Outcomes and Safety of History-Based Screening for Medication Abortion: A Retrospective Multicenter Cohort Study

IMPORTANCE: Screening for medication abortion eligibility typically includes ultrasonography or pelvic examination. To reduce physical contact during the COVID-19 pandemic, many clinicians stopped requiring tests before medication abortion and instead screened patients for pregnancy duration and ectopic pregnancy risk by history alone. However, few US-based studies have been conducted on the outcomes and safety of this novel model of care. OBJECTIVE: To evaluate the outcomes and safety of a history-based screening, no-test approach to medication abortion care. DESIGN, SETTING, AND PARTICIPANTS: This retrospective cohort study included patients obtaining a medication abortion without preabortion ultrasonography or pelvic examination between February 1, 2020, and January 31, 2021, at 14 independent, Planned Parenthood, academic-affiliated, and online-only clinics throughout the US. EXPOSURES: Medications for abortion provided without preabortion ultrasonography or pelvic examination and dispensed to patients in person or by mail. MAIN OUTCOMES AND MEASURES: Effectiveness, defined as complete abortion after 200 μg of mifepristone and up to 1600 μg of misoprostol without additional intervention, and major abortion-related adverse events, defined as hospital admission, major surgery, or blood transfusion. RESULTS: The study included data on 3779 patients with eligible abortions. The study participants were racially and ethnically diverse and included 870 (23.0%) Black patients, 533 (14.1%) Latinx/Hispanic patients, 1623 (42.9%) White patients, and 327 (8.7%) who identified as multiracial or with other racial or ethnic groups. For most (2626 [69.5%]), it was their first medication abortion. Patients lived in 34 states, and 2785 (73.7%) lived in urban areas. In 2511 (66.4%) abortions, the medications were dispensed in person; in the other 1268 (33.6%), they were mailed to the patient. Follow-up data were obtained for 2825 abortions (74.8%), and multiple imputation was used to account for missing data. Across the sample, 12 abortions (0.54%; 95% CI, 0.18%-0.90%) were followed by major abortion-related adverse events, and 4 patients (0.22%; 95% CI, 0.00%-0.45%) were treated for ectopic pregnancies. Follow-up identified 9 (0.40%; 95% CI, 0.00%-0.84%) patients who had pregnancy durations of greater than 70 days on the date the mifepristone was dispensed that were not identified at screening. The adjusted effectiveness rate was 94.8% (95% CI, 93.6%-95.9%). Effectiveness was similar when medications were dispensed in person (95.4%; 95% CI, 94.1%-96.7%) or mailed (93.3%; 95% CI, 90.7%-95.9%). CONCLUSIONS AND RELEVANCE: In this cohort study, screening for medication abortion eligibility by history alone was effective and safe with either in-person dispensing or mailing of medications, resulting in outcomes similar to published rates of models involving ultrasonography or pelvic examination. This approach may facilitate more equitable access to this essential service by increasing the types of clinicians and locations offering abortion care.