Abstract
Purpose: Medication abortion (MAB) provision by family physicians has the potential to expand abortion access. However, there are documented individual and structural barriers to provision. This study investigates how family physicians in the United States (US) navigate the barriers impeding abortion provision in primary care.
Methods: We conducted a qualitative study on the experiences of US family physicians with MAB in primary care. We recruited participants at national conferences and via professional networks. This analysis focuses on the experiences of the subset of participants who expressed interest in providing MAB.
Results: Forty-eight participants met inclusion criteria, with representation from all 4 regions of the US. Participants had diverse experiences related to abortion provision, training, and the environment in which they practice, with a third of participants working in states with hostile abortion policies. We categorized participants into 3 groups: (1) doctors who did not receive training and do not provide abortions (n = 11), (2) doctors who received training but do not provide abortions (n = 20), and (3) doctors who received training and currently provide abortions (n = 17). We found that training, administrative and community support, and internal motivation to overcome barriers help family physicians integrate MAB in primary care practices. Federal and state laws, absence of training, stigma around abortion provision, inaccurate or limited knowledge of institutional barriers, and administrative resistance all contributed to doctors excluding abortion provision from their scope of practice.
Conclusion: Improving medication abortion provision by family physicians requires addressing the individual and system barriers family physicians encounter so they receive the education, training, and support to successfully integrate abortion care into clinical practice.
- Contraception
- Drug-Induced Abortion
- Family Medicine
- Family Physicians
- Health Policy
- Health Services Accessibility
- Mifepristone
- Patient-Centered Care
- Primary Health Care
- Qualitative Research
- Reproductive Health
- Scope of Practice
Introduction
Although abortion is a common medical experience, growing legal restrictions in the United States (US) have decreased availability of abortion services across the country. According to the Guttmacher Institute, 89% of counties in the US do not have a known clinic that provides abortion services.1 In 2000, the Food and Drug Administration (FDA) approved mifepristone for medication abortion (MAB). This approval created the potential to integrate abortion services into primary care,2 which abortion advocates hoped would normalize abortion services and broaden access.3⇓⇓⇓⇓⇓⇓⇓–11
Unfortunately, such an expansion of abortion provision in primary care has not occurred. Prior studies document that US family physicians face extensive barriers to providing abortion services, including absence of training, institutional restrictions, costly liability insurance, unsupportive professional and peer networks, and fear of violence or harassment.2,3,12⇓⇓⇓–16 In addition the Hyde Amendment, which prevents federal funds from being used for abortion, serves as a barrier for clinics receiving federal funding (eg, Federally Qualified Health Centers and Veteran's Affairs clinics) and the FDA's stringent Risk Evaluation and Mitigation Strategy (REMS) criteria on mifepristone limit who can prescribe mifepristone and where it can be dispensed.17,18 As a result, abortion services in the US remain largely confined to specialized abortion clinics, with 95% of abortions taking place within dedicated abortion clinics in 2017.1
Despite knowledge of barriers to abortion provision, we know little about how individual family physicians navigate this landscape, and why some ultimately are successful at providing abortion services and others are not. This study uses qualitative interviews with a national set of family physicians who are already providing or interested in providing MAB to understand the experiences of family physicians related to MAB training and provision.
Methods
Sample
This study draws on a larger project exploring how family physicians consider MAB provision in relationship to the core values of the specialty. We used a multi-pronged recruitment approach to obtain a geographically diverse sample of family physicians. We recruited participants at family medicine conferences and from national listservs. We used professional networks to purposively sample family physicians from states with more hostile abortion policies and to identify family physicians who had successfully integrated abortion services into primary care or had experience in leadership in family medicine.
Inclusion criteria were being either a new career family physician or a family medicine thought leader. We defined new career family physicians as individuals who completed a family medicine residency in the US within the last 10 years. We identified thought leaders as family physician experts with experience motivating other family physicians to expand their scope, or family physicians with experience integrating abortion into family medicine. Because the overarching goal of the project was to develop values-based communication strategies to motivate family physicians to provide MAB, we excluded individuals opposed to abortion, based on their response to, “Are you personally opposed to people getting abortions?”
