Factors affecting willingness to provide medication abortion among North American Society for Pediatric and Adolescent Gynecology members caring for adolescents and young adults following the Dobbs decision

Abstract

Study objective:

To assess willingness to provide medication abortion among North American Society for Pediatric and Adolescent Gynecology (NASPAG) clinicians caring for adolescents and young adults (AYA) following Dobbs v. Jackson Women’s Health Organization

Design:

Cross-sectional online survey

Methods:

Potential participants received an e-mail invitation via the NASPAG listserv. A 43-item questionnaire queried demographics, practice setting, abortion training and practice, willingness to provide medication abortion, potential or real barriers to providing medication abortion, and sentiments of abortion. Descriptive statistics, χ², and Fisher’s exact tests were used.

Results:

Of the 70 participants, 51% were willing to provide a medication abortion for an adolescent who requested it in their clinical practice. The most common barriers to providing medication abortion were legislative restrictions (47%) and dispensing pills from clinic (33%). Participants’ willingness to provide a medication abortion differed by type of practice (p = .001), availability of mifepristone (p = .006), perception of state’s abortion policy (p = .001), concern about legislative restrictions (p = .008), experience providing abortion (p = .04), and receipt of medication abortion training (p = .02). Willingness to provide medication abortion also differed based on various sentiments of abortion measured but not on opinion regarding legality of abortion for adolescents (p = .49).

Conclusions:

Perception of state’s abortion rights and concern about legislative restrictions influenced NASPAG clinicians’ willingness to provide medication abortion for adolescents. Interventions to minimize legislative interference with medical care, increase abortion training, and implement medication abortion in pediatric settings may expand AYA medication abortion access.

Introduction

Abortion is critical healthcare for adolescents and young adults (AYAs).¹ The Supreme Court decision in Dobbs v. Jackson Women’s Health Organization significantly changed the landscape of abortion access in the United States (US) by overturning 1973’s Roe v. Wade.^{2, 3} By the end of 2023, 21 states enacted new abortion restrictions previously prohibited by this standard. While adolescents (19 years and younger) comprise a small proportion of people having abortions in the US (8.4% of abortions in 2021),⁴ this population experiences unique barriers to abortion (e.g., state parental involvement laws, financial constraints, travel challenges, and confidentiality issues).^1,5–8 As such, those who provide sexual and reproductive healthcare must consider the disproportionate impact of abortion restrictions on adolescents and the importance of expanding access.^6–9

Few studies have surveyed US clinicians primarily caring for AYAs to understand their sentiments and practices of abortion, with the most recent published over a decade ago.^10,11 Prior to Dobbs, studies found that few physicians and nurse practitioners belonging to the Society for Adolescent Health and Medicine (SAHM) provided either medication or procedural abortion (1.7% in 1998 and 3% in 2011).^10,11 Medication abortion, which involves use of misoprostol, a prostaglandin, with or without mifepristone, a progesterone receptor antagonist, during the first 11 weeks of pregnancy, is within the skillset of a wide range of clinicians.¹² Notably, almost half of clinicians surveyed in 1998 indicated an interest in providing medication abortion in anticipation of the Food and Drug Administration’s ultimate approval of mifepristone in 2000.¹⁰ However, a 2011 survey of a similar sample of clinicians did not reveal a substantial increase in those providing abortion. While the 2011 study demonstrated SAHM physicians and nurse practitioners had gaps in knowledge about medication abortion, the authors did not examine additional factors preventing these clinicians from offering this service in their practice.¹¹ Furthermore, provision of medication abortion in the US has increased substantially since these studies (24% of abortions in 2011 versus 63% of abortions in 2023,¹³ which may have influenced clinicians’ present day knowledge of and willingness to provide medication abortion. Additionally, over the past decade, the number of obstetricians and gynecologists (OBGYNs) with subspecialty training in pediatric and adolescent gynecology has grown substantially.¹⁴ These clinicians routinely provide gynecologic care to adolescents who may become pregnant; yet, there has not been a full assessment of pediatric and adolescent gynecologists’ sentiments and practices of abortion care.

