Three regimens of emergency contraception--a single 10 mg dose of mifepristone, a single 1.5 mg dose of levonorgestrel and two 0.75 mg doses of levonorgestrel taken 12 hours apart-- appear to be equally effective. According to results of a randomized, double-blind trial among 4,071 women in 10 developed and developing countries who sought emergency contraception within five days of unprotected coitus,1 there was no significant difference in pregnancy rates by emergency contraceptive method. Only 1.5-1.8% of the women became pregnant and 77-82% of expected pregnancies were averted, with no significant difference by type of treatment. There was no significant difference by regimen in most side effects, except for the timing of expected menses after treatment; users of mifepristone experienced delays of at least seven days significantly more often than did users of levonogestrel.

The multicenter study was undertaken to compare the emergency contraceptive efficacy of levonorgestrel and mifepristone, and to determine the feasibility of a single dose of levonorgestrel. The study was conducted in 1998-2000 in 15 family planning clinics in 10 countries--China, Finland, Georgia, Hungary, India, Mongolia, Slovenia, Sweden, Switzerland and the United Kingdom. Women with regular menstrual cycles who presented at a clinic requesting emergency contraception within 120 hours of a single act of unprotected coitus were randomly assigned to one of the three regimens--one 10 mg dose of mifepristone, one 1.5 mg dose of levonorgestrel and two 0.75 doses of levonorgestrel taken 12 hours apart. They were asked to return for follow-up one week after the estimated onset of their next menstrual period; those who had not menstruated by that time or who had had an abnormal period were tested for pregnancy. The investigators assessed pregnancy rates, the proportion of expected pregnancies averted, participants' experience of side effects and delays in the return of menstruation.

A total of 4,071 women provided usable data for analysis, roughly one-third of whom were randomly assigned to each regimen. Fifty-four percent of participants were Chinese, 34% were white and 12% were of other races or ethnicities; their average age was 27 years. There were no significant differences between treatment groups in background characteristics. Overall, 60% of women had been pregnant and 48% had had an induced abortion, but these proportions varied widely by center. Fifty-two percent of women had asked for emergency contraception because they had not used any contraceptive method, 44% because a condom had failed, and 3-4% because another method had failed. Overall, 26% of participants had used emergency contraception before. Forty-four percent of the women presented for treatment within 24 hours of unprotected coitus, 72% did so within 48 hours, and 88% within 72 hours.

The three emergency contraception regimens had the same efficacy, even after the investigators controlled for treatment center and ethnicity. Overall, only 1.5-1.8% of women became pregnant. The regimens prevented 77-82% of pregnancies that would have been expected in the absence of emergency contraceptive use (a nonsignificant difference by treatment group). There were no significant differences in the risks of pregnancy between users of mifepristone and users of both levonorgestrel regimens combined, between users of mifepristone and users of single-dose levonorgestrel, and between users of one dose and users of two doses of levonorgestrel.

Among women in all treatment groups, there was no significant difference in pregnancy rates between those who took their pills within 72 hours of unprotected coitus and those who received treatment later. An increasing trend in pregnancy rates over the first five days after unprotected coitus was significant, but the sample sizes were too small to analyze this trend separately by treatment group. Among women who had unprotected intercourse after treatment but before expected menses, the rate of pregnancy was significantly higher among those who took mifepristone than among those who took levonorgestrel.

There were no differences by treatment group in the proportions of women reporting nausea, vomiting, diarrhea, fatigue, dizziness, headaches or breast tenderness. However, a significantly higher proportion of levo- norgestrel users than of mifepristone users reported bleeding unrelated to menstruation in the first week after treatment (31% vs. 19%). Overall, women in the developed countries reported more side effects than those in the developing countries. In addition, the proportion of women whose first menses after treatment began more than seven days later than expected was significantly higher among mifepristone users than among levonorgestrel users (9% vs. 5% among nonpregnant women only).

The researchers affirm that there were no differences in efficacy among the three treatment groups; moreover, within the limits set by the study, the single dose of levonorgestrel was at least as effective as the split dose. According to the investigators, because the risk of pregnancy continues after treatment, if women have further acts of unprotected intercourse, that risk "should be highlighted, especially if mifepristone is used"; contraceptive use should be encouraged if abstinence is unfeasible. Levonorgestrel's advantage of being associated with early menses means that users of this method will menstruate sooner, and be relieved of pregnancy anxiety sooner, than users of mifepristone. The researchers conclude that all three regimens "prevent a high proportion of pregnancies, even up to [five] days after coitus."--L. Remez


1. von Hertzen H et al., Low dose mifepristone and two regimens of levonorgestrel for emergency contraception: a WHO multicentre randomised trial, Lancet, 2002, 360(9348):1803-1810.