[Skip to Content]
[Skip to Content Landing]
April 17, 1972

Oral Contraceptives and Thromboembolic Disease

Author Affiliations

National Institute of Child Health and Human Development Bethesda, Md

JAMA. 1972;220(3):416-417. doi:10.1001/jama.1972.03200030074025

To the Editor.—  In our opinion, the data Drill used are not suitable for the analysis he undertook and, to the limited extent that they can be used, we interpret them differently.The author's argument is based on the observation that in the studies cited only 63 cases of thromboembolism were observed in drug users, whereas he expected to see 151 based on his review of previously published data. He describes this difference as statistically not significant, but no method of analysis is cited. We applied a rudimentary but appropriate test of significance that leads to an enormous level of chi-square and a probability far less than.001.Clearly the number of cases seen is significantly smaller than expected. Unless one is willing to infer that oral contraceptives are protective from thromboembolic disease (a hypothesis which Drill rejects) then either of two explanations must follow: (a) his estimate of the expected