Abortion Training in Family Medicine Residency Programs: A National Survey of Program Directors 5 Months After the Dobbs Decision

Over the past 30 years, some family medicine residency programs (FMRPs) have integrated early abortion training into their curricula. This training increases residents’ knowledge and skills^{1, 2} and is associated with higher rates of abortion provision after graduation. Studies have found that 24% ³ and 27% ^{4, 5} of family physicians trained at programs with routine abortion training provided abortion after residency, compared to 13% of those trained at similar programs without abortion training ³ and only 3% in a large representative survey conducted by the American Board of Family Medicine. ⁶

Early abortion provision is well within the scope of family medicine, with family physicians routinely providing contraception, prenatal care, and miscarriage management. Furthermore, abortion provision within family medicine is acceptable to patients, with a majority stating they would prefer abortion care with their primary care physician rather than at a freestanding clinic. ^{7, 8}

However, abortion training in FMRPs varies widely. It may be exclusively didactic or include hands-on training; be routine or elective; include only medication abortion (MAB) or also aspiration abortion; and occur only in high-volume clinics or also within the family medicine setting.⁹ While roughly a third of FMRPs in a 2011 study ⁹ reported some abortion training, less than 10% had routine training; but the data did not distinguish between training in different types of abortion. An update is overdue because much has changed in the last 13 years: notably, advances that make MAB easier to integrate into routine care^10-13 and the political landscape of abortion.

After the Dobbs v Jackson Women’s Health Organization Supreme Court decision ¹⁴ removed the constitutional right to abortion in June 2022, abortion training became at risk in unprecedented ways. As of August 2023, 29% percent of FMRPs were in states with bans or very restrictive policies.¹⁵ Understanding the current status of abortion training within FMRPs and projecting implications for the future is vitally important should the legal landscape shift in those states where abortion is currently threatened.

We aimed to determine the scope and details of abortion training at FMRPs throughout the United States and to assess associations between abortion training and program-level characteristics. In addition, we aimed to learn about didactic abortion training, clinical training in miscarriage management, abortion provision within the FM setting, and program directors’ attitudes on routine abortion training. A national analysis on the state of abortion training in family medicine allows family medicine educators, reproductive health advocates, and curricular experts to identify gaps and create strategies for the future.

Our questions were part of a larger omnibus survey conducted by the Council of Academic Family Medicine Educational Research Alliance (CERA). The methodology of the CERA Program Director Survey has been described previously in detail.¹⁶ The CERA steering committee evaluated our questions for consistency with the overall subproject aim, readability, and existing evidence of reliability and validity. Pretesting was done on family medicine educators who were not part of the target population. We modified questions following pretesting for flow, timing, and readability. The project was approved by the American Academy of Family Physicians Institutional Review Board in November 2022. Data was collected from November 16 to December 18, 2022.

The sampling frame for the survey was all directors of US FMRPs accredited by the Accreditation Council for Graduate Medical Education as identified by the Association of Family Medicine Residency Directors. Email invitations to participate were sent with a survey link generated by the online program SurveyMonkey (SurveyMonkey Inc). Five follow-up emails to encourage nonrespondents to participate were sent before the survey closed. The total number of program directors (PDs) was 722 at the time of the survey. Nine email addresses were undeliverable, leaving 713 invitations delivered. Thirty-five PDs indicated that they did not meet eligibility criteria because they had not had three resident classes, reducing the sample size to 678.

Drawing on resources from the Guttmacher Institute ¹⁷, ¹⁸ and the Center for Reproductive Rights ¹⁹, we grouped states and the District of Columbia into three categories based on level of restrictiveness as of December 2022, the time of data collection. States in the first category had banned abortion or had no facilities providing abortion (AL, AR, ID, KY, LA, MS, MO, ND, OK, SD, TN, TX, WI, WV), those in the second category had some restrictions and/or a ban that was enjoined (AZ, FL, GA, IA, IN, NC, NE, OH, SC, UT, WY), while the third group of remaining states were those in which abortion access was protected and/or expanded.

We began our analysis by examining descriptive statistics for all variables to understand the abortion and reproductive health training offered at FMRPs. Following that, we used χ² goodness of fit tests to assess whether differences among programs and residency PDs were significantly related to the presence and types of abortion training available. For the multivariate analysis, performed with binary logistic regression, we included those variables in our model that previous research led us to hypothesize would have a relationship with abortion provision and/or those associated with provision on a bivariate level. Statistical significance was set at two-sided α=0.05. All data analysis was conducted using SPSS version 27 (IBM).

The overall response rate for the survey was 42.2% (286/678). Responding programs were fairly equally distributed regionally and were primarily community-based and university-affiliated, with most sized between 19 and 31 residents and located in communities between 75,000 and 500,000 people (Table 1). More than half of programs were in a state where some protections existed for abortion care, and 17% of programs were Catholic-affiliated. PDs were close to evenly split between male and female; most had been in their current role less than 5 years, and three-fourths were White. Just under half believed that at least some routine abortion training should be standard in FMRPs.

