It has been predicted that physicians of the 21st century will be judged not so much by their ability to cure disease as by their ability to prevent disease.1 Intervening in the absence of signs or symptoms of disease, however, is a troubling conundrum for the profession that aspires above all else to do no harm, as any clinician knows who has had to tell a healthy woman that her mammogram was "not normal." At times, knowing the benefit-risk ratio for a preventive intervention would seem to require the knowledge of a soothsayer. What clinicians use, instead, is the science of clinical epidemiology. The second edition of the Guide to Clinical Preventive Services is a handy compendium of this science as applied to clinical preventive medicine in 1996.
The first edition,2 published in 1989, was a landmark work, evaluating 169 preventive interventions and grading each on the