Strengths and limitations
A key limitation is the low response rate, which appeared to decrease over time. If the likelihood of responding was related to the patient’s experience and outcome, then this would be a source of bias in our results. A loss to follow-up is to be expected and is a persistent challenge within abortion research given the high level of stigma associated and the legally restricted settings(8, 23).  Other limitations include the use of self-reported data. However, given that the sample was overwhelmingly in legally restricted settings and in an informal healthcare setting, using self-reported data is the most feasible method(21). We could not adjust for known confounders, such as gestational age, as this information was not available in this dataset(22). However, as WoW limits the availability of the service to women who are less than 9-10 weeks gestation, the the abortions in this dataset are expected to have taken place in the first trimester. To our best knowledge, this study provides the largest evaluation of the outcome of self-managed abortions provided by telemedicine. Furthermore, it includes the largest sample of women in legally restricted settings. Outcomes are reported in consistency with prior studies for comparability. In addition, we included outcome measures which have never been looked at before such as prior knowledge about medical abortion.