To the Editor.—We are worried that the estimates of risk of seroconversion provided by Lowenfels et al1 in the November 1989 issue of the Archives will lead to unnecessary and extreme concern on the part of the great majority of surgeons. These estimates were used at a recent regional meeting to justify a recommendation for testing all preoperative patients.
Our concern is based on the fact that the authors took care to determine the frequency of puncture injuries in surgeons but estimated prevalence in the population to be 5%. Prevalence in the population and frequency of punctures have an equal effect on risk of seroconversion. The authors do note that this is the prevalence in patients with trauma in some inner city populations; however, they do not indicate that this rate greatly exceeds the prevalence of infection with human immunodeficiency virus (HIV) in the general population. In a