Better understanding of clinicians’ attitudes and willingness to provide medication abortion to AYAs and potential implementation barriers is crucial to inform the broadening of the medication abortion provider network, which may increase AYA access to abortion in the wake of the Dobbs decision.⁹ The North American Society for Pediatric and Adolescent Gynecology (NASPAG) is an organization with the mission “to provide multidisciplinary leadership in education, research, and gynecologic care to improve the reproductive health of youth.”¹⁵ This post-Dobbs cross-sectional survey aims to address gaps in the literature by assessing the prevalence of NASPAG clinicians caring for AYAs willing to provide medication abortion in their current practice settings and examining their sentiments regarding abortion. Secondary aims include identifying potential or real barriers to implementing medication abortion into practice, examining factors affecting NASPAG clinicians’ willingness to provide medication abortion, and assessing differences in abortion practices and sentiments among NASPAG clinicians based on clinician specialty and practice setting.

Methods

Survey

We designed a 43-item questionnaire to examine sentiments regarding abortion and willingness to provide medication abortion among NASPAG clinicians caring for AYAs capable of pregnancy (see Online Supplementary Materials for the questionnaire). We derived questions from prior abortion research and the research team’s experience caring for AYAs and providing abortion. The survey examines barriers to medication abortion,¹⁶ abortion referral practices,¹⁷ availability of mifepristone (despite the risk evaluation and mitigation strategy (REMS) program),¹⁸ and personal, peers’, and states’ sentiments about abortion.^17–20 To measure potential or real barriers to providing medication abortion, we asked participants to select all that applied from a list of 21 options. We also surveyed participant demographics (age, gender identity, type of clinician, religious beliefs) with guidance from a resource aimed at improving diversity, equity, and inclusion in data collection,²¹ abortion training experiences, and practice setting. We did not randomize or alternate survey items. No survey items required a response beyond the initial screening questions to assess eligibility. Participants completed surveys anonymously to limit social desirability bias. We collected and stored de-identified data in Research Electronic Data Capture (REDCap), a secure, HIPAA-compliant platform of the Mass General Brigham Research Computing, Enterprise Research Infrastructure & Services (ERIS) group. An institutional firewall provided data security. The research team consulted the guidelines of Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) during study design and manuscript preparation.²²

We, a multidisciplinary team of pediatricians and OBGYNs at urban academic centers, recruited potential participants from the NASPAG listserv via e-mail. We utilized a convenience sample and estimated a sample size of 85 participants based on recent studies which used the NASPAG listserv as a sampling frame.^14,23 Eligible participants included any medical provider currently caring for AYAs who could become pregnant. We invited potential participants via listserv e-mail. The research team sent a second, identical invitation via the NASPAG listserv four weeks after the initial invitation. Interested participants accessed the survey via a hyperlink provided in the e-mail. The first page of the survey included a consent fact sheet. We assumed consent by survey continuation. Participants completed the survey online using a personal device. Following completion of the survey, participants received resources about abortion care (see Online Supplementary Materials). Survey completion was voluntary, and responses were anonymous. We invited all participants to enter an optional drawing for a gift card via a separate survey link to preserve anonymity following survey completion. We used this gift card drawing as a token of appreciation for participation, similar to a prior survey study utilizing the NASPAG listserv.²³ The institutional review boards at Boston Children’s Hospital (IRB-P00043717) and Mass General Brigham (2022P002823) deemed this study exempt.

Statistical analysis

We used numbers and percentages to analyze data. Except for the inclusion criteria, all questions were optional resulting in some variables with missing data. For variables with missing data, we calculated percentages using the number of respondents for that variable. We used χ² and Fisher’s Exact Test, where appropriate, to assess differences in responses between groups. We performed all analyses in Stata 16.0 (StataCorp, LLC).

