Background The advent of medical abortion has improved access to safe abortion procedures. Medical abortion procedures involve either administering mifepristone followed by misoprostol or a misoprostol-only regimen. The drugs are commonly administered in the presence of clinicians, which is known as provider-administered medical abortion. In self-administered medical abortion, drugs are administered by the woman herself without the supervision of a healthcare provider during at least one stage of the drug protocol. Self-administration of medical abortion has the potential to provide women with control over the abortion process. In settings where there is a shortage of healthcare providers, self-administration may reduce the burden on the health system. However, it remains unclear whether self-administration of medical abortion is effective and safe. It is important to understand whether women can safely and effectively terminate their own pregnancies when having access to accurate and adequate information, high-quality drugs, and facility-based care in case of complications. Objectives To compare the effectiveness, safety, and acceptability of self-administered versus provider-administered medical abortion in any setting. Search methods We searched Cochrane Central Register of Controlled Trials, MEDLINE in process and other non-indexed citations, Embase, CINAHL, POPLINE, LILACS, ClinicalTrials.gov, WHO ICTRP, and Google Scholar from inception to 10 July 2019. Selection criteria We included randomized controlled trials (RCTs) and prospective cohort studies with a concurrent comparison group, using study designs that compared medical abortion by self-administered versus provider-administered methods. Data collection and analysis Two reviewers independently extracted the data, and we performed a meta-analysis where appropriate using Review Manager 5. Our primary outcome was successful abortion (effectiveness), defined as complete uterine evacuation without the need for surgical intervention. Ongoing pregnancy (the presence of an intact gestational sac) was our secondary outcome measuring success or effectiveness. We assessed statistical heterogeneity with Chi(2) tests and I-2 statistics using a cut-off point of P < 0.10 to indicate statistical heterogeneity. Quality assessment of the data used the GRADE approach. We used standard methodological procedures expected by Cochrane. Main results We identified 18 studies (two RCTs and 16 non-randomized studies (NRSs)) comprising 11,043 women undergoing early medical abortion (>= 9 weeks gestation) in 10 countries. Sixteen studies took place in low-to-middle income resource settings and two studies were in high-resource settings. One NRS study received analgesics from a pharmaceutical company. Five NRSs and one RCT did not report on funding; nine NRSs received all or partial funding from an anonymous donor. Five NRSs and one RCT received funding from government agencies, private foundations, or non-profit bodies. The intervention in the evidence is predominantly from women taking mifepristone in the presence of a healthcare provider, and subsequently taking misoprostol without healthcare provider supervision (e.g. at home). There is no evidence of a difference in rates of successful abortions between self-administered and provider-administered groups: for two RCTs, risk ratio (RR) 0.99, 95% confidence interval (CI) 0.97 to 1.01; 919 participants; moderate certainty of evidence. There is very low certainty of evidence from 16 NRSs: RR 0.99, 95% CI 0.97 to 1.01; 10,124 participants. For the outcome of ongoing pregnancy there may be little or no difference between the two groups: for one RCT: RR 1.69, 95% CI 0.41 to 7.02; 735 participants; low certainty of evidence; and very low certainty evidence for 11 NRSs: RR 1.28, 95% CI 0.65 to 2.49; 6691 participants. We are uncertain whether there are any differences in complications requiring surgical intervention, since we found no RCTs and evidence from three NRSs was of very low certainty: for three NRSs: RR 2.14, 95% CI 0.80 to 5.71; 2452 participants. Authors' conclusions This review shows that self-administering the second stage of early medical abortion procedures is as effective as provider-administered procedures for the outcome of abortion success. There may be no difference for the outcome of ongoing pregnancy, although the evidence for this is uncertain for this outcome. There is very low-certainty evidence for the risk of complications requiring surgical intervention. Data are limited by the scarcity of high-quality research study designs and the presence of risks of bias. This review provides insufficient evidence to determine the safety of self-administration when compared with administering medication in the presence of healthcare provider supervision. Future research should investigate the effectiveness and safety of self-administered medical abortion in the absence of healthcare provider supervision through the entirety of the medical abortion protocol (e.g. during administration of mifepristone or as part of a misoprostolonly regimen) and at later gestational ages (i.e. more than nine weeks). In the absence of any supervision from medical personnel, research is needed to understand how best to inform and support women who choose to self-administer, including when to seek clinical care.
