SAN FRANCISCO — A review of 11 studies examining the impact of increased access to emergency contraceptive pills finds no evidence that this increased access reduces the number of pregnancies or the number of abortions at a population level, James Trussell, Ph.D., reported at a meeting on contraceptive technology sponsored by Contemporary Forums.
In an often-cited 1992 paper, Dr. Trussell of Family Health International, Research Triangle Park, N.C., used modeling to suggest that access to emergency contraception could prevent half of all unintended pregnancies and abortions in the United States (Fam. Plann. Perspect. 1992;24:269–73).
But in nine randomized trials and one cohort study in which a total of 11,830 women were enrolled, and in one demonstration project in which 17,831 women were given emergency contraceptive pills (ECPs), not one of the studies found clinically or statistically significant differences between intervention and control groups in pregnancy or abortion.
Dr. Trussell noted that although many of these studies had substantial flaws, including small sample size, huge losses to follow-up (up to 62%), weak interventions, good access to ECPs among the comparison group, a low baseline risk of pregnancy, and a lack of randomization, none of the studies had all of those problems and some were very good.
“The consistency of these findings is hard to ignore,” Dr. Trussell said.
Dr. Trussell considered several other possible explanations for the failure of access to ECPs to reduce unintended pregnancies and abortions. Some have suggested that easy availability of emergency contraception would increase risk taking among women. But the studies found no evidence of an increase in unprotected sex or decrease in use of regular contraception with enhanced ECP access. Furthermore, two of the studies fail to find any increase in sexually transmitted infections with increased ECP access, also arguing against the increased-risk-taking hypothesis.
A third hypothesis is that emergency contraceptive pills have intrinsically low efficacy. The labeling on Plan B (two doses of levonorgestrel) quotes an 89% reduction in pregnancy risk after a single act of coitus, and estimates of risk reduction in the published literature range from 60% to 94%. But Dr. Trussell called all these estimates into question because it's difficult to accurately estimate the expected number of pregnancies that would have occurred without emergency contraception.
Expected pregnancies are calculated by determining the day of the menstrual cycle when the coital act occurred, and by using published charts listing probabilities of pregnancy by cycle day. But those charts might not apply in this population since they were constructed from data in women who wanted to become pregnant. Women using ECP don't want to become pregnant, and that could lead to possible differences in fecundity, the frequency of unprotected sex, and the accuracy of self-reports. Dr. Trussell concluded that the numbers of expected pregnancies reported by studies are probably too high and that most published efficacy figures are probably overestimates. Nevertheless, it's clear that emergency contraceptives produce physiologic effects that are incompatible with pregnancy and that at the very least they are more effective than nothing.
The fourth hypothesis is that even when they have easy access to ECPs, women don't use them sufficiently. The studies showed that repeated use of ECP was uncommon, that many unprotected acts remained uncovered by ECPs, and that no ECPs were used in most pregnancy cycles.
Women in the studies cited several reasons for non-use including a failure to perceive pregnancy risk, a lack of motivation to use emergency contraception, forgetfulness, and inconvenience. (Dr. Trussell expressed amazement that some women cited inconvenience as an excuse in view of the fact that they had free, study-supplied ECPs in their possession.)
Expense and side effects further reduce ECP use in the real world. Dr. Trussell concluded that insufficient use is definitely a problem and likely contributed to the failure of ECPs to reduce unintended pregnancies.
As for how physicians should respond to this new information, Dr. Trussell urged honesty. Physicians should not oversell emergency contraception by implying that Plan B will reduce unintended pregnancy at a population level, he said.
Contemporary Forums and this news organization are both wholly owned subsidiaries of Reed Elsevier.