Results

Eighty-one individuals started the study survey. We excluded two potential participants for failing to meet our eligibility criteria and nine participants for incomplete survey responses, with almost all the questions unanswered. Our study included 70 NASPAG clinicians currently caring for AYAs who could become pregnant in their clinical practice. As seen in Table 1, the majority of our sample identified as a cisgender woman, female, or feminine (94%) and were physicians (99%). Few participants (14%) identified as clinicians in training. Of physicians, 26% had training in pediatrics and/or adolescent medicine, and 71% had training in OBGYN. Twenty percent of our sample identified as somewhat, quite a bit, or a great deal religious. As seen in Table 2, participants practiced in various settings, with half (51%) primarily at pediatric hospitals. Most participants practiced in urban settings (77%), and few practiced in facilities with religious affiliations (4% Catholic affiliation and 6% other religious affiliation). Participants reported practicing throughout the US, with even representation of participants who perceived their state’s abortion rights to be “supportive or very supportive” (42%) and “hostile or very hostile” (42%).

Table 1:

Demographics of participants (n = 70)^a, comparing those willing to provide medication abortion to an adolescent in their current practice setting and those who were not.

	Total sample (n = 70) n (%)	Willing to provide MAB (n = 36) n (%)	Not willing to provide MAB (n = 34) n (%)	p Value
Age (years)				.34
26–35	13 (19)	8 (22)	5 (15)
36–45	23 (33)	12 (33)	11 (33)
46–55	15 (22)	9 (25)	6 (18)
56–65	14 (20)	4 (11)	10 (30)
66 and over	4 (4)	3 (8)	1 (3)
Gender identityb				1.00
Cisgender woman, female, feminine	65 (94)	34 (94)	31 (94)
Cisgender man, male, masculine	4 (6)	2 (6)	2 (6)
Type of clinician				.48
Physician	68 (99)	36 (100)	32 (97)
Advanced practice nurse	1 (1)	0 (0)	1 (3)
Clinician in training				.31
Yes	10 (14)	7 (20)	3 (9)
No	59 (85)	29 (80)	30 (91)
Type of physicianc				.51
Pediatrics - adolescent medicine	15 (22)	8 (22)	7 (22)
Pediatrics - other	2 (3)	1 (3)	1 (3)
OBGYN - generalist	6 (8)	4 (11)	2 (6)
OBGYN - pediatric and adolescent	40 (59)	19 (53)	21 (66)
OBGYN - other	3 (4)	3 (8)	0 (0)
Internal Medicine - adolescent medicine	1 (1)	0 (0)	1 (3)
Other	1 (1)	1 (3)	0 (0)
Religious beliefs				.98
No religious identifica ion	31 (45)	15 (42)	16 (49)
A little or not at al rer nous	24 (34)	13 (36)	11 (33)
Somewhat religious	7 (10)	4 (11)	3 (9)
Quite a bit or a great deal religious	7 (10)	4 (11)	3 (9)

MAB, medication abortion; OBGYN, obstetrics and gynecology.

^aOne participant did not complete the demographic data. For variables with missing data, percentages were calculated using the number of respondents.

^bThe survey included 7 potential responses for gender identity (cisgender woman, female, feminine; transgender woman, female, feminine; cisgender man, male, masculine; transgender man, male, masculine; non-binary, gender queer, gender expansive; prefer to self-describe; prefer not to answer). Options with no responses were deleted from the table for space.

^cType of physician only includes physician participants (n = 68). Options with no responses were deleted from the table for space.

Table 2:

Characteristics of participants’ main practice settings (n = 70)^a, comparing those willing to provide medication abortion to an adolescent in their current practice setting and those who were not.

