key: cord-0050689-9cty2uho
authors: MOSESON, Heidi; FIX, Laura; RAGOSTA, Sachiko; FORSBERG, Hannah; HASTINGS, Jen; STOEFFLER, Ari; LUNN, Mitchell R.; FLENTJE, Annesa; CAPRIOTTI, Matthew R.; LUBENSKY, Micah E.; OBEDIN-MALIVER, Juno
title: Abortion experiences and preferences of transgender, nonbinary, and gender-expansive people in the United States
date: 2020-09-25
journal: Am J Obstet Gynecol
DOI: 10.1016/j.ajog.2020.09.035
sha: 66e08fccec81ff3d906f83a9d2dcdf94d8e32579
doc_id: 50689
cord_uid: 9cty2uho

BACKGROUND: Transgender, nonbinary, and gender-expansive (TGE) people who were assigned female or intersex at birth experience pregnancy and have abortions. No data have been published on individual abortion experiences or preferences of this understudied population. OBJECTIVE(S): To fill existing evidence gaps on the abortion experiences and preferences of TGE people in the United States to inform policies and practices to improve access to and quality of abortion care for this population. STUDY DESIGN: In 2019, we recruited TGE people assigned female or intersex at birth and aged 18 years and older from across the United States to participate in an online survey about sexual and reproductive health recruited through The PRIDE Study and online postings. We descriptively analyzed closed- and open-ended survey responses related to pregnancy history, abortion experiences, preferences for abortion method, recommendations to improve abortion care for TGE people, and respondent sociodemographic characteristics. RESULTS: The majority of the 1,694 respondents were less than 30 years of age. Respondents represented multiple gender identities and sexual orientations and resided across all four United States Census Regions. Overall, 210 (12%) respondents had ever been pregnant; these 210 reported 421 total pregnancies, of which 92 (22%) ended in abortion. For respondents’ most recent abortion, 41 (61%) were surgical, 23 (34%) were medication, and 3 (4.5%) used another method (primarily herbal). Most recent abortions took place at or before nine weeks gestation (n=41, 61%). If they were to need an abortion today, respondents preferred medication abortion to surgical abortion three to one (n=703 versus n=217), but 514 (30%) respondents did not know which method they would prefer. Reasons for medication abortion preference among the 703 respondents included a belief that it is the least invasive method (n=553, 79%) and the most private method (n=388, 55%). To improve accessibility and quality of abortion care for TGE patients, respondents most frequently recommended that abortion clinics adopt gender-neutral or gender-affirming intake forms, that providers utilize gender-neutral language, and that greater privacy be incorporated into the clinic. CONCLUSION(S): These data contribute significantly to the evidence base on individual experiences of and preferences for abortion care for TGE people. Findings can be used to adapt abortion care to better include and affirm the experiences of this underserved population.

Transgender, nonbinary, and gender-expansive (TGE) people experience pregnancy and 103 need abortions. [1] [2] [3] Transgender is an umbrella term that describes a person whose gender identity 104 (e.g., man, nonbinary, woman) differs from the sex they were assigned at birth (i.e., female, 105 intersex, male) which is typically based on external genitalia. Cisgender describes a person 106 whose gender identity aligns with the sex they were assigned at birth. Nonbinary and gender-107 expansive are also umbrella terms that describe gender identities that are not limited to man or 108 woman -they could be a combination of both or neither. Transgender people are thought to 109 make up at least 0.6% of the total United States population or 1.4 million people. 4 This 110 proportion may be higher among younger people, especially when including nonbinary and 111 gender-expansive identities: a recent study found that 2% of 18-34-year-olds identified as 112 transgender, 8% identified as agender, bigender, genderfluid or genderqueer, and another 2% 113 identified as unsure or questioning. 5 In short, 12% of those in this age group identified as 114 transgender or gender non-conforming. 5 Population level data do not exist on the number of TGE 115 people in the United States capable of pregnancy. The majority of TGE individuals assigned 116 female sex at birth do not have surgeries to remove their internal reproductive organs (i.e., 117 uterus, ovaries, and fallopian tubes), 6,7 and some report having sperm-producing sexual 118 partners. 3, 8, 9 As a result, a substantial proportion of TGE individuals assigned female sex at birth 119 may need pregnancy and/or abortion care during their lives. Similarly, people with intersex 120 conditions or differences in sex development (DSD) -a heterogeneous group that may or may 121 not also be TGE identified -may also need pregnancy and/or abortion care during their lives. 10,11 122 Although current studies estimate that one quarter of all (presumably cisgender) women will 123 have an abortion in the United States, 12 no corresponding population-level data exist on the 124 J o u r n a l P r e -p r o o f abortion rate among TGE people who can get pregnant. The best approximation, from all known 125 abortion-providing facilities in the United States, estimated that there were between 462 and 530 126 transgender and nonbinary abortion patients nationwide in 2017. This incidence estimate, 127 however, is likely an underestimate as not all providers collected data on the patients' gender 128 identities and/or sex assigned at birth -necessary to identify TGE people. 2,13 129

