The Yuzpe regimen 1,2 of emergency contraception is approved for the prevention of pregnancy when treatment is started up to 72 hours after unprotected intercourse. The 72-hour cutoff, however, is neither evidence based nor convincing biologically. 3 Even if effectiveness declines the longer after unprotected sex a woman begins the therapy (as some observational studies of emergency contraception timing suggest 4,5), it seems unreasonable that effectiveness would drop to zero precisely at 72 hours. A large new study of the closely related levonorgestrel-only regimen 6 showed that that method was effective when started up to 120 hours after unprotected intercourse.
Certainly, even if starting therapy after 72 hours confers less protection than more prompt initiation, some women could still benefit. Women might need more than 72 hours to obtain emergency contraception, particularly when facing a prescription requirement, as in the United States. They may not be able to get a doctor's appointment quickly, or might need the therapy over the weekend or a holiday when their doctor is not available. Young women, who are more likely to have unprotected sex, may be especially reluctant to go to a new health care provider to seek emergency contraception.
The standard therapy for women seeking emergency contraception more than 72 hours after unprotected sex is copper intrauterine device (IUD) insertion. Intrauterine devices, however, can cost $400 7 and have other drawbacks, so many women refuse them. In many places, access to experienced providers who can insert IUDs may be difficult, and many providers who do insert IUDs do not provide them to young or nulliparous women. Even if hormonal emergency contraception is less effective than an IUD after 72 hours has elapsed, some women might still prefer to try this therapy. The purpose of our study was to determine whether the window for hormonal emergency contraception can be extended to 120 hours.
MATERIALS AND METHODS
In parallel to a double-masked randomized controlled trial of two modifications to the Yuzpe regimen of emergency contraception, 8 we conducted a prospective, observational study to evaluate the effectiveness of the Yuzpe regimen when started between 72 and 120 hours after unprotected intercourse (ie, on days 4 or 5) among women refusing copper IUD insertion. We compared failure rates observed in this cohort to rates among otherwise similar women assigned to the control group of the companion randomized trial. Women in the observational study were enrolled at the same clinics as in the randomized trial (Iowa, United States; Liverpool, Glasgow, Edinburgh, and Sheffield, United Kingdom) during the same period (September 1997 through February 2000). Eligibility criteria were the same as for the randomized trial: age 16–45 (18–45 in the United States), regular menstrual cycles between 21 and 35 days in length, having had unprotected intercourse between 10 days before and 6 days after the estimated day of ovulation (defined as typical cycle length minus 13) in their current cycle, willingness to abstain from further unprotected intercourse during the study cycle, willingness to come for follow-up and fill in a diary of side effects, and refusal of a copper IUD. We excluded women who tested positive using urine pregnancy tests, were breast-feeding, had not had two normal periods before presenting to the clinic, or had used any hormonal contraception in the last 2 months.
All women received 100 μg of ethinyl estradiol plus 1.0 mg of norgestrel, repeated after 12 hours. Women were asked to return for a follow-up visit approximately 1 week after the expected start of their next menses.
We compared perfect-use and typical-use failure rates and effectiveness rates among women in the Yuzpe control and the days 4–5 groups. We used StatXact-3 (Cytel Software Corp., Cambridge MA) to compute P values for differences in failure rates based on Fisher exact tests and 95% binomial confidence intervals for failure rates. P values for differences in effectiveness rates and 95% confidence intervals for effectiveness rates were based on z tests. Our small sample size of 111, however, gave us just 25% power to detect a doubling in the failure rates (2% to 4%) and 59% power to detect a tripling in the failure rates (2% to 6%).
The Population Council's institutional review board and local ethical review committees at each site approved the protocol. All women gave written informed consent.
Women enrolled in the observational days 4–5 cohort were similar to those in the standard Yuzpe control arm from the randomized study (Table 1). In both groups, the mean adjusted cycle day of unprotected sex was approximately the same (adjusted cycle day 13). Women in both groups were in their early 20s (mean age 22 years in the days 4–5 group and 24 years in the standard Yuzpe control group). Body mass index was similar (mean 24 in the days 4–5 group and 23 in the standard Yuzpe group). Three of the 116 women in the days 4–5 group and four of the 696 women in the standard Yuzpe control group turned out later to have been enrolled by mistake (primarily because their cycle day of exposure was incorrectly calculated); none became pregnant, but all were retained in our intention-to-treat analyses. Five (4.3%) of the women in the days 4–5 cohort were lost to follow-up, as were 21 (3.0%) of the women in the standard Yuzpe control group. The analysis of effectiveness is based on the remaining 111 and 675 women in the two groups.