	Total sample (n = 70) n (%)	Willing to provide MAB (n = 36) n (%)	Not willing to provide MAB (n = 34) n (%)	p Value
Type of practice				.001
Private practice	8 (12)	3 (8)	5 (15)
Community-based health center	4 (6)	4 (11)	0 (0)
Pediatric hospital	35 (51)	11 (30)	24 (73)
General hospital, pediatric department	10 (15)	8 (22)	2 (6)
Other	12 (17)	10 (28)	2 (6)
Region of United States				.007
West	11 (17)	9 (27)	2 (6)
Midwest	21 (33)	5 (15)	16 (52)
Southwest	6 (9)	3 (9)	3 (10)
Southeast	11 (17)	5 (15)	5 (19)
Northeast	15 (23)	11 (33)	4 (12)
Population density				.33
Urban	53 (77)	27 (75)	26 (79)
Midsize town/suburb	13 (19)	6 (17)	7 (21)
Rural	3 (4)	3 (8)	0 (0)
Religious affiliation				.20
Nonreligious facility	61 (88)	32 (91)	29 (88)
Catholic facility	3 (4)	0 (0)	3 (9)
Other religious facility	4 (6)	3 (9)	1 (3)
Distance from nearest abortion clinic (miles)				.02
Less than 5	38 (55)	25 (69)	13 (39)
5–25	13 (19)	6 (17)	7 (21)
26–50	5 (7)	3 (8)	2 (6)
>50	11 (16)	2 (6)	9 (27)
Unsure	2 (3)	0 (0)	2 (6)
Mifepristone available at current setting				.006
Yes	20 (29)	16 (44)	4 (12)
No	44 (63)	17 (47)	27 (79)
Unsure	6 (9)	3 (8)	3 (9)
Perception of state’s abortion rights				.001
Supportive or very supportive	27 (42)	20 (61)	7 (23)
Leans supportive	5 (8)	4 (12)	1 (3)
Middle ground	1 (1)	0 (0)	1 (3)
Leans hostile	4 (6)	2 (6)	2 (6)
Hostile or very hostile	27 (42)	7 (21)	20 (64)

MAB, medication abortion.

^aFor variables with missing data, percentages were calculated using the number of respondents.

Abortion practices and training

As demonstrated in Table 3, almost all the participants had ever provided pregnancy options or abortion counseling to an adolescent (98%), and 61% ever provided abortion (5% medication abortion only; 16% procedural abortion only; 79% both). More than half of participants practiced less than five miles from the nearest abortion clinic (55%) and did not have mifepristone available in their current practice setting (63%) (Table 2). Three-quarters of participants reported referring patients for abortion outside of their current practice setting (n = 50), such as to an independent abortion clinic (46%), gynecologist (23%), or another specialist (3%). Most participants knew where to seek abortion training (87%); however, of those who had no prior training (n = 12), only two sought out training. Of those who sought out any or additional training (n = 18), only three participants did so since May/June 2022, corresponding to the leak of and the final Supreme Court decision in Dobbs v. Jackson Women’s Health Organization.

Table 3:

Participants’ clinical practice and training in abortion (n = 70)^a, comparing those willing to provide medication abortion to an adolescent in their current practice setting and those who were not.

	Total sample (n = 70) n (%)	Willing to provide MAB (n = 36) n (%)	Not willing to provide MAB (n = 34) n (%)	p Value
Ever provided pregnancy options/abortion counseling to an adolescent (yes)	69 (98)	36 (100)	33 (97)	.49
Ever provided an abortion				.04
Yes	43 (61)	27 (75)	16 (47)
No - no opportunity	22 (31)	8 (22)	14 (41)
No - chose not to	5 (7)	1 (3)	4 (12)
Knows where to seek abortion training (yes)	61 (87)	30 (83)	31 (91)	.48
Type of abortion training received
Medication	51 (72)	31 (86)	20 (59)	.02
Procedural	46 (66)	27 (75)	19 (56)	.08
Pre-abortion or post-abortion care	52 (75)	30 (83)	22 (65)	.10
None	12 (17)	4 (11)	8 (24)	.21
Has sought out additional training (yes) (if had any, n = 58)	16 (28)	12 (38)	4 (15)	.08
Has sought out any training (yes) (if had none, n = 12)	2 (17)	2 (50)	0 (0)	.09
When did you seek training? (n = 18)
Since May/June 2022	3 (4)	3 (8)	0 (0)	.24
Before May/June 2022, within the last 2 years	4 (6)	4 (11)	0 (0)	.12
Greater than 2 years ago	12 (17)	8 (22)	4 (12)	.34

MAB, medication abortion.

