The Effectiveness of the Yuzpe Regimen of Emergency Contraception

TLDR

A review of 10 clinical trials shows that conventional failure rates for the Yuzpe emergency contraception are unreliable because they rely on assumptions about lost follow‑up and equal risk, leading researchers to adopt an effectiveness rate that incorporates both observed and expected pregnancies. The effectiveness rate is calculated as the proportionate reduction in conception probability by comparing observed and expected pregnancies for each menstrual cycle day. The Yuzpe regimen (200 mcg ethinyl estradiol + 2 mg norgestrel) produced a pooled failure rate of 1.5 % and an effectiveness rate of 74 %, demonstrating high contraceptive efficacy.

Abstract

A review of the 10 clinical trials of the Yuzpe method of emergency contraception that reported the data required to calculate effectiveness rates suggests that this may be a more accurate measure of efficacy than the failure rate. The Yuzpe regimen, which involves the administration of 200 mcg of ethinyl estradiol and 2.0 mg of norgestrel, was associated with failure rates ranging from 0.2% to 2.8%; the pooled rate was 1.5% (95% exact confidence interval, 1.2-1.9%). The equality of failure rates across studies was compromised by two assumptions: women lost to follow-up (as high as 22%) became pregnant at the same rate as women observed, and all women in the trials had an equal probability of failure. The effectiveness rate--the proportionate reduction in the probability of conception caused by emergency contraception use--avoids these sources of error by including data on both the observed and expected number of pregnancies and computing the risk of conception for each day of the menstrual cycle. These estimates range from 55.3% to 94.2%, with a pooled effectiveness rate of 74.0% (95% exact confidence interval, 68.2-79.3%). On the other hand, four methodological issues are inherent in use of the effectiveness rate: the assumption of homogeneity implicit in pooling observations, bias introduced by the unknown pregnancy rate among women lost to follow-up, the probability some women violated study protocol and had more than one unprotected act of intercourse during their cycle, and possible underestimation of the expected number of pregnancies.