Obstetrician-gynecologist willingness to provide medication abortion with removal of the in-person dispensing requirement for mifepristone

Medication abortions have been steadily increasing in recent years, accounting for nearly 40% of abortions in the United States []. However, access to timely abortion care is hampered by various state and federal policies, including the Food and Drug Administration's Risk Evaluation and Mitigation Strategy (FDA REMS) for mifepristone, the primary drug used for medication abortion []. The mifepristone REMS requires the drug be dispensed directly to the patient, by a certified provider in a clinic, medical office, or hospital []. As such, clinicians who want to provide medication abortions must stock the drug in their practice, adding up-front costs and administrative burdens to clinics [,]. This requirement may be particularly burdensome to patients in the Midwest, South, and in rural areas because those regions have a limited number of abortion providers [].

Obstetrician-gynecologists (ob-gyns) are well situated to provide timely abortion care as they may be the first clinicians to confirm an unintended pregnancy. A national survey of ob-gyns from 2016 to 2017 found that 72% had a patient in the prior year who needed or wanted an abortion; however, only 24% provided abortion care in their practice []. The survey also found that the mifepristone in-person dispensing requirement was a barrier to provision []. The purpose of this analysis was to explore physician and practice setting characteristics associated with willingness to provide abortion with removal of the in-person dispensing requirement.

We used data from a cross-sectional survey conducted between August 2016 and March 2017 of a national sample of practicing ob-gyns, which we have previously described [,]. We invited 2500 Fellows and Junior Fellows of the American College of Obstetricians and Gynecologists (ACOG) to respond to an online survey on “selected ob-gyn practices.” The sample included 1,000 members of a demographically representative research network of practicing ACOG members (the Collaborative Ambulatory Research Network) and 1500 non-network members selected using a proportionate stratified sample by geographic distribution using ACOG districts. Of note, the prior published national analysis of ob-gyn medication abortion provision focused only on the research network sample [].

For this analysis, our primary outcome measure examined willingness to provide medication abortions if the mifepristone in-person dispensing requirement were not in place. Specifically, respondents who reported having patients seeking abortion care in the prior year but had not performed medication abortions were asked: “Currently, if you want to provide medical abortion, you must stock the medications in your office. Would you offer medical abortion to your patients if you could write a prescription for mifepristone and misoprostol, and your patient could obtain both medications at a pharmacy?” Response options were “yes,” “no,” and “not sure.”

We limited our analyses to ob-gyns who reported they “had any patients in the last 12 months who wanted or needed an abortion or termination of pregnancy.” We calculated abortion provision in the previous year by physician characteristics (age, gender, race and ethnicity, years in practice, and research network membership) and practice characteristics (region, practice setting, and practice location). We estimated willingness to provide medication abortion among those that did not provide medication abortion in the previous year, as well as the theoretical percent increase in medication abortion provision if the in-person dispensing requirement were removed. We performed bivariable and multivariable logistic regression of willingness to provide medication by physician and practice characteristics among all respondents that did not provide medication abortion in the previous year, including those that did not respond to the willingness question. We present adjusted odds ratios (aOR) and 95% confidence intervals (CI) by practice characteristics adjusting for gender and years in practice. We removed age group from our model as it was collinear with practice years and included all variables from the bivariate analyses that had a p-value <0.20. Reference category was selected based on sample sizes or meaningful comparison groups. To assess generalizability of our findings, we conducted a sensitivity analysis by limiting our sample to the demographically representative research network. We performed statistical analysis using Stata 16 (StataCorp LP, College Station, TX). The Allendale Investigational Review Board approved this study. We provided additional information on sample design and study procedures, as well as variables for physician and practice setting characteristics, in previous publications [,,].

Our final sample of 1,280 respondents (52% response rate; 62% for research network members vs. 42% for non-network members) was restricted further to respondents who reported having abortion-seeking patients in the last 12 months (n = 868, 68% of sample). Among those, 31% (n = 267) provided abortion in the past year, including 19% (n = 164) who provided medication abortion (Table 1). Among the 704 ob-gyns that did not provide medication abortion in the previous year, 24% (n = 171) reported that they would provide medication abortions if they could write a prescription for mifepristone and misoprostol, and their patient could obtain both medications at a pharmacy. The additional 171 ob-gyns expressing willingness to provide medication abortion if the in-person dispensing requirement were removed represents a potential doubling (+104%, 95% CI: 97%−112%) of the number of medication abortion providers from the existing 164. The largest theoretical increases in ob-gyns providing medication abortion were in the Midwest (+189%, 95% CI: 172%−207%) and South (+118%, 95% CI: 103%−134%) (Table 1). In the multivariable logistic regression, female ob-gyns, those working in a university faculty or “other” practice (compared to in a partnership or group), and those who were members of the research network who did not provide medication abortion in the past year had higher odds of reporting they would provide the method if the mifepristone in-person dispensing requirement were removed; those who had been in practice >10 years (compared to those in practice ≤10 years) had lower odds of reporting they would start providing (see Table 2). A sensitivity analysis including only the research network sample yielded results that were consistent with our primary analysis (data not shown).

This analysis indicates that provision of medication abortion by ob-gyns would increase if the in-person dispensing requirement for mifepristone were removed, and this increase would be seen across all regions of the country and in all practice settings. Ob-gyns in the Midwest, South, and West were just as likely as those in the Northeast to report willingness to start providing medication abortion if the dispensing requirement were removed. This finding is notable given that ob-gyns in the Midwest and South were significantly less likely to currently provide abortion compared to those in the Northeast []. In fact, the largest theoretical increases in ob-gyn medication abortion providers were seen in the Midwest and South.

It is important to note that an additional 22% (n = 155) of those who had abortion-seeking patients but did not provide medication abortion indicated that they were “not sure” if they would offer medication abortion if the dispensing requirement were removed. As such, it is possible that we have underestimated the theoretical increase in medication abortion provision.

This analysis has several limitations. Our primary outcome is based on a hypothetical question and may not reflect actual practice if the policy were changed. Our unweighted findings and non-response bias may limit generalizability. In addition, the survey was conducted in 2016 to 2017, and physician perspectives may have changed over time. In particular, the COVID-19 pandemic has seen a marked increase in telemedicine [], and ob-gyns might be more interested in providing medication abortion without an in-person visit. Alternatively, ob-gyns focused on providing their usual care during the pandemic may be less willing to introduce a new service in the near future, even if the dispensing requirement were removed.

The in-person dispensing requirement for mifepristone codified in the drug's REMS is a barrier to clinician provision of the method. Removing this requirement could increase the number of medication abortion providers across the country, including in settings with limited access.

The authors declare no conflict of interest.

This research was supported by the Society of Family Planning Research Fund (SFPRF), grant SFPRF7-9. The views and opinions expressed are those of the authors and do not necessarily represent the views and opinions of SFPRF. Additional support was provided by the Maternal and Child Health Bureau (Title V, Social Security Act, Health Resources and Services Administration, and Department of Health and Human Services), grant UA6MC19010. The funding sources had no role in the study design; the collection, analysis, and interpretation of the data; nor the preparation, writing, or submission of this manuscript.

The authors thank Kate Grindlay for her assistance in preliminary data cleaning and management.