^aFor variables with missing data, percentages were calculated using the number of respondents.

Willingness to provide medication abortion

Thirty-six (51%) participants indicated they were willing to provide a medication abortion for an adolescent who requested it in their clinical practice, while 34 (49%) indicated they were not willing to do so (Table 1). Notably, participants’ willingness to provide a medication abortion differed by type of practice, region of the US, distance from nearest abortion clinic, availability of mifepristone, and perception of state’s abortion rights (Table 2). Participants willing to provide medication abortion were more likely to have experience providing abortion (p = .04) and to have received medication abortion training (p = .02) (Table 3). The most common potential or real barriers to providing a medication abortion were legislative restrictions (n = 33; 47%) and dispensing pills from clinics (n = 23; 33%) (Figure 1). Participants willing to provide medication abortion cited legislative restrictions as a barrier less frequently than those not willing to provide (30% vs. 65%; p = .008). These groups had a similar proportion of clinicians who considered dispensing pills from clinic a barrier to providing medication abortion (39% willing vs. 26% not willing; p = .32).

Figure 1.

Potential or real barriers to participants’ ability to provide medication abortion in their current practice setting (n = 70).

Footnote: Participants were asked to select all that applied from a list of 21 barriers.

When comparing participants with OBGYN training (n = 50) to those with other training (n = 18), there was no difference in ever providing pregnancy options or abortion counseling to an adolescent (p = 1). Significantly more OBGYN-trained participants ever provided abortion compared to participants trained in other specialties (84% vs. 5%; p = .001). Significantly more OBGYN-trained participants received training in medication abortion (88% vs. 33%; p = .001), procedural abortion (86% vs. 17%; p = .001), and pre- and post-abortion care (90% vs. 39%; p = .001) than those trained in other specialties. Notably, one percent of participants with training in a specialty other than OBGYN reported providing medication abortion despite 33% reporting having received medication abortion training. There were no differences among participants with OBGYN training and those with other training with regards to perception of how relevant abortion is to their clinical practice (OBGYN-trained: 8% not at all, 54% a little or somewhat, 36% very or extremely vs. other-trained: 5% not at all, 50% a little or somewhat, 39% very or extremely; p = 1) or willingness to provide a medication abortion (54% OBGYN-trained vs. 50% other-trained; p = .79).

Additionally, when comparing participants practicing in pediatric hospitals (n = 35) versus other practice settings (n = 34), we found that participants practicing in pediatric hospitals were more likely to refer patients for abortion outside their main practice setting (86% vs. 59%; p = .02) and were less likely to have mifepristone available at their main practice setting (6% vs. 53%; p = .001).

Sentiments regarding abortion

Few participants indicated the provision of abortion care was not at all relevant to their clinical practice (9%) or that they were morally opposed to abortion (6%) (Table 4). Eighty-eight percent of participants supported unrestricted access to abortion a lot, while only 60% thought their workplace peers shared this same level of support. Almost all participants responded that adolescents should be able to have an abortion legally (99%), laws should be changed to make it easier for an adolescent to have an abortion (94%), people should be able to use their health insurance to help cover the cost of receiving an abortion (97%), and public funding should cover abortion services (97%). The majority of participants disagreed that adolescents requesting abortion should experience a mandatory waiting period before receiving an abortion (89%) and that minor adolescents should be required to obtain parent/guardian consent before obtaining an abortion (94%).

Table 4:

Participants’ sentiments of abortion care (n = 70)^a, comparing those willing to provide medication abortion to an adolescent in their current practice setting and those who were not.

