Are Obstetrics and Gynecology Residents Equipped to Care for Transgender and Gender Nonconforming Patients? A National Survey Study

Abstract

Purpose: Our study aims to assess three self-reported outcomes: (1) comfort of, (2) competency in, and (3) curricular satisfaction of OB-GYN residents in caring for transgender and gender nonconforming (TGNC) patients.

Methods: This was a cross-sectional survey of a convenience sample of OB-GYN residents consisting of 28 questions on a 4-point Likert scale. The survey was distributed to OB-GYN residents via residency program directors and coordinators. Descriptive statistics and multivariate linear regression modeling were performed to identify demographic and training characteristics associated with differences in comfort, competency, and curricular satisfaction.

Results: One-hundred twenty-six surveys were completed by OB-GYN residents (response rate=12.6%). Composite mean scores were calculated in the three self-reported outcome domains: comfort (2.8±0.67), competency (2.7±0.61), and satisfaction (2.2±0.82) which correlate to being “somewhat not” and “somewhat” comfortable, competent, and satisfied. Trainees who identified as lesbian, gay, bisexual, or queer were found to have higher comfort scores. Older age and male gender identity were associated with higher competency scores. No significant differences in comfort, competency, and satisfaction scores between residency training level were observed. The majority (78.1%, N=89) of trainees “strongly agreed” that it was important for them to obtain training in TGNC care topics.

Conclusion: OB-GYN residents strongly agreed that learning about care for TGNC patients was important. Residents reported being more competent and comfortable than satisfied, which suggests that further curricular and clinical exposure is necessary to address the unique health care needs of this underserved patient population and to meet the educational needs of OB-GYN residents.

Introduction

Over 1 million transgender and gender nonconforming (TGNC) people live in the United States with this number likely underestimating the true size of the population given societal stigmatization.^1,2 TGNC individuals face profound harassment and discrimination in nearly all aspects of their lives. Almost a quarter of TGNC people reported being refused medical care, and half reported needing to teach their medical providers about transgender care.³ As a consequence of these negative experiences, many patients report postponing preventative and necessary medical care.⁴

The care of TGNC people may include primary care physicians, plastic and reconstructive surgeons, urologic surgeons, endocrinologists, psychiatrists, clinical psychologists, pediatricians, and OB-GYNs.^5–9 In 2011, the American Congress of Obstetricians and Gynecologists (ACOG) published a committee opinion that voiced support for OB-GYNs participating in the care for transgender patients and opposed gender identity-based discrimination.¹⁰ This support was followed in 2013 with the Council on Resident Education in Obstetrics and Gynecology (CREOG) updating the educational objectives for residents to include training in the care for transgender patients.¹¹ A survey of OB-GYN residency program directors conducted in 2016 showed that greater than 80% believed that the CREOG objectives regarding the care for TGNC patients were important. However, many felt that they lacked the expertise to provide training in the care of TGNC people to residents.¹²

Our study aims to assess three self-reported outcomes: (1) comfort of, (2) competency in, and (3) curricular satisfaction of OB-GYN residents in caring for TGNC patients.

Methods

A validated survey tool to assess knowledge of care for TGNC patients does not currently exist. For this study, an interdisciplinary research team was formed that consisted of a women's health nurse practitioner, a medical student with interest and experience in transgender health and advocacy, a current OB-GYN resident physician, an attending in urogynecology who coordinates the OB-GYN medical school clerkship, and an attending in allergy and immunology who is the medical director of the Northwell Health Center for Transgender Care.

The overall structure of our survey was modeled after Zelin et al.¹³ that assessed medical students' comfort and competency using a series of parallel questions on a 4-point Likert scale. The survey tool used for our study consisted of 28 questions designed to assess the comfort, competency, and curricular satisfaction of OB-GYN residents with the care of TGNC patients on a 4-point Likert scale (Fig. 1). Our Likert scale was defined as “not” (1), “somewhat not” (2),” “somewhat” (3), and “very” (4) comfortable, competent, and satisfied.

FIG. 1.

Survey tool distributed to OB-GYN residents. Questions 1–12 and 14–16 were asked on a 4-point Likert scale defined as “not” (1), “somewhat not” (2),” “somewhat” (3), and “very” (4) comfortable, competent, and satisfied for self-reported comfort, competency, and curricular satisfaction.

Demographic and training characteristics of participants were collected and included respondent age (in years), gender identity, sex assigned at birth, sexual orientation, region of origin, medical school location, training year, residency program name, and postgraduate fellowship interests. Gender identity, sex assigned at birth, and sexual orientation questions were answered in short free-text form. Region of origin (i.e., home state) was coded according to the U.S. Census Bureau regions: Northeast, Midwest, South, and West.¹⁴ Residency program location was coded according to the CREOG districts: District 1 (Connecticut, Maine, Massachusetts, Newfoundland, New Hampshire, New York, Nova Scotia, Quebec, Rhode Island, and Vermont); District 2 (Delaware, Indiana, Kentucky, Michigan, New Jersey, Ohio, Ontario, and Pennsylvania); District 3 (District of Columbia, Florida, Georgia, Maryland, North Carolina, Puerto Rico, South Carolina, Virginia, and West Virginia); District 4 (Alabama, Arkansas, Illinois, Iowa, Kansas, Louisiana, Manitoba, Minnesota, Mississippi, Missouri, Nebraska, Oklahoma, Tennessee, Texas, and Wisconsin); and District 5 (Alberta, Arizona, Armed Forces District, British Columbia, California, Colorado, Hawaii, Nevada, New Mexico, Oregon, Utah, and Washington).¹⁵

Our study was determined to be exempt by the Northwell Health Institutional Review Board (IRB). Before distribution, the survey tool was piloted by selected residents for terminology comprehension, length, and readability. The survey tool was then distributed electronically to all allopathic and osteopathic OB-GYN residency program coordinators and program directors in the United States via publicly available email addresses on the Accreditation Council for Graduate Medical Education (ACGME) and Doximity lists.^16,17 An initial participation request was emailed to program coordinators and directors on July 2019 with follow-up participation requests sent both 1 and 2 months later. All attempts were made to find the most recent and up-to-date email addresses for program coordinators and program directors, respectively. The survey response rate was calculated according to the number of program coordinators that responded to the initial and/or follow-up emails with confirmation of receipt and the number of residents to whom the survey had been distributed.

Survey results were collected and managed using Research Electronic Data Capture (REDCap) hosted at Northwell Health. Descriptive statistics, Cronbach's alpha statistic for internal reliability, paired-samples t-tests, and multivariate linear regression with stepwise modeling were performed using IBM SPSS Statistics, version 27 (IBM Corp., Armonk, NY). Variables were entered into our stepwise modeling for F≤0.05 and excluded for F≥0.100. Statistical significance was determined at p<0.05.

Results

Of the 269 residency programs surveyed, 37 program directors (13.8%) and/or coordinators responded to our initial email request with confirmation of receipt and delivery to their residents and the total number of residents in their program (N=1002). The overall resident survey response rate for this study was 12.6% (N=126). Twelve survey responses were excluded from our final analysis (N=114) due to incompletion of greater than 75% of the survey questions. Of these, 10 survey responses had partially incomplete demographic and training data but were included for analysis. Survey questions related to the three self-reported domains were found to be internally reliable as all alpha statistics were greater than 0.70 (questions 1–6 for comfort, α=0.820; 7–12 for competency, α=0.771; and 14–17 for curricular satisfaction, α=0.870).

The average age (in years) of the survey respondents was 29±2.3 with 30.7% (N=35) postgraduate year (PGY)-1, 22.8% (N=26) PGY-2, 17.5% (N=20) PGY-3, and 21.1% (N=24) PGY-4 (Table 1). Of the respondents, 15.8% (N=18) identified as male and 76.3% (N=87) identified as female. There were no respondents whose gender identity differed from their sex assigned at birth. In regard to sexual orientation, 73.7% (N=84) of respondents identified as heterosexual and 18.4% (N=21) identified as lesbian, gay, bisexual, or queer (LGBQ). All five CREOG districts were represented: District 1 (30.7%, N=35), District 2 (16.6%, N=19), District 3 (12.3%, N=14), District 4 (17.5%, N=20), and District 5 (12.3%, N=14).

Table 1.

Demographic and Training Characteristics of Survey Respondents

	Mean (SD, median)
Age	29 (2.3, 29)
	N (%)
Sex assigned at birth
Female	87 (76.3)
Male	18 (15.8)
Missing response	9 (7.9)
Gender identity
Female	87 76.3)
Male	18 (15.8)
Missing response	9 (7.9)
Sexual orientation
Heterosexual	84 (73.7)
Bisexual	9 (7.9)
Gay	10 (8.8)
Queer	2 (1.7)
Missing response	9 (7.9)
Region of origin
Northeast	38 (33.3)
Midwest	21 (18.4)
South	24 (21.1)
West	19 (16.7)
Non-United States	3 (2.6)
Missing response	9 (7.9)
Medical school location
Northeast	38 (33.3)
Midwest	23 (20.2)
South	30 (26.3)
West	11 (9.6)
Non-United States	3 (2.6)
Missing response	9 (7.9)
Residency CREOG district
1	35 (30.7)
2	19 (16.6)
3	14 (12.3)
4	20 (17.5)
5	14 (12.3)
Missing response	12 (10.5)
Residency year
PGY-1	35 (30.7)
PGY-2	26 (22.8)
PGY-3	20 (17.5)
PGY-4	24 (21.1)
Missing response	9 (7.9)
Presence of TGNC care center at residency site
Yes	52 (45.6)
No	59 (51.8)
Missing response	3 (2.6)
Fellowship interests
Female pelvic medicine and reconstructive surgery	5 (4.4)
Gynecologic oncology	10 (8.8)
Maternal-fetal medicine	9 (7.9)
Reproductive endocrinology and infertility	10 (8.8)
Family planning	12 (10.5)
Pediatric and adolescent gynecology	1 (0.9)
Minimally invasive gynecologic surgery	8 (7.0)
Other (i.e., specialist in OB-GYN, global health, administration, etcetera)	49 (43.0)
Missing response	10 (8.8)

Region of origin and medical school location were coded as per U.S. Census Guidelines. Residency locations were coded according to CREOG districts: (1) Connecticut, Maine, Massachusetts, Newfoundland, New Hampshire, New York, Nova Scotia, Quebec, Rhode Island, Vermont; (2) Delaware, Indiana, Kentucky, Michigan, New Jersey, Ohio, Ontario, Pennsylvania; (3) District of Columbia, Florida, Georgia, Maryland, North Carolina, Puerto Rico, South Carolina, Virginia, West Virginia; (4) Alabama, Arkansas, Illinois, Iowa, Kansas, Louisiana, Manitoba, Minnesota, Mississippi, Nebraska, Oklahoma, Tennessee, Texas, Wisconsin; (5) Alberta, Arizona, Armed Forces District, British Columbia, California, Colorado, Hawaii, Nevada, New Mexico, Oregon, Utah, Washington.

CREOG, Council on Resident Education in Obstetrics and Gynecology; PGY, postgraduate year; TGNC, transgender and gender nonconforming.

The average self-reported comfort score was 2.8±0.67. The average competency score was 2.7±0.61. The average curricular and didactic satisfaction score was 2.2±0.82. A score between two and three on our Likert scale correlated with “somewhat not” and “somewhat” comfortable, competent, and satisfied, respectively. The full Likert score range (1–4) was utilized in all questions.

Residents' comfort and competency levels were compared in the specific content areas. There were significant differences identified in residents' comfort and competency scores with counseling patients on hormone therapy, surgical options, and natal organ screening guidelines. Residents felt more comfortable (M=2.0, SD=1.03) than competent (M=1.7, SD=0.87) in regards to hormone therapy (t[3.281], p=0.001). Regarding surgical options, comfort scores (M=2.2, SD=1.03) were higher than competency scores (M=2.0, SD=0.94); t(2.838), p=0.005. Comfort with natal organ screening guidelines (M=3.1, SD=0.95) were also higher than competency scores (M=2.9, SD=0.97); t(2.557), p=0.012.

Multivariate regression models were constructed to identify demographic and training characteristics associated with differences in self-reported comfort, competency, and curricular satisfaction. No characteristics were significant to residents' curricular satisfaction scores. In regard to the comfort score, sexual orientation was significant in our regression model [F(1, 80)=7.418, p=0.008] with an adjusted R² of 0.073 (Table 2). Residents' comfort scores increased (β=0.497, confidence interval [95% CI] 0.134–0.860, p=0.008) if they identified as LGBQ. In regard to the competency score, age and gender identity were significant in our regression model [F(2, 79)=9.012, p<0.001] with an adjusted R² of 0.165 (Table 3). Residents' competency scores increased if they were older (β=0.082, 95% CI 0.029–0.135, p=0.003) and if they identified as male (β=0.478, 95% CI 0.155–0.800, p=0.004).

Table 2.

Predictors for Self-reported Comfort Score in a Stepwise Multivariable Linear Regression Model

Characteristic	β	95% CI	p
Age	0.146	—	0.212
Gender identity	0.103	—	0.368
Sexual orientation	0.497	0.134 to 0.860	0.008
Region of origin	−0.100	—	0.358
Medical school region	0.037	—	0.731
CREOG region	0.109	—	0.331
Presence of TGNC care center at institution	−0.08	—	0.940
PGY training level	0.020	—	0.852
Fellowship interests	0.130	—	0.227
Adjusted R2=0.073

Of note, sex assigned at birth was not included in our regression model as there was no discordance between sex assigned at birth and gender identity in our survey population.

CI, confidence interval.

Table 3.

Predictors for Competency Score in a Stepwise Multivariable Linear Regression Model

Characteristic	β	95% CI	p
Age (years)	0.082	0.029 to 0.135	0.003
Gender identity	0.478	0.155 to 0.800	0.004
Sexual orientation	−0.056	—	0.638
Region of origin	−0.036	—	0.726
Medical school region	0.065	—	0.553
CREOG region	0.151	—	0.137
Presence of TGNC care center at institution	0.034	—	0.744
PGY training level	0.008	—	0.947
Fellowship interests	0.086	—	0.401
Adjusted R2=0.165

Of note, sex assigned at birth was not included in our regression model as there was no discordance between sex assigned at birth and gender identity in our survey population.

The majority of respondents (78.1%, N=89) “strongly agreed” that it was important for them to obtain training in TGNC care, followed by 15.8% (N=18) “somewhat agree,” 1.8% (N=2) “somewhat disagree,” and 1.8% (N=2) “strongly disagree.” Approximately half (45.6%, N=52) of residents reported that their health system had a transgender clinic/program in place whereas 51.8% (N=59) reported that there was not.

Discussion

Transgender and gender nonconforming patients face many struggles in obtaining comprehensive and compassionate care due to a combination of factors, including but not limited to societal stigmatization, health care provider harassment, and lack of adequate health care provider education.^4,18,19 OB-GYNs play a crucial role in the medical and surgical management of TGNC patients.^10,20,21 There is minimal research focused on the ability of OB-GYN residents to care for this unique patient population. In this study, we assessed OB-GYN residents' self-reported comfort of, competency in, and curricular satisfaction with caring for TGNC patients.

Barriers to improving OB-GYN resident readiness to care for TGNC patients may include factors such as lack of institutional resources and minimal access to this patient population. A cross-sectional survey of OB-GYN program directors highlighted that approximately half of programs stated that they offered TGNC care didactics as a component of their residency curricula. In programs that did not offer transgender health education, only 43% stated that they were aware of a transgender population in their communities requiring services despite population research illustrating that TGNC individuals are present in every state in the United States.^1,2,22 Studies have been conducted in other medical and surgical disciplines have yielded similar results in which there appears to be a high degree of trainee interest in caring for TGNC patients juxtaposed with minimal educational exposure.^8,23–25

Our study results demonstrate that OB-GYN residents feel it is important to obtain training in the care for TGNC patients. The majority of residents “strongly agreed” that it was important for them to receive training in TGNC patient care. These results are similar to the responses from trainees in other fields, such as plastic surgery and urology.^26,27 Our results also suggest that there is concordance between OB-GYN residents and program directors regarding the importance of education pertinent to the care for patients who identify as TGNC.²² Of note, almost half of residents reported the presence of a transgender health center at their institution, but its presence or absence was not associated with changes in self-reported comfort, competency, or curricular satisfaction scores. Future survey studies may inquire about the involvement of OB-GYN residents with these health centers (i.e., opportunities for elective rotations, specialty clinics, etcetera) if present.

