Home-based medication abortions may be as effective as those administered in a clinic, and they may be more acceptable to women, according to an analysis of pooled data from nine studies in which women took a combination of mifepristone and misoprostol to induce abortion.1 Misoprostol, the second drug administered in a two-part medication abortion regimen, is commonly taken at home by women obtaining such abortions in the United States and France. Past research has established that its administration at home may have several advantages over clinic-based protocols, including allowing for greater privacy and lessening the burden on both women and service providers by reducing the number of clinic visits. However, evidence on the comparative safety, effectiveness and acceptability of home-based and clinic-based administration had not previously been reviewed. Analysis of the pooled data showed that an average of 90% of home-based procedures and 93% of clinic-based procedures resulted in a complete abortion; the difference was not statistically significant. Compared with women who took misoprostol at a clinic, women who took it at home experienced pain for 0.3 days longer and were more likely to be open to using the method again (odds ratio, 7.1).

The data used in the analysis come from prospective cohort studies that had been published in 1990 or later, compared outcomes of home-based and clinic-based medication abortions, assessed outcomes prospectively and reported the proportion of abortions that were completed successfully. The nine studies that met these criteria provided data on a total of 4,522 women (3,478 home users and 1,044 clinic users) aged 15–49 in Albania, France, India, Nepal, Tunisia, Turkey and Vietnam. A dose of 400 mcg of misoprostol was administered in all cases, while mifepristone dosages varied between 200 mg (in seven studies) and 600 mg (in two). All women were given the choice of taking misoprostol at home or at the clinic and were advised to use painkillers. Women were followed up 10–20 days after mifepristone administration.

The average age of study participants ranged from 24.7 to 32.2 years. The maximum length of gestation was 49 days in two studies and 56 days in seven studies.

Among women who took misoprostol at home, the proportion who had a complete abortion ranged from 86% (in India) to 97% (in Albania); the average success rate was 90%. Among those who took misoprostol at a clinic, between 80% (in Turkey) and 99% (in France) had a complete abortion, and the average success rate was 93%. Although three studies had found that success rates were higher among women who received clinic-based abortions than among those who had home-based abortions, and two studies had come to the opposite conclusion, the difference in rates based on pooled data was not significant.

Pain and cramping were assessed in three studies; in each, at least 90% of women reported this side effect. Pooled results showed that pain lasted significantly longer among women who took misoprostol at home than among those who took it at a clinic (weighted mean difference, 0.3 days). Vomiting, also assessed in three studies, was reported by 12–34% of women and also lasted 0.3 days longer among home users. Reports of other complications did not differ according to the medication abortion protocol that was used.

Across eight studies, an average of 88% of women who took misoprostol at home and 86% who took it at a clinic were satisfied with the medication abortion method. An average of 94% of home users and 62% of clinic users in all nine studies reported that they would choose medication abortion again. The difference between home- and clinic-based users was significant only for the second measure (odds ratio, 7.1).

The researchers explain several limitations of their analysis, most of which stem from flaws inherent in the studies on which the review is built. For instance, only one study adjusted for potential confounders, and none was able to minimize selection bias through randomization. They also acknowledge that the success of a home-based protocol may depend on the medical resources available: “The effectiveness of home-based [medication] abortion in non-research settings without the precautionary measures and support systems that were most likely in place in these and other studies, essential for compliance with ethical norms, may be less satisfactory.” Yet, the researchers point out, where appropriate support is available, “simplified protocols could give greater access to [medication] abortion to women living in restrictive and/or resource-limited settings where mortality related to unsafe abortion remains high.”—H. Ball