Question/Statement	Total sample (n = 70) n (%)	Willing to provide MAB (n = 36) n (%)	Not willing to provide MAB (n = 34) n (%)	p Value
How relevant do you think your provision of abortion care is (or would be) to your clinical practice?				.03
Not at all	6 (9)	1 (3)	5 (16)
A little or somewhat	37 (54)	17 (47)	20 (62)
Very or extremely	25 (37)	18 (50)	7 (22)
How do YOU feel about unrestricted access to abortion?				.01
Oppose a lot	1 (1)	0 (0)	1 (3)
Oppose somewhat	1 (1)	1 (3)	0 (0)
Oppose a little	1 (1)	0 (0)	1 (3)
Neither oppose nor support	0 (0)	0 (0)	0 (0)
Support a little	0 (0)	0 (0)	0 (0)
Support somewhat	5 (7)	0 (0)	5 (15)
Support a lot	61 (88)	34 (97)	27 (79)
How do YOUR PEERS at your workplace generally feel about unrestricted access to abortion?				.38
Oppose a lot	2 (3)	0 (0)	2 (6)
Oppose somewhat	2 (3)	1 (3)	1 (3)
Oppose a little	2 (3)	2(6)	0 (0)
Neither oppose nor support	2 (3)	1 (3)	1 (3)
Support a little	8 (12)	4 (11)	4 (13)
Support somewhat	11 (16)	5 (14)	6 (19)
Support a lot	40 (60)	23 (64)	17 (55)
Leaving aside what you think about abortion for yourself or for those close to you, do you think an adolescent should be able to have an abortion legally?				.49
Yes	69 (99)	36 (100)	33 (97)
No	1 (1)	0 (0)	1 (3)
Unsure	0 (0)	0 (0)	0 (0)
Leaving aside whether you think abortion should be legal, are you morally opposed to abortion?				.02
Yes	4 (6)	0 (0)	4 (12)
No	63 (90)	33 (92)	30 (88)
Unsure	3 (4)	3 (8)	0 (0)
Do you think that laws should be changed to make it easier for an adolescent to have an abortion?				.05
Yes	66 (94)	36 (100)	30 (88)
No	1 (1)	0 (0)	1 (3)
Unsure	3 (4)	0 (0)	3 (9)
Do you think people should be able to use their health insurance to help cover the cost of receiving an abortion?				.23
Yes	68 (97)	36 (100)	32 (94)
No	1 (1)	0 (0)	1 (3)
Unsure	1 (1)	0 (0)	1 (3)
Do you think public funding [Medicaid] should cover abortion services?				.23
Yes	68 (97)	36 (100)	32 (94)
No	2 (3)	0 (0)	2 (6)
Unsure	0 (0)	0 (0)	0 (0)
When an adolescent requests an abortion, there should be a mandatory waiting period before the abortion is performed.				.001
Strongly disagree	39 (56)	27 (75)	12 (35)
Disagree	23 (33)	4 (11)	19 (56)
Neither agree/disagree	3 (4)	3 (8)	0 (0)
Agree	3 (4)	2 (6)	1 (3)
Strongly agree	2 (3)	0 (0)	2 (6)
Minors under the age of 18 should be required to obtain consent from their parent or guardian before obtaining an abortion.				.001
Strongly disagree	40 (57)	28 (78)	12 (35)
Disagree	26 (37)	8 (22)	18 (53)
Neither agree/disagree	0 (0)	0 (0)	0 (0)
Agree	4 (6)	0 (0)	4 (12)
Strongly agree	0 (0)	0 (0)	0 (0)

MAB, medication abortion.

^aFor variables with missing data, percentages were calculated using the number of respondents.