While complete demographic data for nonresponding programs was not available, we found no significant differences in the sampling frame on regional distribution and abortion-restrictedness when comparing respondents and all programs, with P values of .64 and .52 for regional distribution and restrictiveness, respectively. We also compared the percentage of Catholic-affiliated respondents (16.8%) to recent research on FMRPs (14%) ²⁰ and found no significant differences (P=.27). Finally, Ray Biggs of the Society of Teachers of Family Medicine shared membership data from the Association of Family Medicine Residency Directors (email, August 24, 2023) on 560 of the 678 PDs, and we found no significant differences between the gender and race/ethnicity of their 560 PDs and that of the PDs in this sample (P=.70 and P=.26, respectively).

Medication abortion training was more common than aspiration abortion training (Table 2). Programs with routine aspiration training were primarily a subset of those with routine MAB training; almost all (28/30) programs with routine aspiration abortion training had routine MAB training, but only about half (28/53) of the programs with routine MAB training had routine aspiration training.

Elective training was also far more common than routine training. Nineteen percent had routine MAB training and 10% had routine aspiration training, compared to the 58% with elective MAB training and 52% with elective aspiration training. Nearly all programs with routine training in a skill also had elective training, typically for those who desired increased proficiency.

While training in medical management of miscarriage was fairly common, at 48% of programs, aspiration for miscarriage was less commonly taught, offered only at 21% of responding programs. As expected, abortion training and miscarriage training were closely related. Ninety-eight percent of programs with routine MAB training had routine training in medical management of miscarriage, compared to 43% of those with elective MAB training only and 31% of those without any MAB training (P<.001). Patterns were similar for aspiration for miscarriage: 90% of programs with routine aspiration abortion training offered aspiration miscarriage training, compared to 18% of programs with elective aspiration abortion training and 9% of programs with no aspiration training (P<.001).

Bivariate analysis clearly showed that region was a strong predictor of abortion training, with programs in the Northeast and West far more likely to have both routine and elective training (Table 3). Furthermore, and relatedly, the more restrictive the abortion laws in a state were, the less likely programs were to offer abortion training.

Routine abortion training was more common at larger residency programs and at programs with no Catholic affiliation, and residency programs located in urban areas were significantly more likely to have any abortion training.

When the PDs believed that abortion training should be standard, the program was far more likely to offer both routine abortion training and any type of abortion training. However, we found no other associations with the presence of either routine or elective training and PD demographics.

In adjusted analysis, multiple factors were significantly associated with the presence of routine abortion training (Table 4). In this model, residency programs without a Catholic affiliation, located in states where abortion access was protected, and with PDs who believed abortion training should be routine, were far more likely to have routine abortion training. Larger residency programs, those with more than 31 residents, were also more likely to have routine training. Although region and community size were associated with abortion training on a bivariate level, this trend was not evident in multivariate analysis. The latter model was robust when testing alternate and additional variables of interest, such as community size and PD characteristics, but these variables did not ultimately contribute to the overall strength of the model.

Because PDs’ beliefs on abortion were strongly associated with the presence of abortion training, we also examined factors associated with the belief that training should be standard. As with abortion training, both region and level of abortion restrictions played key roles (P<.001), with more PDs believing training should be standard in the Northeast and West, and in states with less restrictive laws. Community size also was associated (P=.033), with 35% believing training should be standard in communities with fewer than 75,000 people compared to 53% in larger communities, as was PD gender (P<.001), with 67% of female PDs supporting routine training compared to 31% of male PDs.

At 81% of programs, residents were not able to provide abortion in the continuity clinic (CC) or any other family medicine (FM) setting. The 53 programs offering provision in the CC/FM setting were split evenly between those that offered only medication abortion and those where residents provided both medication and aspiration abortion in that setting. Programs with routine training were far more likely to offer these services in the CC/FM setting. Seventy percent of programs with routine MAB training (vs 37% of those with only elective MAB training) and 50% of the programs with routine aspiration abortion training (vs 30% of the programs with elective aspiration abortion training) allowed residents to provide these services in the CC/FM setting.

Other factors associated with abortion provision in the CC/FM setting on a bivariate level included region, restrictiveness, and Catholic affiliation (Table 5). In addition, PD gender and belief in the importance of routine training were significantly associated with abortion training in the CC/FM setting.

More than 60% of responding FMRPs reported some training in abortion, and 19% reported at least one type of routine training. These data reveal a marked increase from the last similar research, published in 2011. ⁹ Overall, we found that MAB training was far more common than procedural abortion training, and elective training was far more common than routine training.

