Competition between spontaneous and induced abortion.

Because of their similar timing in pregnancy, spontaneous and induced abortions can be viewed as competing outcomes. Some intended abortion operations are anticipated by earlier miscarriages while some potential miscarriages are forestalled by earlier deliberate interruptions of pregnancy. Previous treatments of this competition are reviewed, and a new analysis is made on the basis of New York data. A simple rule for approximating the reduction in apparent incidence of spontaneous abortion in the presence of induced abortion is presented. The effects of nonreporting and misreporting of pregnancy outcomes upon this rule are examined by means of the Perrin-Sheps renewal process. Two expectations are tested on Taiwanese data.