Characteristics of people obtaining medication vs procedural abortions in clinical settings in the United States: Findings from the 2021–2022 Abortion Patient Survey

Abstract

ObjectivesThis study aimed to compare the characteristics of people obtaining medication and procedural abortions in clinical settings in the United States. Study designWe conducted a cross-sectional survey of people obtaining abortions at a sample of 56 US facilities between June 2021 and July 2022. We restricted analyses to patients with pregnancies less than 11 weeks’ gestation presenting at 43 clinics offering both medication and procedural abortion. We conducted bivariate analyses and multivariable logistic regression to identify factors associated with obtaining a medication vs procedural abortion. ResultsOur analytic sample includes 4717 respondents, 57% of whom obtained a medication abortion. In bivariate analyses, individuals who identified as Asian or White, had no prior births or abortions, or were paying out of pocket were all more likely to have a medication abortion. Non-Hispanic Black individuals, those with incomes at or below the poverty level, and those paying with insurance were more likely to have a procedural abortion. Some 24% of respondents chose the facility because it offered medication abortion, but even after controlling for this proxy for method preference in a logistic regression model, Black respondents and those with poverty-level incomes were less likely to have a medication abortion. ConclusionsThe findings of the study suggest that Black individuals and those with low incomes—who often face systemic barriers to care—are less likely to have medication abortions. When medication abortion is the only option available, for example, at a medication-only clinic or from an online source, these groups may be most impacted by the lack of options. ImplicationsTo the extent possible, offering both medication and procedural abortion and increasing access to both types are important to meet patients’ individual needs and preferences.