Participant willingness to provide medication abortion differed with regards to perceived relevance of abortion care to their clinical practice (p = .03), personal feelings about unrestricted access to abortion (p = .01), and being morally opposed to abortion (p = .02) (Table 4). Participants willing and not willing to provide medication abortion differed regarding support for policies regulating adolescent abortion access, such as requiring adolescents requesting an abortion to experience mandatory waiting periods before an abortion (p = .001) and requiring a minor adolescent to obtain parent/guardian consent before an abortion (p = .001) (Table 4). The two groups’ responses were similar when asked whether an adolescent should be able to have an abortion legally (p = .49), how their peers feel about unrestricted access to abortion (p = .38), whether laws should make it easier for an adolescent to access an abortion (p = .05), whether people should be able to use their health insurance to help cover the cost of receiving an abortion (p = .23), or whether public funding [Medicaid] should cover abortion services (p = .23) (Table 4).

Discussion

Calls to expand the medication abortion provider network have existed over the last decade²⁴ and have become amplified in the wake of the Dobbs decision, with particular emphasis on the inclusion of clinicians caring for AYAs.^6,9,25 The increased use of medication abortion in the US¹³ and reduced abortion access following Dobbs may impact clinicians’ perception of and willingness to provide medication abortion. Our study is the first, to our knowledge, to survey a specialty organization of clinicians caring for AYAs about their practices of and sentiments toward abortion following the Dobbs decision.

Guidelines from the American College of Obstetricians and Gynecologists (ACOG) state clinicians who can screen for medication abortion eligibility, provide appropriate follow-up care, and perform uterine evacuation or refer to a clinician with these skills, are qualified to provide medication abortion.¹² As such, clinicians who care for AYAs, such as NASPAG clinicians, are suitable to provide this highly efficacious, safe, and low resource abortion option.¹² Roughly half of our participants were willing to provide medication abortion to an adolescent in their current practice setting, which is similar to the proportion of SAHM clinicians expressing interest in providing medication abortion more than 25 years ago.¹⁰ Very few NASPAG clinicians in our sample opposed unrestricted access to abortion (n = 3), were morally opposed to abortion (n = 4), or thought adolescents should not be able to have an abortion legally (n = 1). Compared to a 2001 study, our study demonstrates a low proportion of clinicians personally opposed to abortion (3% in our sample of NASPAG clinicians versus 40% of gynecologists and 37% of general practitioners in a 2001 national sample of women’s healthcare providers).²⁶ While interventions aimed at addressing clinician attitudes towards and perceptions of abortion may be helpful at increasing provision of medication abortion among some groups of clinicians, such interventions would likely be less useful among NASPAG clinicians who are highly supportive of abortion.

Barriers related to governmental policies significantly influenced NASPAG clinicians’ willingness to provide medication abortion, as participants willing and not willing to provide medication abortion differed based on perception of state’s abortion rights and concerns about legislative restrictions. When compared to data pre-Dobbs, more clinicians in our sample cited legislative restrictions as a barrier to medication abortion provision (47% in our sample of NASPAG clinicians versus 36% of general physicians and 39% of gynecologists in 2001).²⁶ These results emphasize the need for state-level policy interventions to limit legislative interference in provision of this critical healthcare.

Our data also suggest that various systems-level factors influence NASPAG clinicians’ willingness to provide medication abortion to adolescents in their current practice, including prior training in medication abortion, prior experience providing abortion care, and availability of mifepristone in their practice setting. These results are similar to pre-Dobbs studies, which indicate a connection between abortion knowledge and training of non-OBGYN clinicians^16,27 and challenges dispensing mifepristone^18,27,28 and willingness to provide medication abortion. These systems-level barriers were most prevalent among clinicians who practice in pediatric hospital settings or trained in specialties other than OBGYN. As such, interventions aimed at the individual level, such as increasing abortion training for non-OBGYN clinicians, particularly those trained in pedatrics,²⁵ and at the institutional level, such as assisting pediatric hospital systems with implementing medication abortion provision in this practice setting (e.g., overcoming barriers to stocking mifepristone) may play critical roles in expanding the AYA abortion provider network.

Focusing on expanding access to medication abortion in pediatric hospital systems may be particularly important, as NASPAG clinicians practicing in this setting in our sample were more likely to refer AYAs outside of their main practice setting for an abortion compared to those practicing in other settings. It is within the scope of practice for pediatricians, adolescent medicine physicians, pediatric and adolescent gynecologists, and other clinicians practicing in pediatric hospital settings to provide medication abortion.¹² A lack of clinicians capable of performing procedural abortion, in the rare circumstance it is required following a medication abortion, should not be a barrier to providing medication abortion in pediatric settings. A collaboration within the healthcare system (e.g., clinicians capable of performing procedural abortion at an affiliated or nearby adult hospital) may provide a pathway for emergent access. Prior to Dobbs, a national survey of adolescents (younger than 20) accessing abortion demonstrated that many (69.5%) of adolescents wanted their abortion sooner and that adolescents were more likely to report not knowing where to obtain an abortion when compared to older patients.²⁹ While no studies exist, to our knowledge, comparing outcomes of AYAs who must receive referrals to access any abortion care compared to those who do not, it is likely that these external referrals cause delays in accessing this critical care. AYAs have unique barriers to attending referral appointments, especially when needed to travel far distances, such as lack of financial independence, transportation barriers, and confidentiality concerns. Furthermore, referral practices for abortion likely exacerbate abortion stigma and disrupt continuity of care for AYAs. As such, interventions at the institutional level to educate pediatric hospital leadership on the importance of providing medication abortion in these settings and assist pediatric hospital systems in creating the necessary infrastructure for providing medication abortion may help increase access and decrease barriers to abortion for AYAs. Most importantly, more research is urgently needed focusing on the experience of adolescents accessing abortion in the evolving policy landscape post-Dobbs to provide evidence to interventions aimed at expanding AYA access to abortion.

Our study has several limitations worth noting. Our survey includes a relatively small number of NASPAG clinicians who care for AYAs and notably includes only one non-physician clinician, few pediatricians without adolescent subspecialty, and no family medicine clinicians; therefore, results may not be generalizable to all providers who care for AYAs. Our study includes participants throughout all regions of the US and has equal representation of participants practicing in states perceived to be supportive and hostile to abortion rights, although notably has few clinicians practicing in rural settings. We do not know the number of subscribers to the NASPAG listserv, thus are not able to estimate our response rate. Our study included a similar number of participants as prior research using this listserv.^14,23 Nonresponse or selection bias is possible, which could result in our data over or underestimating the true prevalence of observed outcomes, particularly with regard to sentiments of abortion, as clinicians highly supportive of abortion may have been more likely to complete the survey; however, clinicians with strong sentiment against abortion may have been equally as motivated to complete the survey. Additionally, social desirability bias may have influenced answers related to sentiments of abortion, however, survey anonymity may have mitigated this effect. Lastly, given the survey design, we are limited to reporting associations between variables and willingness to provide medication abortion. As such, we are not able to prove causality or account for confounding variables.

Despite these limitations, our survey of NASPAG clinicians provides important preliminary information about the sentiments and practices of abortion among clinicians caring for AYAs following the Supreme Court decision in Dobbs v. Jackson Women’s Health Organization. Our participants were highly supportive of adolescent abortion provision, and the presence of and concern about legislative restrictions following the Dobbs decision affected clinician willingness to provide medication abortion for adolescents in their current practice. These results underscore the importance of policy interventions aimed at reducing state legislative interference in providing this critical healthcare. Most importantly, our findings provide insight about interventions to expand medication abortion access for AYAs, such as increasing abortion training for clinicians trained in specialties other than OBGYN and designing interventions to increase medication abortion provision in pediatric hospital settings. A robust repository of educational and informational materials exists online to aid providers in expanding their clinical practice to include medication abortion (see Online Supplementary Materials). Future research directions include investigating the impact of abortion restrictions on AYAs and the clinicians that serve them across multiple states and settings, as well as the feasibility and impact of interventions aimed at increasing AYA access to abortion in states where this care remains legal.