To the Editor.—We were most interested in Maki's1 review of the literature on risk factors for nosocomial infection. Maki refers to the work of Craven et al2 as implicating histamine2 (H2) receptor antagonists as a risk factor. This work has weaknesses apparently not taken into consideration by Maki.
Initially, Maki refers to the uncontrolled trial by Craven et al2 on patients receiving continuous mechanical ventilation. In this study, pneumonia developed in 18% (24/134) of patients receiving antacids alone, 38% (18/48) in patients receiving antacids and cimetidine, and 36% (4/11) of patients receiving only cimetidine. There were not enough patients in the trial to estimate the true incidence of pneumonia in the cimetidine populations; the 95% confidence intervals for the cimetidine alone sample were so large (8% to 64%) that the true incidence in this population could well be less than in the population taking antacids.