Antibiotic prophylaxis for herniorrhaphy and breast surgery

To the Editor: Our criticism of the paper by Platt et al. (Jan. 18 issue)¹ is based on four considerations: the statistical weakness of the data because of the small sample (alpha or Type I error²); inherent bias caused by overestimating the incidence of infection; changing surgical trends, which may diminish the clinical relevance of the data; and the potential hazards of widespread use of prophylaxis. The statistical weakness stems from the relatively small sample chosen.² The study was designed for an estimate of “base-line risk,” or an infection incidence rate of 10 percent in the placebo group —.