Several studies have published data on abortions experienced by TGE people in the United 130

States. 14,15 A survey of 450 transgender and gender non-conforming adults who were assigned 131 female sex at birth found that 28 (6%) reported having at least one unplanned pregnancy, and of 132 these, 10 (32%) ended in abortion. 15 In a mixed-methods study of 197 masculine identified 133 people who were assigned female sex at birth, 32 (16%) participants reported 60 lifetime 134 pregnancies, of which 7 (12%) ended in abortion. 14 We are not aware of any studies that describe 135 the abortion types that TGE patients have had, the gestational ages at which abortion care was 136 accessed, or preferences for abortion care. 137

There are well-established barriers to general health care for TGE people, including 138 discrimination based on gender identity in clinics, limited provider knowledge, refusal of care 139 provision, lower rates of insurance coverage than the general United States population, and more 140 frequent discrepancies between gender presentation/identity and sex/gender indicated on 141 administrative documents compared to cisgender women. [16] [17] [18] [19] [20] [21] [22] [23] These barriers result in delays, 142 denials, and extra charges for care. 18, 21, 22, 24 These same barriers likely hinder access to abortion 143 care. [24] [25] [26] [27] [28] [29] To begin addressing these barriers to care, foundational epidemiological data on 144 abortion -a major pregnancy and reproductive health outcome 30 -among TGE individuals are 145 needed to inform the adaptation of abortion care. Stakeholders -including researchers, health 146 care providers, and community members -have called for these data. 24, 31, 32 design and format have been described in detail elsewhere. 35 All survey questions allowed for a 185 "Prefer not to say" or "I don't know" response option to ensure completeness of responses. To 186 prevent multiple responses from any participants, we enabled the "Prevent Ballot Box Stuffing" 187 feature and reviewed participant IP addresses; IP address data were subsequently deleted. 188

Participants who completed the survey were entered into a randomized drawing to win a $50 189 electronic gift card ($6,700 in gift cards were distributed in total). 190

Key variables included experiences with abortion, recommendations for improving 193 abortion care, measures of abortion method preference, and respondent sociodemographic 194 characteristics. To evaluate experiences of abortion, the survey included a pregnancy history 195 module that prompted respondents to enter each pregnancy they had experienced. For each 196 pregnancy, participants were asked whether they were trying to get pregnant and to indicate how 197 each pregnancy had ended. For respondents that reported a prior abortion, survey questions 198 assessed how many abortions and the types of abortions that they had experienced. For a 199 respondent's most recent abortion, additional survey questions inquired about the abortion type 200 and gestational age at which the abortion took place. Among those who reported a prior abortion, 201 respondents had the opportunity to indicate recommendations for improving abortion care from a 202 list of ten options, including the option to write-in a recommendation. To measure abortion 203 method preference, all respondents were asked: "If you needed an abortion now, what type of 204 abortion would you prefer?" The response choices included "medication abortion", "surgical 205 abortion", "not listed" (with an option to write-in a method), or "I don't know". The survey then 206

prompted respondents to answer the question: "What are the main reasons that this is your 207 preferred method of abortion?" Respondents could select up to three options from a multiple-208 choice list of reasons related to method privacy, cost, accessibility, pain, familiarity, and more, 209 including a write-in response. The full text of the survey has been published elsewhere. 35 210 Specific sociodemographic characteristics included age at the time of survey initiation, gender 211 identity, sex assigned at birth, intersex identity, sexual orientation, race/ethnicity, education 212 level, health insurance coverage, and region of residence. For gender identity, sexual orientation, 213 and race/ethnicity, respondents could select all options that applied, or write-in their own option. proportion of the total number exposed to study information), and 4,207 from The PRIDE Study 235 (35.3% of PRIDE participants likely eligible due to reporting female sex assignment at birth, or 236 with missing data for assigned sex at birth). In response to a question on sex assigned at birth in 237 this current survey, 2,704 of these 4,207 PRIDE participants reported having been female sex 238 assigned at birth, 1,400 reported male, eight each reported neither or preferring not to say, and 87 239 did not respond to the question. Approximately half of the PRIDE participants who responded to 240 this survey and reported having been female sex assigned at birth (50.8%) identified as cisgender 241 sexual minority women, and thus, their results are not presented here. Among all respondents to 242 the survey, 1,694 expressed a gender identity that aligned with the larger umbrella of TGE and 243 were female or intersex assigned at birth. The majority of these participants (n=1,281, 76%) were 244 recruited through The PRIDE Study, and the rest from the general public (n=413, 24%). Details 245 of study screening and recruitment are reported elsewhere. 35 246