There were four pregnancies observed in the days 4–5 group, compared with 17 pregnancies in the standard Yuzpe control group. Based on a review of ultrasound information, serial pregnancy tests, reports from women's physicians, and from the women themselves, we determined that two and 12 of these pregnancies, respectively, occurred despite perfect use of the method, whereas the other two and five pregnancies, respectively, occurred during imperfect use. When in doubt, pregnancies were classified as occurring during perfect use.
The failure rates during perfect use were 1.9% and 2.0% in the days 4–5 group and in the standard Yuzpe control group, respectively (Table 2). Typical-use failure rates were 3.6% in the days 4–5 group and 2.5% in the standard Yuzpe control group, respectively. The failure rates during either perfect or typical use for the days 4–5 group relative to the standard Yuzpe control group did not statistically differ, but the sample sizes in the days 4–5 group were small and the confidence intervals are correspondingly large.
Table 2 also shows effectiveness when the therapy is started on days 1–3 and on days 4–5, respectively, during both perfect and typical use. Differences between the standard Yuzpe control group and the days 4–5 group are not statistically significant. During perfect use, taking the Yuzpe regimen on days 1–3 and on days 4–5 after unprotected intercourse prevented 73% and 77% of expected pregnancies, respectively. Again, however, the small sample in the latter group leads to large confidence intervals around the point estimates.
Our study suggests that the 72-hour cutoff after unprotected intercourse for starting the Yuzpe regimen of emergency contraception is unnecessarily restrictive. This result is consistent with the finding from a recent, similar Canadian study, 9 as well as with a study of levonorgestrel recently published by the World Health Organization. 6 Women who present seeking emergency contraception later than the 72-hour cutoff should be counseled that insertion of a copper IUD is the most effective way to prevent pregnancy. 10 If they do not want an IUD or an IUD is contraindicated or not available, however, they should be offered hormonal therapy. The Yuzpe regimen initiated 4–5 days after unprotected sex reduces the risk of pregnancy substantially.
Unfortunately, enrollment in the observational piece of our study proceeded much more slowly than expected, and we were able to enlist only 116 women (we had planned for more than 600). Perhaps our small, slow enrollment arose from the fact that most participants enrolled at clinics in the United Kingdom, where information about emergency contraception is widespread and knowledge of the 72-hour time limit is relatively high. The small numbers in the cohort led to wide confidence intervals around our estimates of failure and effectiveness rates, as was the case with the Canadian study. 9 Nevertheless, the point estimates themselves provide evidence of at least partial effectiveness when the Yuzpe regimen of emergency contraception is started after 72 hours after unprotected sex.
1. Yuzpe AA, Lancee WJ. Ethinylestradiol and dl-norgestrel as a postcoital contraceptive. Fertil Steril 1977;28:932–6.
2. Yuzpe AA, Percival Smith R, Rademaker AW. A multicenter clinical investigation employing ethinyl estradiol combined with dl-norgestrel as a postcoital contraceptive agent. Fertil Steril 1982;37:508–13.
3. Grou F, Rodrigues I. The morning-after pill–how long after? Am J Obstet Gynecol 1994;171:1529–34.
4. Task Force on Postovulatory Methods of Fertility Regulation. Randomised controlled trial of levonorgestrel versus the Yuzpe regimen of combined oral contraceptives for emergency contraception. Lancet 1998;352:428–33.
5. Piaggio G, von Hertzen H, Grimes DA, Van Look PFA. Timing of emergency contraception with levonorgestrel or the Yuzpe regimen. Lancet 1999;353:721.
6. Von Hertzen H, Piaggio G, Ding J, Chen J, Song S, Bartfai G, et al. Low dose mifepristone and two regimens of levonorgestrel for emergency contraception: A WHO multicentre randomized trial. Lancet 2002;360:1803–10.
7. Trussell J, Koenig J, Ellertson C, Stewart F. Preventing unintended pregnancy: The cost-effectiveness of three methods of emergency contraception. Am J Public Health 1997;87:932–7.
8. Ellertson C, Webb A, Blanchard K, Bigrigg A, Haskell S, Shochet T, et al. Modifying the Yuzpe regimen of emergency contraception: A multicenter randomized controlled trial. Obstet Gynecol 2003;101:1160–7.
9. Rodrigues I, Grou F, Joly J. Effectiveness of emergency contraception pills between 72 and 120 hours after unprotected sexual intercourse. Am J Obstet Gynecol 2001;184:531–7.
10. Trussell J, Ellertson C. Efficacy of emergency contraception. Fertil Control Rev 1995;4:8–11.