Steroids and Peptic Ulcer

Publisher Website
Google Scholar
Abstract
To the Editor: We believe that the conclusions made by Conn and Blitzer (N Engl J Med 294:473–479, 1976) in their article on the "nonassociation" of steroid therapy and peptic ulcer are incorrect. Although they claim to produce "impressively negative" evidence against the "steroid-ulcer myth," the data that they present actually provide statistically significant evidence that steroids increase the risk of proved peptic ulcer. Unfortunately, the authors do not define the statistical methods that they employed to label the differences between control and steroid groups "insignificant." Since we are concerned with whether steroids increase ulcer risk, a one-sided test is . . .