In a presurvey, participants ranked their interest in providing MAB on a scale of 1 to 10, 1 being not at all interested and 10 being extremely interested. For the purposes of this analysis, we included individuals scoring a 6 or higher to develop an understanding of the experience of barriers among those interested in providing abortions to focus the analysis and tailor recommendations to integrating MAB to those who want to provide this care.
Data Collection
We developed an interview guide with input from the research team, which included family physicians, educators, advocates, a social and behavioral scientist, and a communications specialist. Interview questions explored family physicians' perspectives and experiences related to MAB provision in family medicine practice, drawing on the Theory of Planned Behavior (TPB), a well-described approach to understanding influences on behavior.19 The research team iterated the interview guide after initial interviews and added probes based on emergent areas identified through participants' responses to questions about their views on abortion and family medicine values.
Research staff obtained verbal informed consent and participants completed surveys with questions about demographics, training, and clinical experience ahead of the interviews. Three team members (CD, EF, SW), trained in qualitative interview methods and with experience working in reproductive health, conducted the interviews. Interviews took place either in-person or virtually over video conferencing software and lasted 60 to 75 minutes. All interviews were audio-recorded. We compensated participants for their time with $100 gift cards. Recruitment ended when data saturation was reached, and no new information emerged regarding the primary study's themes. The study was approved by the UCSF Institutional Review Board (#18-26392).
Analysis
Audio recordings were transcribed verbatim and deidentified by a HIPAA-compliant professional transcription service and research team members reviewed transcripts to ensure accuracy. Members of the research team read transcripts, discussed themes with the entire study team, and developed a preliminary codebook based on components of the TPB including norms, perceived control, and intentions.
Two researchers (SW, IS) double coded an initial set of transcripts using NVivo 12 to assess intercoder agreement, clarify codes, and resolve disagreements. Through this iterative process the research team revised the codebook and inductively added additional codes not captured by TBP. Three members of the study team (CP, NR, SW) coded an approximately equal number of transcripts and met regularly to achieve consensus on the interpretation and meaning of codes and resolve any discrepancies.
Researchers took a deductive-inductive content analysis approach and used memos to identify broad impressions and additional themes.20 After coding a transcript, a memo was drafted to document and describe impressions of the interview, which became more structured to highlight key domains over the course of the process. Through regular meetings to discuss major themes, researchers clarified the central attitudes and challenges shaping family physicians' perspectives on providing MAB.
Results
Between January and October 2019, we conducted 56 in-depth semistructured interviews with family physicians in the US, with 48 interview participants meeting the inclusion criteria for this analysis of expressing an interest of ≥6 in providing abortion services on the presurvey (Table 1). Of the included physicians, 41 were early career family physicians and 7 were physician thought leaders. Most of the included participants, 65% (n = 31) were not currently abortion providers and 35% (n = 17) were currently abortion providers. Through the analysis process, we divided participants into 3 main groups: 1) physicians who did not receive abortion training and do not provide abortions (n = 11), 2) physicians who received abortion training (medication and/or surgical abortion) but do not provide abortions (n = 20), and 3) physicians who received abortion training and currently provide abortions (n = 17). Participant quotes are labeled by their study number (P#), geography, and the state's abortion policy landscape.21
Group 1: Absence of Training and No Provision
Family physicians who did not receive abortion training and were not abortion providers discussed their lack of training as the most significant barrier to abortion provision. As one participant explained to us, “I myself do not [provide abortions], but that is… more of a reflection of lack of training. It is not by no means anything I am philosophically opposed to” (P32, West, Supportive).
Other participants agreed that while family physicians should provide abortions, the lack of universal training made them uncomfortable with prescribing MAB. For example, one physician working in a Southern state felt “completely uneducated on the approach.” As a result, she felt unable to prescribe the medication necessary for a medication abortion. As she explained to us: “I do not prescribe any medications that I do not understand” (P35, South, Hostile). Another physician emphasized the absence of training on abortion provision as a unique problem that family physicians face:
I treat diabetes, I treat thyroid medication, why wouldn't I provide that service? … I think that the challenge is - is that unlike diabetes, which pretty much all family medicine physicians have like bought into and can agree like, okay, this is a chronic disease, we need to manage it, these are the typical ways we do it, abortion does not necessarily elicit that same universal approach (P30, West, Supportive).
