Abstract
The article by Schoenbaum et al. (Sept. 28 issue)* provides some very interesting demographic observations about some people in a section of New York City who are positive for the human immunodeficiency virus (HIV). There is, however, at least one seriously flawed conclusion. The authors write, "Factors with significant and independent associations with the presence of HIV antibody included black race [and] Hispanic background." This statement is a new version of the yellow-fingers-cause-lung-cancer school of statistical conclusion. African American and Hispanic people are disproportionately represented among intravenous drug users and would be expected to be overrepresented among the HIV-positive patients the authors studied.

This publication has 4 references indexed in Scilit: