Personal experience with abortion influences how we practice as obstetrician–gynecologists

Abstract

Purpose

To evaluate whether a personal history of having an abortion, or knowing someone close who did, has an impact on how a physician practices.

Methods

A web-based survey was sent to Ob/Gyn residents asking about personal abortion experience, demographics, religious background, political views, residency program metrics, and intent to provide abortion. Pearson Chi-square and Mann–Whitney U tests were performed to assess whether or not personal history of abortion (PHA) was significantly associated with personal characteristics or training patterns.

Results

Ob/Gyn residents with a personal history of abortion were older in age (p = 0.002), sought out additional training in family planning (p = 0.009), considered it extremely important that their program offers family planning training (p < 0.001), and were much more likely to intend to become an abortion provider themselves post-residency (p < 0.001).

Conclusion

Personal experience is associated with training and practice patterns among Ob/Gyn residents, a finding that has implications for an Ob/Gyn workforce under pressure from increasing legislative restrictions to abortion care. Ob/Gyn residents should be supported in their training goals to combat moral injury associated with the inability to provide evidence-based care.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s00404-026-08369-4.

What does this study add to the clinical work

Personal experience with abortion—either personally or through someone close—significantly influences an Ob/Gyn resident's desire for family planning training and their commitment to becoming a future abortion provider. These personal histories are a primary driver of practical goals, which is critical for workforce planning as legislative restrictions on abortion care and training continue to increase.

Introduction

Family planning is one of the most common reasons reproductive-aged women visit their obstetrician–gynecologist, and abortion is one of the most common medical and surgical procedures performed in the United States [1]. Because of the need for safe and accessible abortion and contraception care, the American College of Obstetricians and Gynecologists has supported abortion education for residents to increase the availability of trained providers [2]. In 1996, the Accreditation Council for Graduate Medical Education (ACGME) required residency training programs to integrate abortion training within the curriculum on an “opt-out” or “opt-in” basis [3]. After the US Supreme Court’s Dobbs v. Jackson Women’s Health Organization overturned Roe v. Wade, the ACGME revised Obstetrics and Gynecology (Ob/Gyn) program requirements in September of 2022 to necessitate that all programs offer abortion training on an “opt-out” basis [4].

Despite this accreditation requirement, access to abortion training remains variable, as nearly half of Ob/Gyn residency programs are in restrictive jurisdictions for abortion care [5]. Recent studies have found that Ob/Gyns in states with abortion bans are already experiencing moral injury, as these restrictions hinder their ability to provide evidence-based abortion care [6, 7].

Each individual resident carries unique personal experiences that have the potential to inform their approach to care. In this paper, we seek to determine if a personal history of having had an abortion or knowing someone close who did have one (PHA) influences how that individual physician practices.

Materials and methods

Web-based survey

Under IRB approval (IRB#074617MP4X), a link (https://waynestate.az1.qualtrics.com/jfe/form/SV_7VssqjccfquwnVb) to an anonymous, voluntary Qualtrics (Qualtrics International, Provo, UT, USA) open survey (Fig. S1) was sent to each Obstetrics and Gynecology residency training program director in the United States with a request to forward the survey to their residents. This dataset was used in a recently published study which examined a resident’s intent to provide abortions post-residency [8]. Informed consent was obtained electronically prior to the start of the survey. The two-page survey was developed by the authors and the technical functionality of the survey was tested prior to distribution. The survey asked if the resident had personally had an abortion or knew someone close to them that did, as a single question (independent variable). Collected data included the following: demographics, religious background, political views, residency program metrics, and intent to provide abortion (Table 1). Survey items were not randomized or alternated, with no adaptive questions. There were 15 questions per page displayed in the survey, which was in total 2 pages long. There were no questions marked as required to check for survey completeness upon submission. Respondents were able to review or revise their responses by navigating backwards prior to submission. Unique survey responses were determined through IP addresses (69% of surveys came from unique addresses) and enforced through personalized cookies, preventing duplicates among the respondents who used a shared work computer. In addition, prevention of duplicates was performed using Qualtrics internal system rather than simply discriminating by IP address, and log file analysis was not applicable. Participants were given the option of submitting an email if they desired to partake in a random drawing for a Starbucks gift card. States were grouped into regions as defined by the U.S. Census Bureau [9], political affiliation was grouped as republican or democrat based on choice of Presidential candidate, religion was analyzed as categorical data, and case numbers (contraceptive visits, tubal sterilizations, long-acting reversible contraceptive insertions, and abortions) were analyzed as continuous data. Data were collected over a 3-month period. All data, including demographic information, were protected from unauthorized access by password protection and were encrypted and stored on a secure server only accessible to the principal investigator.

Table 1.

Cohort of residents responding to the electronic survey

		Count	Column N (%)
Age	25–30	230	56
	31–35	165	40
	36–45	12	3
	46 +	1	0
Gender	Female	362	89
	Male	45	11
	Trans/intersex/nonbinary	1	0
Sexual orientation	Heterosexual	370	92
	Bisexual	19	5
	Homosexual	15	4
Marital status	Married	210	52
	Single	90	22
	Long-term relationship	46	11
	Engaged	29	7
	Dating	30	7
	Decline to answer	2	0
Region of origin	Midwest	112	28
	West	101	25
	South	92	23
	Northeast	82	20
	Outside US	16	4
Size of origin	Major metro area of 1mill +	121	30
	Large city of 0.5–1mill	49	12
	City of 100,000–0.5mill	82	20
	Small urban area of 50–100,000	64	16
	Rural (less than 50,000)	91	22
Region of residency	Midwest	125	31
	Northeast	112	28
	West	105	26
	South	65	16
Year of training	PGY1	101	25
	PGY2	100	25
	PGY3	110	27
	PGY4	97	24
Current religion	Atheist/agnostic	90	22
	No religion identified	77	19
	Catholic	76	19
	Nondenominational	52	13
	Protestant	35	9
	Jewish	28	7
	Other	19	5
	Hindu	13	3
	Baptist	8	2
	Muslim	4	1
	Evangelical	3	1
	Buddhist	3	1
Family’s degree of religiosity	Yes, both parents attend services regularly	186	46
	Parents only attend during major holidays	110	27
	Parents are not practicing religion but do believe	53	13
	Parents non-religious/not at all	57	14
Current religious practice	Not at all	202	50
	Only on major holidays	140	34
	Actively (attend daily/weekly)	65	16
Political party affiliation	Democratic	358	88
	Republican	34	8
	No answer	17	4
Had an abortion (or someone close)	Yes	190	50
	No	193	50

Statistical analysis

Deidentified survey data were pasted into an Excel spreadsheet and this convenience sample was analyzed using Statistical Package for the Social Sciences (SPSS) (IBM Analytics). Partially completed surveys were included in the analysis. Descriptive statistics were used to characterize the sample. The Pearson Chi-square test was performed to test the association between PHA (a categorical variable) and other categorical variables. The nonparametric Mann–Whitney U test was performed to test the association between PHA and ordinal variables. A raw p value < 0.05 was considered significant. To control the Family-Wise Error Rate (FWER, false positives) across multiple independent comparisons, the Holm correction was applied.

Results

At the time of this survey, there were 4991 Ob/Gyn residents training in 295 ACGME-accredited programs in the United States [10]. The convenience sample of survey respondents includes 429 respondents (8% of the U.S. total) who started the survey, with 409 completing the majority of questions (a 95% completion rate). Eighty-nine percent (89%) of respondents identified their gender as “Female”. There was a near equal distribution of responses by geographic region and year in training (Table 1). A total of 383 survey respondents (93.6% response rate) answered the single question regarding their personal history of an abortion experience (either their own or someone they know) with a nearly even split between “yes” and “no” responses.

Those that either had a personal history of abortion or knew someone who did tended to be older in age (p = 0.002), lean democrat (p = 0.005), train at opt-out programs (p = 0.040), seek out additional training in family planning (p = 0.009), and were more likely to intend to provide abortion care post-residency (p < 0.001). When considering residency applications, those with a personal abortion experience considered it “extremely” or “very” important that their residencies have family planning training (p < 0.001). Gender, sexual orientation, marital status, geographic region of residency training, year of training, religion, degree of religiosity (as defined by regular attendance), hospital religious affiliation, presence of a Ryan program, or family planning case volumes were not significantly different between these two groups (Table 2).

Table 2.

Association between abortion history and demographics/beliefs/practice patterns

Variable (p value of association test)		Personal history of abortion or know someone close who did (PHA)
		No n = 190		Yes n = 193
		Count	Row (%)	Count	Row (%)	Significant after holm correction?
Gender (p = 0.056)	Male	25	64	14	36	No
	Female	164	48	178	52
Age (p = 0..002)	25–30	122	56	94	44
	31–35	65	42	90	58	Yes
	36–45	3	27	8	73
Current religious practice (p = 0.019)	Not at all	88	45	106	55
	Only on major holidays	69	53	61	47	No
	Actively (attend daily/weekly)	33	58	24	42
Political party affiliation (p = 0.005)	Democratic	159	47	179	53
	No Answer	9	60	6	40	Yes
	Republican	22	73	8	27
Importance of Family planning in choosing a residency program (p < 0.001)	Chose because it did not	1	33	2	67
	Not at all important	22	67	11	33
	Slightly or moderately	65	63	38	37	Yes
	Very or extremely	102	42	141	58
Presence of a Ryan program (p = 0.621)	No	50	54	43	46
	Unsure	15	50	15	50	No
	Yes	124	48	135	52
Opt-in or opt-out program (p = 0.040)	Opt-in	26	65	14	35	Yes
	Opt-out	164	48	179	52
Sought out additional family planning training outside residency (p = 0.009)	No	178	52	165	48
	Yes	12	30	28	70	Yes
OBGYN career choice factor—social aspects of care (p = 0.036)	No	107	55	88	45
	Yes	83	44	105	56	No
Intent to provide abortions (p < 0.001)	No	72	59	50	41
	Yes, but only medical or 1st tri	58	58	42	42	Yes
	Yes, 2nd trimester	60	38	100	63

Discussion

Despite recent restrictions to abortion access, abortion remains one of the most common medical and surgical procedures in Ob/Gyn care [11]. According to the Guttmacher Institute’s Monthly Abortion Provision Study, over 1 million abortions were performed by a clinician in states without bans in 2023 [12]. Notably, there has been an 11% increase in clinician-provided abortions in these states since 2020. While all residents are required to obtain similar levels of knowledge on contraception and abortion, approximately 40% of Ob/Gyn residents are in states with restrictive abortion legislation [13]. Variable opportunities to receive adequate abortion training add complexity to the implications of this study’s findings.

Each resident brings a personal experience which impacts practice patterns [14]. In this survey of Ob/Gyn residents, we found that approximately half of respondents had a personal experience with abortion care, as patients themselves or as someone who knows a patient closely. When interviewing for residencies, individuals with PHA were significantly more likely to consider it “extremely” or “very” important that the program has family planning training (p < 0.001). However, the personal and clinical experience cumulatively gained from each year of residency did not impact PHA. This suggests that the personal experience with abortion happens before residency and that this experience influences program selection and future practice goals. Ob/Gyn residents with PHA were also more likely to be committed to offer abortion services themselves after graduation. As such, they sought out residency programs that were “opt-out” and sought additional training in family planning to gain more experience. Possibly, these individuals may have realized the effect that abortion had on their careers, their lives, or on the life of those close to them. Further research is needed to follow these residents through training and beyond.

Notably, despite abortion being a common procedure, affecting approximately one in three individuals assigned female at birth, nearly half of surveyed Ob/Gyn residents reported no personal experience with abortion for themselves or someone close to them. This discrepancy raises important educational and cultural considerations, suggesting that a proportion of residents may lack exposure to the lived experiences of patients seeking abortion care. Additionally, this finding raises questions about the perceived approachability of Ob/Gyn residents for abortion-related disclosures, as limited dialog may perpetuate science and stigma even within medical training environments.

While data suggested that those leaning republican had significantly lower PHA (27% vs. 53%), it is unclear if this group had overall lower abortion rates or if the subject of abortion was less discussed within that group. Unlike other investigators, we found that religion [15] and religious upbringing, as measured by regular attendance at services, size of the town and geographic area where the resident grew up, and region of the country where they are undergoing training were not correlated with PHA. These findings suggest that having had an abortion, or knowing someone close who did, is a universal experience that is not bound by religion or geography.

Implications

The findings of this study have implications for the Ob/Gyn workforce in the US. The legal landscape for abortion has significantly restricted access to care, training opportunities, and the ability to provide evidence-based abortion care. It is important to emphasize the finding that nearly half of survey respondents indicated PHA, among which training and practice goals are informed by personal experiences. For these residents, the ability to access abortion training and provide abortion care has become uncertain in the last several years. As of August 2024, 14 states have banned abortion and 14 others have implemented restrictive legislation [16]. Given the proportion of Ob/Gyn residents that have an interest in family planning training and abortion care informed by PHA, the lack of training opportunities has the potential to further exacerbate the well-documented burden of moral injury and distress in the field of Ob/Gyn [6, 17].

Limitations

This survey has several limitations. The study did not adequately address the impact of race and ethnicity on PHA. The data, collected for a study on intention to provide abortion, also did not differentiate between having a personal history of abortion and knowing someone close who did. We deliberately chose to be inclusive of female partners of male residents as such information can help gain additional insights. While our sample size was well distributed geographically and represented every region of the United States nearly equally, we did not, however, sample every program. Furthermore, as a cross-sectional survey, causality could not be established. It is also possible that there may be residual confounders that we were unable to control for, such as a bias in residents who chose to partake in this survey.

Conclusion

Having personally had an abortion, or knowing someone close who did, can influence an Ob/Gyn’s training and practice goals. These findings are important in the context of a legal landscape that restricts abortion care in jurisdictions in which approximately 40% of Ob/Gyn residents practice. Residents should be supported in their pursuit of family planning training to address the ongoing moral injury crisis in the field.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization: KJK and MAR; formal analysis: SO and CB; funding acquisition: MAR; methodology: SO and MAR; resources: KJK, CB, and MR; software: SO and CM; supervision: MAR; validation: MR and CM; writing–original draft: KJK; writing–review: MAR, CM, and DG.

Funding

This research was supported by an NIH-Women’s Reproductive Health Research Career Development Award (K-12HD001254).

Data availability

The data used in this paper are available from the corresponding author (M.-A.R.) upon reasonable written request.

Declarations

Ethics approval and consent to participate

This study was approved by the institutional review board at Wayne State University, Detroit, MI (IRB#074617MP4X) and involves an anonymous Internet survey. Informed consent was obtained electronically prior to initiating the survey. All data were immediately deidentified. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

Consent for publication

Informed consent was obtained electronically prior to the start of the survey.

Competing interest

The authors declare no competing interests.