To determine the effectiveness and safety of repeated precoital and postcoital use of levonorgestrel for pregnancy prevention.
We searched eight computerized databases for studies that evaluated oral hormones taken for contraception immediately before or after each coital act during one or more menstrual cycles. We reviewed reference lists and contacted experts.
METHODS OF STUDY SELECTION:
Of 31 studies identified, we excluded 16 from this review because they contained incomplete information on the drug regimen, number of pregnancies, or duration of follow-up or because they did not evaluate levonorgestrel. The remaining 15 studies collectively included 20 groups of women treated with levonorgestrel: 10 groups received 0.75 mg and 10 received other doses.
TABULATION, INTEGRATION, AND RESULTS:
Two authors independently extracted data from the included studies. We calculated Pearl indices from individual studies and pooled data and estimated confidence intervals (CIs) using a Poisson distribution. Most studies had methodological limitations. The studies of the 0.75-mg levonorgestrel dose included 2,628 patients who used the method for 1,033 woman-years and yielded a pooled Pearl index of 5.1 pregnancies per 100 woman-years (95% CI 3.8–6.7). The studies of other doses added 5,787 patients who used the method for 53,347 cycles. The pooled Pearl index for all studies was 4.9 pregnancies per 100 woman-years (95% CI 4.3–5.5). The main side effect was frequent menstrual irregularities. No serious adverse events were reported. Most women liked the pericoital method.
The data suggested that pericoital oral levonorgestrel is safe and moderately effective. However, the quality of the studies was suboptimal. A pressing need exists for rigorous research to evaluate pericoital use of levonorgestrel as a primary means of contraception.