life expectancy of nearly 80 years for women in the United States? In comparison to an average gain in life expectancy of 14 hours (a 28:1 risk benefit ratio), the risk seems trivial. Again, the numbers miss the point entirely: Neither a 14-hour gain nor a 25-minute loss is meaningful, but that 25-minute reduction in life expectancy results from some young woman dying prematurely and unnecessarily, whereas many other women undergo surgery uneventfully and are reassured by normal findings. Averages do not begin to compute the calculus of this risk–benefit tradeoff.

How can we most optimally resolve the tension between risk and benefit in such decisions for the patients under our care? Surely not by a government or managed care organization fiat that looks only at the averages and ignores the individualities! The physician's duty is to provide the best available current information to individual patients and to assist them in making their own best decisions. The emotional content of these decisions is far more important than the bloodless statistics.