Spontaneous abortion incidence in the treatment of infertility.

Analysis of published series indicates that, irrespective of the method of data collection, close agreement exists on empirically derived incidences of spontaneous abortion in North American populations, provided that age, previous abortion history, and gravidity are controlled. The normal incidence of clinically apparent abortion among first pregnancies in women under age 30 years is in the range 8.3% to 11.0%. A comprehensive analysis of published series on pregnancies after infertility treatment indicates that only three therapeutic methods are attended by an abortion incidence that approaches this asymptote: ovulation induction with bromocriptine in hyperprolactinemic anovulation (11.8%; n = 1,233 pregnancies); artificial insemination with donor semen for azoospermia (11.4%; n = 326 first pregnancies); and operation for endometriosis (9.3%; n = 768 pregnancies). Abortion incidences accompanying other modes of therapy are higher. Because increased abortion incidence is not generally recognized as a specific reproductive difficulty in infertile couples, as are the other two: i.e., refractory infertility despite technically adequate therapy and ectopic pregnancy, plausible physiologic mechanisms for abortions in specific categories of disease or treatment type are described and discussed in detail. Moreover, abortion incidence is proposed to be a sensitive and objective parameter with which to assess distortions in human reproductive physiology, especially when competing methods of infertility treatment have overall pregnancy outcomes that are thought to be similar.