A New Approach to Estimating the Efficacy of Medical Abortion

Allison A. Hedley, Princeton University
Charlotte Ellertson, Ibis Reproductive Health
James Trussell, Princeton University
Abigail Norris Turner, Population Council
Beverly Winikoff, Population Council

We propose and implement a new method for evaluating the efficacy of medical abortion. Previous research on the efficacy of medical abortion has followed the conceptual model for the analysis of surgical abortion; the failure rate is computed by simple division and expressed as a proportion. While this method of calculating efficacy is appropriate for the discrete event of surgical abortion, it is inappropriate for medical abortion because it fails to account for the temporal process of medical abortion. We assert that life tables should be used to calculate the efficacy of medical abortion because they incorporate the important dimension of time in success and create unbiased estimates of efficacy. Using data on more than 6,500 women from six previously conducted medical abortion studies, we use life table procedures to calculate the efficacy of medical abortion and hazard models to assess the impact of various demographic characteristics on efficacy.