American journal of epidemiology
-
The hypothesis that spontaneous abortion is associated with prior induced abortion is tested in an epidemiologic study. The reproductive histories of a consecutive series of women admitted to hospital with spontaneous abortions were compared with those of a control series of women who delivered after 28 weeks gestation. ⋯ The power of the analysis to detect a 30% excess in the frequency of induced abortion in cases as compared to controls is 82% at alpha = .05. The design and analysis effectually controlled for all potentially confounding factors.