Prevalence and clinical correlates of unsuccessful use of drugs to induce menstruation.

Despite its prohibition, illegal abortion is widely practiced in Brazil, with important adverse health consequences. This report aims to document the prevalence and correlates of the unsuccessful use of drugs to "induce menstrual flow" in a cohort of pregnant Brazilian women. In a cross-sectional study, 6,102 pregnant women between gestation weeks 21 and 28 were interviewed in prenatal clinics of the Brazilian National Health Care System from April 1991 to November 1995. When asked "In order to know if you were pregnant, did you take any medication to induce menstrual flow", 874 (14.4%) responded "yes." The most frequently used drugs were herbal teas (41%), estrogens and/or progestogens (30%), and misoprostol (16%). As demonstrated through logistic regression analysis, independent correlates of such use were unplanned pregnancy (odds ratio [OR] 4.3), low educational attainment (OR 3.3), absence of a husband or partner (OR 1.8), number of children (one or more) (OR 1.5), a history of a previous induced abortion (OR 1.4), and use of oral contraceptives at the time (OR 1.4). Misoprostol use occurred in 2.2% of pregnancies, and showed a very strong association with an unplanned pregnancy (OR 16.0), absence of a husband or partner (OR 3.5), and a history of a previous induced abortion (OR 2.2). It was not associated with a history of menstrual irregularity. In contradistinction, the use of medroxyprogesterone was strongly associated with previous menstrual irregularity (OR = 5.0). The use of drugs and teas, many of which are unknown in terms of fetal risk, in early pregnancy to induce menstrual flow is quite common in women in the Brazilian National Health System. Although the objective of such drug use appears to be varied, analysis of the clinical correlates of use suggest attempted abortion to be the principal aim.