Residents who identified as LGBQ had higher comfort scores compared with residents who identified as heterosexual. This may be due to increased personal and professional levels of familiarity and involvement with the LGBQ and TGNC communities.²⁸ In our study population, residents who identified as male were found to have higher self-reported competency scores compared with their female-identifying colleagues. Of note, half of the respondents who identified as male also identified as being either gay or queer. Nonetheless, these results are similar to previous studies in which individuals identifying as male tend to overestimate measures such as self-esteem, confidence, and competence compared to their female-identifying counterparts.^29,30 Specific to obstetrics and gynecology, female residents underestimate their surgical ability when compared with their male counterparts despite no objective differences found on standardized skills exams.³¹ Our results are similar to those of Zelin et al.¹³ in which medical students who identified as LGBQ had significantly higher mean comfort and competency scores compared with their colleagues who identified as heterosexual. In their study population, medical students who identified as male also had higher levels of self-reported competency and curricular satisfaction.

Overall, OB-GYN residents felt an intermediate level of comfort and competency in caring for patients identifying as TGNC. These results are similar to those observed in medical students surveyed by Zelin et al.¹³ Of note, residents felt more competent than comfortable in regard to counseling their patients on options for hormone therapy, gender-affirming surgeries, and natal organ screening guidelines. This may represent a need for further standardized curriculum development in these areas.

Our study is limited inherently by its low response rate, lack of survey validation, survey design, and low response rate. Responder bias cannot be ignored given the low survey response compared with other uncompensated survey studies. It is possible that residents who were more familiar at baseline with the content material were more intrinsically motivated to respond to the survey. Conversely, transphobia may have negatively impacted our response rate and residents' self-reported comfort, competency, and curricular satisfaction scores. This effect was observed in work by Stroumsa et al.³² that found transphobia was associated with decreased provider knowledge of care for patients who identify as TGNC.

The timing of our survey distribution may also have negatively impacted our survey response rate and generalizability. Residency programs transition between their training classes during the middle of June and July. The survey tool was distributed at the beginning of July, and as a result, we may not have captured the cumulative didactic sessions, clinical experiences, or knowledge gained by respondents in their fourth PGY (PGY-4). The effect of this may be negligible as our multivariate models did not show any significant difference in comfort, competency, and curricular satisfaction by PGY level. This indicates that the self-reported scores of these three domains do not increase with training year.

To our knowledge, a validated survey tool assessing self-perceived comfort and competency and their relationships to positive clinical outcomes in the TGNC patient population does not exist. As such, it is difficult to comment on how associations between demographic characteristics and differences in self-reported scores translate into residents' ability to care for patients who identify as TGNC.

In regard to survey design, the authors also acknowledge that the survey tool may not adequately reflect health care issues salient to the TGNC community as the survey was not reviewed by an individual of TGNC experience. Questions inquiring about gender identity and sex assigned at birth were asked in a short free-text form as opposed to providing specific response options (e.g., male, female, transgender male-to-female, transgender female-to-male, gender nonconforming). In our survey population, none of the respondents identified as transgender based on their free-text responses but future studies should use the standardized “two-step method” for ascertaining transgender status (i.e., step one: assigned sex at birth, step two: current gender identity) with specific response options and one free-text option as recommended by The Williams Institute.³³

In conclusion, this survey study reports on the comfort of, competency in, and curricular satisfaction of OB-GYN residents in caring for TGNC patients. Residents have a high level of interest in learning about this patient population, have intermediate self-reported levels of comfort of and competency in caring for them, but are comparatively less satisfied with the education they have received thus far regarding the care for TGNC patients. The findings from our study may serve as a platform for further research and curricular development in helping OB-GYN residents provide more comprehensive and compassionate care for this underserved population.

Abbreviations Used

ACGME

Accreditation Council for Graduate Medical Education

ACOG

American Congress of Obstetricians and Gynecologists

CREOG

Council on Resident Education in Obstetrics and Gynecology

PGY

postgraduate year

REDCap

Research Electronic Data Capture

TGNC

transgender and gender nonconforming

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

Cite this article as: Qin LA, Estevez SL, Radcliffe E, Shan WW, Rabin JM, Rosenthal DW (2021) Are obstetrics and gynecology residents equipped to care for transgender and gender nonconforming patients? A national survey study, Transgender Health 6:4, 194–200, DOI: 10.1089/trgh.2020.0063.