Among the 1,694 participants, most were younger than 30 years (median=27; Table 1 ). 247

The most common gender identity was nonbinary (51%), followed by transgender man (39%), 248 and genderqueer (39%); 61% of respondents reported more than one gender identity. Most (99%) 249 respondents reported having been female sex assigned at birth, with 4% identifying as intersex. 250

Respondents reported a range of sexual orientations, most frequently queer (68%), followed by 251 bisexual (34%) and pansexual (25%). Respondents were primarily white (87%), well-educated, 252 and most (89%) had health insurance coverage. 253

For the 421 lifetime pregnancies reported across 210 (12%) respondents, 233 (55%) were 256 retrospectively reported as unintended. Of these 210 ever-pregnant respondents, 67 (32%) 257 reported at least one pregnancy ending in abortion. These 67 respondents reported a total of 92 258 abortions. Fifty-two respondents reported a single abortion, nine reported two abortions, and six 259 reported three or more (Table 2) . For respondents' most recent abortion, 41 (61%) were surgical, 260 23 (34%) were medication, and 3 (4.5%) were another method (primarily herbal). Nearly two 261 thirds of respondents' most recent abortions took place at or before nine weeks gestation (n=41, 262 61%) ( Table 2) . 263 J o u r n a l P r e -p r o o f

The 67 respondents who reported a pregnancy ending in abortion offered gender-related 266 recommendations to improve the abortion care experience as a TGE person. Specifically, 267 respondents most frequently recommended that clinics adopt gender-neutral intake forms that are 268 gender and sexual orientation affirming, and that staff utilize gender-neutral language (Table 3) . 269

Other respondent recommendations related to specific ideas for increasing the availability of 270 affirming abortion care, as well as increasing patient privacy within and outside of abortion 271 facilities. 272

When asked about abortion method preference, 703 respondents (42%) preferred 275 medication abortion over surgical (n=217, 13%) or an unlisted method (n=28, 2%) (Figure 1) , 276 while 514 respondents (30%) did not know what type of abortion they would prefer. Among the 277 28 respondents who wrote-in an unlisted method, 12 indicated that they would never get an 278 abortion because of opposition to abortion or inability to get pregnant; five indicated that they 279 would base the decision on the provider's recommendation; two stated that either method was 280 fine; and two indicated a preference for an herbal method. While medication abortion was the 281 most preferred method among both those who had experienced an abortion and those who had 282 not (45% versus 41% respectively), a higher proportion of respondents who had experienced 283 abortion reported a preference for surgical abortion than among respondents who had not 284 experienced abortion (28% versus 12%); while a lower proportion of those who had experienced 285 abortion did not know what type they would prefer (13% versus 31%). Among the 67 most 286 recent abortions, 89% of people who preferred surgical abortion had obtained a surgical abortion, 287 while only 50% of those who preferred medication abortion had obtained a medication abortion. 288

Overall, the most common reasons given for preferring medication abortion included 289 "This method is the least invasive" (n=553, 79%); "This method feels the most private" (n=388, 290 55%); and "This method does not require anesthesia" (n=231, 33%) ( Table 4 ). Thirty-one 291 respondents wrote-in a reason for preferring medication abortion, which included a desire to 292 avoid interactions with medical providers where they could be misgendered or traumatized (n=9, 293 1.3%), and the ability to manage the abortion themselves in the privacy of their own homes 294 without having to face protestors (n=6, 0.8%). 295

Among the 217 respondents who indicated a preference for surgical abortion, the most 296 common reasons included "I feel most comfortable with the type and number of medical staff 297 present for this option" (n=105, 48%); "This method would take the least amount of time (is 298 fastest)" (n=88, 41%); and "The method is the least painful" (n=40, 18%) (Table 4) . Write-in 299 responses from 38 participants who preferred surgical abortion included an aversion to the 300 hormones contained in medication abortion (n=10, 5%), a greater certainty that the abortion 301 would be a success (n=7, 3%), a desire to avoid passing the pregnancy at home (n=7, 3%), and a 302 sense that surgical would be less traumatizing than medication abortion (n=6, 3%). 303 304

