Risk of Transmission of HTLV-III by Needle Stick

Abstract

To the Editor: Hirsch et al. (Jan. 3 issue)¹ suggest that there is little additional risk of transmission of human T-cell lymphotropic virus III (HTLV-III) by accidental needle sticks. However, further statistical analysis of their data raises questions about such a conclusion.The authors base their conclusion on their finding of a zero numerator — that is, the occurrence of "no events" in looking for HTLV-III antibodies in health workers with accidental needle sticks. Inferences about the size of a risk in studies with zero numerators may be calculated with the "rule of three."² This rule states that if the . . .