Multiple factors were associated with the presence of abortion training. Our results confirmed earlier research that found associations with Northeast and West regions, no Catholic affiliation, and larger program size. ^{9, 18, 21} We theorize that larger residency programs may be more likely to have training because having more faculty overall increases the likelihood that a faculty member at the program would be able and interested to train residents in abortion.

We were surprised to see a small number of PDs in ban states and at Catholic-affiliated institutions report that their programs had routine abortion training. This finding could be due to having misunderstood the question, having arranged for all residents to travel or rotate through another institution for training, or responding in one particular moment in a rapidly shifting political and legal climate.

While demographics of residency program directors had little association with training, programs whose directors believed that abortion training should be routine were significantly more likely to have both routine and elective abortion training. Although the direction of the association is unclear, because many seek employment in departments that align with their values, this finding may speak to the potential impact of a champion for abortion training in the PD role. However, at 28% of surveyed programs, PDs believed that abortion training should be routine, but their programs lacked abortion training; providing technical assistance, as RHEDI (Reproductive Health Education in Family Medicine) does, ²² is one way to assist these programs. Financial support to protect faculty time to institute abortion training, and development of partnerships with high-volume clinics to support resident training also would help ensure the success of these initiatives.

The largest proportion of training, and the largest area of growth, is in elective abortion training. While elective training can be a pathway to competency and postresidency provision, that training likely varies widely; no details are available on the length and content of training, nor the percentage of residents who participate. The category potentially includes primarily observational training and/or quite low volume abortion care, or merely the permission for residents to seek/attend outside opportunities on their own. Surprisingly, 50% of Catholic-affiliated programs and 41% of programs in states with abortion bans reported elective training. More research is needed to better understand exactly what those opportunities entail so that medical students who prioritize abortion training in their residency selection process could be better informed. Increasing substantive away elective training opportunities in abortion, such as those offered by the Midwest Access Project, ²³ would be helpful for both residents in restricted states and those in unrestricted states where limited training slots make accessing enough training to achieve competence difficult. Because many more residents apply to these programs than can be accommodated, ²⁴ the creation of additional training partnerships is key.

For programs without routine abortion training, training in miscarriage management can be a way to gain exposure to overlapping clinical skills. While not nearly enough to meet the need, 31% of programs without MAB training had training in medical management of miscarriage, and 9% of programs without aspiration abortion training had aspiration miscarriage training. Both types of miscarriage training can serve as foundations in the clinical skills of abortion care to increase the possibility of providing these types of care; projects such as the Miscarriage Care Initiative, led by the Reproductive Health Access Project, ²⁵ could be adapted for a residency context.

Abortion training within the family medicine context has been shown to lead to higher rates of postresidency abortion provision, both overall and within family medicine. ³ Providing abortion within the family medicine setting allows trainees to more easily integrate abortion provision into a postresidency FM context, as well as offers opportunities to reflect on systems-based practice. ²⁶ Because residents can provide abortion within the continuity clinic or another FM setting at only 30% of programs with abortion training, integrating abortion training and provision into these settings should be a key strategy moving forward. The low percentage may reflect the fact that many FM residency settings are Federally Qualified Health Centers (FQHCs), but this research was unable to examine that possibility. A common misperception is that FQHCs are not able to offer abortion due to the Hyde Amendment, ²⁷ when in fact abortion is possible if billing streams are kept clearly separated. ²⁸ Continuing to normalize abortion care as a routine part of FM through didactic teaching, practice, research, and advocacy in FM professional organizations such as the American Academy of Family Physicians can lay the cultural groundwork for more FMRPs to institute routine clinical abortion training.

Limitations to this study include the 42% response rate. Possibly programs with abortion training were more motivated to respond to the survey. However, the likelihood of response bias is reduced because the survey was an omnibus survey on several topics and the comparative analyses we were able to conduct found no significant differences between residency programs overall and our respondents. Additionally, “elective training” was not defined in detail, so respondents may have meant different things when indicating this training was available. Finally, because the legal landscape around abortion is rapidly evolving, this data captures only one particular moment. Subsequent rulings and laws likely will impact abortion training dramatically as well as abortion access generally; an additional 23% of FMRPs would be impacted if abortion became fully inaccessible in the states we categorized as “seriously restricted,” as described in the Methods section.

Particularly in this post-Dobbs moment, as differences in abortion access become more stark along socioeconomic and regional lines, abortion training at FMRPs is more crucial than ever. Clearly, barriers to instituting routine abortion training during residency persist, including lack of access to training sites, interspecialty conflict, ²⁹ and restrictions on medical liability insurance. ³⁰ In addition, collaboration with legal experts is necessary to clarify what is allowable in restricted states. However, particularly because family physicians are more likely to work in the underserved and marginalized communities most affected by abortion restrictions, ^{31, 32} shoring up the abortion-providing family medicine workforce could mitigate the human rights violations being perpetrated against pregnant individuals seeking abortion.