These results demonstrate that TGE people assigned female or intersex at birth in the 306 United States have medication, surgical, and herbal abortions. Respondents reported nearly one 307 in five abortions occurring past the gestational limits for medication abortion (10 weeks), 37 308 which may account for the higher number of surgical abortions reported as compared to 309 medication abortions, despite a three to one preference for medication abortion. Notably, nearly 310 one third of respondents did not know what type of abortion they would prefer if they were to 311 need one today. To improve abortion care for TGE patients, respondents recommended that 312 abortion providers incorporate affirming intake forms into clinics and that staff and clinicians use 313 gender-inclusive language. 314 315

The primary limitation of this study is the lack of representativeness of the study 317 population. Because no known sampling frame exists for recruiting TGE people assigned female 318 or intersex at birth, we relied on convenience sampling. The extent to which these findings are 319 generalizable to all TGE people assigned female or intersex at birth is unknown. Additionally, 320 although 381 (22%) respondents indicated a race or ethnicity other than "white", some racial and 321 ethnic groups had low representation, and more specific studies focused on the experiences of 322 TGE people of color and the intersection of various sociodemographic characteristics is 323 warranted. Lower numbers of participants from multiple racial groups precluded our ability to 324 assess if and how these abortion experiences and preferences represent a diversity of experiences 325 -particularly when disparities in abortion care along racial lines are well established. 38 326 These limitations are balanced by strengths. This is the first quantitative study to report 327 on abortion experiences and preferences of TGE people in the United States. Further, the large 328 number of respondents, several orders of magnitude larger than prior sexual and reproductive 329 health studies among this population, 14, 15, 39, 40 provides more descriptive information than 330 previously available. The study was performed in a community-dwelling sample rather than a 331 clinical sample. The survey instrument, as well as recruitment efforts, were co-created by our 332 interdisciplinary research team in close collaboration with a Community Advisory Team 35 ; 333 community engagement was essential to reaching respondents and to ensuring that the survey 334 centered the experiences of the target populations. 335

Clinical Implications 337

The implications of these findings are that people of various gender identities and 338 experiences have abortions, and thus abortion providers must ensure that systems serve the 339 abortion needs of people with varying gender identities and experiences. Revising clinic intake 340 forms to assess capacity and desires for pregnancy in a gender-neutral way, as well as 341 systematically incorporating similar questions into conversations between providers and patients, 342 may help to identify patients capable of pregnancy and prompt pregnancy options 343 counseling. 41, 42 Several studies evaluating clinician knowledge and comfort with care provision 344 for TGE populations found self-identified gaps in provider knowledge about TGE health care, 43 345 as well as a lack of confidence, sense of preparedness, or experience with providing care to these 346 populations. 44-46 Therefore, clinicians should seek out training on how to provide gender-347 affirming sexual and reproductive healthcare for TGE patients to improve the appropriateness 348 and quality of care. Perhaps relatedly, many respondents in this study did not know which 349 abortion type they preferred, suggesting that clinicians and counselors should incorporate more 350 information about abortion options in conversations with TGE patients, including advocating for 351 and distributing abortion education materials that are inclusive of many genders, not only 352 cisgender women. 31 353

Clinicians should also consider that reasons for preferring one method of abortion over 354 another may differ for TGE patients as compared to cisgender women patients. Prior studies of 355 abortion method preference among (presumably) cisgender women, although most published 356 following the introduction of medication abortion in the United States, found that women's 357 preferences for abortion were motivated primarily by fears of bleeding, complications, or 358 anesthesia, as well as beliefs about which method was more "natural", and the time involved for 359 either method. 47 While TGE respondents shared some reasons consistent with those reported by 360 cisgender women previously, the importance of privacy and minimizing the invasiveness of the 361 experience emerged more strongly among those who preferred medication abortion -362 considerations central to TGE patients, a community commonly subjected to unnecessary 363 medical questioning, exams, or even assault on the part of providers. 16 Despite a strong preference for medication abortion, more than twice as many 373 respondents had accessed surgical abortion as compared to medication abortion. These data 374 highlight a gap between preferred abortion method and obtained abortion method -a gap that 375 future research should explore. Further, while most respondents obtained an abortion prior to ten 376 weeks gestation, one in five obtained an abortion ten weeks or later. Future research should 377 explore barriers and facilitators to abortion care generally as well as potential delays throughout 378 the process of obtaining an abortion. Finally, most abortion care research in the United States 379 focuses almost exclusively on the experiences of cisgender women, despite these and other 380 recent findings 2 that demonstrate that TGE people want, seek, and obtain abortions. These results 381 emphasize the need for greater awareness and sensitivity to the inclusion of TGE people in 382 research on abortion preferences and experiences and there is growing operational guidance 383 towards these aims. 31 More support from the clinic staff 10 15