The absence of training on medication abortion provision prevented this group from possessing the technical skills to provide abortions. Physicians in this group shared that their lack of exposure to abortion during training solidified for them the exclusion of abortion from family physicians' scope of practice. Furthermore, their lack of training resulted in this group of physicians not being exposed to or aware of the systemic barriers physicians face in integrating medication abortion into primary care.
These physicians, more than participants in other groups, mentioned how location – with subsequent state politics, hospital system restrictions, and poor community support—dissuades them from providing abortions.
I think it's just the region I live in… I mean if I lived in an area where it was routine to provide those services, it would be definitely different than living somewhere where, where you're not able to (P13, South, Hostile).
Group 2: Received Training but Do Not Provide
Family physicians who received abortion training but do not provide abortions primarily described external barriers that prevented or made it difficult to offer abortion in their clinical practice. Unlike the first group of physicians who lacked technical skills and knowledge around institutional barriers, the second group generally had more concrete knowledge of state and institutional restrictions preventing them from providing MAB. For example, one participant discussed how the state laws where she practiced restricted her ability to provide.
There are a ton of… ridiculous abortion restriction laws that happened. Like, you have to have a sign to a certain dimension in the waiting room…to be able to perform abortions…If I lived anywhere else that was more liberal and had better state laws, I would a hundred percent be doing this [providing abortions] (P15, Midwest, Hostile).
Another participant similarly focused on the legal barriers on abortion provision that made it impossible to provide abortion in her clinic: “The setting of my clinics is not an ambulatory - quote unquote, ambulatory surgical center. So, based on state laws… I would not be able to provide it in my family medicine clinic” (P85, Midwest, Hostile).
Many of the physicians in this group chose to work in Federally Qualified Health Centers (FQHCs). Due to federal restrictions such as the Hyde Amendment, that prevents federal funds from being used for abortion, these physicians felt a lack of administrative support to provide MAB. One physician we spoke with expressed her frustration with not being able to provide abortions at her FQHC, despite robust training during residency:
The red tape to be able to provide abortions in an FQHC setting is just so enormous… Like why would they [clinic administration] do that when they could just send them to a Planned Parenthood? …[I]t's just become so normative that abortions occur at specialized clinics that a lot of administrators in the primary care world feel like there isn't a huge reason to get into the abortion business (P6, South, NA).
This group also brought up the logistic barriers imposed by the REMS criteria on mifepristone. As one participant lamented, “I would be doing it in a heartbeat if I could prescribe mifepristone, and my patient could pick it up at a commercial pharmacy, but she cannot because of the way it is regulated by the FDA” (P6, South, NA).
Although most physicians in this group were frustrated by not being able to provide, some felt that they did not need to provide abortion services because patients could access care at local reproductive health clinics. As one physician described:
Part of why I haven't, like, tried really hard to integrate this into my practice is that I have somewhere to send the patients… [if that weren't true] that would be more motivating for me to push myself to do something that I'm not already doing (P26, West, Supportive).
Finally, a few participants in this group, including two thought leaders, had provided abortion care since residency but were not providing abortions at the time of the interview due to competing demands from their current leadership roles and other clinical responsibilities.
Group 3: Received Training and Provide Abortion Services
Most of the physicians who provided abortions (n = 10, 59%) in the study were not providing MAB in primary care settings, and many split their time working between primary care and reproductive health clinics to provide abortions.
As with group 2, some physicians worked at institutions with policies prohibiting abortion provision at their sites or in states with hostile abortion policies that made it more difficult to provide. For example, leadership at one participant's clinic created a policy prohibiting abortion care at their site after hiring her, knowing that she worked part-time at an abortion clinic. Another participant was hired by an academic Family Medicine department for her reproductive health expertise with the goal of expanding their abortion training for residents. After she joined faculty, the Obstetrics and Gynecology (OBGYN) department prevented her from providing abortion care and training.
Two participants who provided abortions in a reproductive health clinic felt like they were too junior to take on the process of getting their primary care clinics to provide MAB. One participant who previously lost a job because her primary care employer found out she was providing abortions at another clinic shared, “I would really need to establish myself there,” before she would be willing to try and integrate abortion care (P1, Midwest, Supportive). Another participant who had recently graduated from residency reflected on how residents never received training on how to navigate the barriers to integrate MAB in primary care.
Like in residency or in medical school there's a lot of just like, oh, learn to do this procedure…. But there's not a lot of training in what are the systemic barriers or issues that you would face if you were trying to actually do this procedure in your office (P9, West, Supportive).
