Among first-time expectant mothers, the risk of vaginal bleeding early in pregnancy is greater for those who have had a medication abortion than for those who have never had any type of abortion, according to a prospective cohort study of Chinese women.1 The risk was especially elevated among women whose medication abortion resulted in curettage or complications (odds ratios, 1.6 and 2.0, respectively). The risk of vaginal bleeding early in pregnancy did not differ between women who had had a first-trimester medication abortion and those who had had a first-trimester surgical abortion.
Vaginal bleeding is a sign of a high-risk pregnancy and has been associated with past surgical abortion. To explore whether a similar association exists for nonsurgical abortions, this study compared rates of vaginal bleeding among pregnant women who had had a medication abortion and those who had had a surgical or no abortion.
From 1998 to 2001, investigators recruited women who were 4–16 weeks pregnant from 83 antenatal clinics in Beijing, Chengdu and Shanghai. Women were eligible for the study if they were aged 20–34, had not given birth before and had previously undergone one first-trimester abortion (surgical or medication) or no abortion. Participants completed questionnaires at enrollment, at 28–30 weeks and at delivery, providing information on their reproductive and medical history, including the current pregnancy. Women’s reports of vaginal bleeding were examined overall and for two periods of pregnancy: before enrollment (the first period) and during follow-up (the second period). Log binomial regression analysis was used to calculate the relative risk of vaginal bleeding while controlling for study center, age, income, residence, season at conception and history of chronic disease.
A total of 14,399 women participated—of whom 4,841 had had a medication abortion, 4,705 a surgical abortion and 4,853 no abortion. The mean age at recruitment was 26. Most of the women had at least a high school degree; lived in a city; and were industrial or service workers, or farmers. Levels of tobacco and alcohol use, as well as of chronic disease, were low.
On average, women in the three groups enrolled in the study at 10–11 weeks’ gestation and had their initial follow-up at 29 weeks. Up to the first follow-up, rates of vaginal bleeding among women with a history of medication abortion, surgical abortion and no abortion were 17%, 17% and 14%, respectively. After adjustments for potential confounders, the risk of vaginal bleeding was significantly higher for women who had had a medication abortion than for those who had had no abortion (relative risk, 1.2). When the relationship between medication abortion and vaginal bleeding was examined according to period of pregnancy, the increased risk was observed only in the first period (1.3). However, no differences in the risk of vaginal bleeding were found between women who had undergone a medication abortion and those who had terminated their pregnancy surgically.
When rates of vaginal bleeding during the first period of pregnancy were compared by gestational age at recruitment, women who had had a medication abortion had a higher rate at every week of gestation than women who had not had an abortion. Moreover, their rates were similar to those of women who had had a surgical abortion. No differences in rates were observed among the three groups later in pregnancy.
Finally, the investigators stratified women who had had a medication abortion according to characteristics of the abortion and then compared their risk of vaginal bleeding before enrollment with that of women with no abortion history. The risk of bleeding was elevated to a similar extent among women who had had a medication abortion before age 25 and those who had had one later; the risk also was elevated among both women with an interpregnancy interval of less than 12 months and those with longer intervals (relative risks, 1.2–1.3). However, while women who had had a medication abortion during the first seven weeks’ gestation were more likely than those who had not had an abortion to report vaginal bleeding (1.3), no association between medication abortion and vaginal bleeding was apparent among women who had had a medication abortion after seven weeks’ gestation. The odds of bleeding were especially high among women who had had curettage or complications following a medication abortion (1.6 and 2.0, respectively); they were much lower among women who had not had curettage or complications, though they were still higher than the odds among women without a history of abortion (1.2 for each).
The investigators acknowledge that women who did not experience bleeding during their current pregnancy may have underreported prior abortions. They also note that given the young age and limited abortion experience of most study participants, the results may not apply to women who are older or have had multiple medication abortions. Nonetheless, the researchers conclude that the risk of vaginal bleeding early in pregnancy is more strongly associated with the previous use of mifepristone than with no abortion, especially when the medication abortion took place before seven weeks’ gestation and resulted in curettage or complications.
1. Yuan W et al., Mifepristone-induced abortion and vaginal bleeding in subsequent pregnancy, Contraception, 2011, 84(6):609–614.