Seven participants were providing abortions in primary care, of which 5 exclusively provided in primary care and 2 provided in primary care and reproductive health settings. Four of these participants joined practices where MAB was already integrated. One participant (P52) had almost integrated MAB into her primary care clinic at the time of the interview and confirmed afterward that she was providing MAB to patients. All participants who provided MAB in primary care settings were either based in the Northeast or West and none were providing abortions in states with hostile abortion restrictions.
Participants who successfully integrated MAB into primary care described facilitators that helped them to provide. Those who provided abortions in FQHCs shared how persistence and supportive administration contributed to their ability to provide in that setting.
It's something my organization is pretty deeply committed to. But places where there aren't big champions, I think those extra costs or those extra administrative hassles may not…the people that run other organizations may not see it as worthwhile (P10, Northeast, Supportive).
A thought leader who started working on MAB integration before mifepristone was even approved by the FDA, shared how she continued to provide MAB at her FQHC despite numerous barriers:
There was always another barrier, another barrier. But, you know, by then, we were already offering abortions. We were offering the pills and the MVAs in the office. And, you know, that seemed like the right thing to do and it seemed like the right thing to advocate for (P88, Northeast, Supportive).
In addition to understanding the laws and policies around MAB provision, participants described the importance of having physicians who were committed to providing MAB and willing to overcome challenges.
Well, you really have to have a champion. You really have to have someone who's just gonna dig in and work at it and work at it and work at it (P88, Northeast, Supportive).
P52 lived in a rural town where abortion was difficult to access and described how her community inspired her to integrate MAB into her primary care clinic:
[In] at least half of the hospital board meetings somebody stands up and talks about how important it is to have reproductive choice… a couple community members would even call the clinic manager and try to demand time to talk about how to implement this… It's [community advocacy that was] really key to having me finally pull it together and realize that whatever the risk is, I can't just keep ignoring my community's need and I need to do something (P52, West, Supportive).
P52 and P8, the only new career family physician in our sample who successfully integrated MAB into primary care, both initiated the process of integrating MAB by sharing their intentions with their colleagues. They discussed how not having seniority at an organization impacted their approach. P52 waited until she had been at the clinic for a few years to begin conversations around providing MAB, whereas P8 initiated the conversation earlier and got approval to provide MAB from OBGYN leadership.
In my first year or two here, I just didn't wanna buck the system… Because I knew it was going to alienate me from her [the clinic director]. And I wanted to establish that I'm competent and a go getter (P52, West, Supportive).
You know, I'm pretty new…I'm a small fish in a big, big pond. I'm not gonna make any waves. So, I took what they gave me, and I was like, all right, I'll do IUDs and endometrial biopsies and medication abortions, and I'll get to the rest later (P8, Northeast, Neutral).
P52 faced opposition from antiabortion leadership. Despite unsupportive leadership, P52 joined hospital committees to change the policy.
I finally had to create a policy and procedures for the addition of this service line… And I eventually clawed my way onto the medical executive committee for the hospital (P52, West, Supportive).
P52 partnered with advocacy organizations and worked with the Reproductive Health Access Project to address each barrier that came up. P8 leveraged his relationship with an OBGYN colleague who helped him expand his scope and eventually also provide aspiration abortions to his patients. When leadership changed at P52's institution, she moved forward with the policies and procedures she had been working on and started prescribing mifepristone to her patients.
Discussion
Our study provides insights into how family physicians navigate providing medication abortion, with respect to training, the presence of institutional and policy barriers, and the context in which they practice. Participants described challenges including lack of training, federal restrictions such as the Hyde Amendment and REMS criteria on mifepristone, state laws restricting where abortion services can be provided, and the absence of institutional support. By examining where different groups of family physicians meet roadblocks, our analysis explores how individuals navigate barriers and ultimately succeed or fail to overcome them. Drawing on our study's findings, we outline interventions to address these multi-level barriers to improve abortion services in primary care.
First, education remains an essential first step for family physicians to be able to provide abortions. Participants who did not receive training or exposure during medical school or residency felt unable to provide abortion services. There remains an ongoing need to strengthen education on abortion in family medicine residencies.22⇓–24 Prior research emphasizes that exposure not only improves the likelihood to provide abortions, but also strengthens contraception counseling.9,25,26
Although training is critical for abortion provision, our findings highlight that for many family physicians, such as those in group 2, clinical training and interest in providing MAB is not sufficient. Recent survey data from the American Board of Family Physicians similarly highlights the persistent gap between training and provision of abortions: 13% of participants felt prepared to provide abortion care after their training but only 3.7% currently provide abortions.22,23
Some early career family physicians, including those who provide abortion care in reproductive health settings, felt unprepared and too junior to take on integrating MAB into a new setting. Study participants demonstrated a need for continued support after residency and training during residency on how to navigate state, federal, institutional, and logistic challenges to integrate MAB into primary care. Several organizations specifically target these system level barriers and exposing family physicians and trainees to them could further support integration of abortion services. For example, ExPAND Mifepristone is a learning collaboration providing evidence-based knowledge on the clinical use of mifepristone and expertise on how to navigate the administrative barriers to abortion provision in primary care settings.27 The Reproductive Health Access Project (RHAP) provides training, resources, and support in sexual and reproductive health (including contraception, abortion, and miscarriage management) for clinicians post formal training. We also found a lack of understanding by family physicians of if and how FQHCs can provide abortion care. FQHCs that currently provide abortion care can serve as models and physicians at these clinics can share best practices on how to overcome challenges of providing abortions in FQHCs and the REMS criteria for mifepristone prescribing.28
Based on our findings, facilitators to integrating MAB into primary care include states with neutral or supportive abortion policy, institutions where abortion is not prohibited, supportive leadership, community support, allies — including advocacy organizations and other physicians—and internal motivation to overcome the barriers to MAB provision. Not all facilitators are needed to successfully integrate MAB and study participants demonstrated how physicians can influence their settings and change policy. In places with state and institutional restrictions, family physicians can play a role in advocating for legal and policy changes to support patients getting the care they need.
Because of the significant institutional barriers, and the availability of specialized abortion care settings, some family physicians felt they personally did not need to offer abortion services within family medicine. This represents a self-perpetuating cycle, in which the marginalization of abortion services into reproductive health clinics justifies the exclusion of abortion from primary care for these physicians. Further education on how providing abortion in family medicine can reduce barriers, decrease stigma, and help normalize abortion within health care is needed.
Our study has several strengths. First, our purposeful sampling provided a broad sample of early career family physicians across the US. This highlights that despite differences in clinical and geographic setting, there are consistent themes these physicians encounter in how they navigate MAB provision. Second, insights into the three main groups we characterize can help build an agenda to tailor interventions with the goal of improving training and access of medication abortion. A key limitation in our study was the difficulty identifying early career family physicians who integrated medication abortion into primary care clinics. Our use of the TPB framework may also be an additional limitation as it may have constrained participants' responses, however we did iterate the interview guide throughout the recruitment process to capture emerging themes. Although we successfully recruited geographically diverse participants who could share the varied experience of abortion provision across the United States, our sample was not racially or ethnically diverse. This likely reflects the ongoing challenge of racial diversity in the physician workforce and the need for more efforts to recruit and retain students and faculty from diverse racial and ethnic backgrounds, particularly those providing abortion care.29⇓–31
Family physicians remain the first contact for many individuals with the health care system. The unique position of family physicians as individuals who center care around building relationships and trust can have a profound impact in increasing access to abortion. Our study highlights the persistent individual, system, and policy level barriers family physicians face in integrating medication abortion into primary care. Improving medication abortion provision by family physicians requires addressing the individual and system barriers family physicians encounter so they receive the education, training, and support to successfully integrate abortion care into clinical practice.
Acknowledgments
We would like to thank Shelly Rodrigues and Rassidatou Konate for their input on the study. Thank you to the many physicians who spoke with us and shared their time, insights, and persistence.
Notes
This article was externally peer reviewed.
Funding: The research reported in this publication was funded through the Society of Family Planning Research Fund (grant award number SFPRF12-MA9).
Conflicting and Competing interests: None declared.
To see this article online, please go to: http://jabfm.org/content/35/3/579.full.
- Received for publication July 1, 2021.
- Revision received October 29, 2021.
- Revision received November 6, 2021.
- Revision received December 3, 2021.
- Accepted for publication December 9, 2021.