More privacy within the clinic 9 13

More support from my provider 9 13

Better pain management during abortion 1 2

More time in recovery 1 2

None of these 14 21 538 539 540 J o u r n a l P r e -p r o o f What are the main reasons this is your preferred method of abortion? n % n % n % This method is the least invasive 556 33 553 79 1 1 Figure 1 . Abortion method preference among an online sample of transgender, nonbinary, 555 or gender expansive people assigned female or intersex at birth in the United States 556 (n=1,694) 557

Family planning and contraception use in 393 transgender men

Transgender abortion patients and the provision of 395 transgender-specific care at non-hospital facilities that provide abortions

Contraceptive 398 use and pregnancy intentions among transgender men presenting to a clinic for sex 399 workers and their families in San Francisco

Accelerating Acceptance 2017: a Harris Poll survey of Americans' acceptance of 403 LGBTQ people

Gender-Affirming Surgery and Age of Hormone Therapy Initiation Among Transgender 406 Adults. Transgend Health

Motivators and Barriers to 408

Accessing Sexual Health Care Services for Transgender/Genderqueer Individuals 409 Assigned Female Sex at Birth. Transgend Health

A mixed methods study of the sexual health needs 411 of New England transmen who have sex with nontransgender men

Bisexual, or Who Have Sex with Men: Results from Ontario, Canada

Fertility in disorders of sex development: A review

Preserving the reproductive potential of transgender and intersex 419 people

Abortion incidence and service availability in the United 421 States

A two-question method for assessing gender 423 categories in the social and medical sciences

Family planning and contraception use in 425 transgender men

Contraceptive use and abortion views among transgender 427 and gender-nonconforming individuals assigned female at birth

DC: National Center for Transgender 431 Equality

Gender-concordant identity documents and 433 mental health among transgender adults in the USA: a cross-sectional study. The Lancet 434 Public Health

Standards of Care for the 438 Health of Transsexual, Transgender, and Gender-Nonconforming People -Version 7. 439

Barriers to healthcare for transgender individuals

Self-reported discrimination in health-care 445 settings based on recognizability as transgender: A cross-sectional study among 446 transgender U.S. citizens. Arch Sex Behavior

Health and access to care 448 and coverage for lesbian, gay, bisexual, and transgender individuals in the

Stakeholder Perceptions and Experiences 452 Regarding Access to Contraception and Abortion for Transgender, Non-Binary, and 453 Gender-Expansive Individuals Assigned Female at Birth in the U.S. Arch Sex Behav

From erasure to opportunity: a qualitative 456 study of the experiences of transgender men around pregnancy and recommendations for 457 providers

Frederiksen 459 BN. Providing Quality Family Planning Services to LGBTQIA Individuals:A Systematic 460

Trans-Inclusive Abortion Care: A Manual for Operationalizing Trans-462

Inclusive Policies and Practices in an Abortion Setting, United States. FQPN, National 463 Abortion Federation

Motivators and Barriers to 466

Accessing Sexual Health Care Services for Transgender/Genderqueer Individuals 467 Assigned Female Sex at Birth. Transgender Health

A 21st-Century Public Health 472 Approach to Abortion

The Imperative for Transgender and Gender 474 Nonbinary Inclusion: Beyond Women's Health

Time for OBGYNs to care for people of all genders

Using mobile technology to engage sexual and 478 gender minorities in clinical research

A digital health research platform for community 480 engagement, recruitment, and retention of sexual and gender minority adults in a national 481 longitudinal cohort study--The PRIDE Study

Development of an affirming and customizable 484 electronic survey of sexual and reproductive health experiences for transgender and 485 gender nonbinary people

Disparities in abortion rates: a public health 493 approach

Transgender men who experienced 495 pregnancy after female-to-male gender transitioning

Contraception across the transmasculine 500 spectrum

Contraceptive counseling for transgender and 502 gender diverse people who were female sex assigned at birth. Contraception. 2020

Nurse Practitioner Knowledge, Attitudes, and Beliefs When 504 Caring for Transgender People

Transgender health in 506 endocrinology: Current status of endocrinology fellowship programs and practicing 507 clinicians

Lesbian, Gay, Bisexual, and Transgender Patient 510 Care: Medical Students' Preparedness and Comfort

Exposure to and attitudes regarding transgender education among urology 514 residents. The journal of sexual medicine

Women's reasons for choosing abortion 516 method: A systematic literature review

Commentary: No-test medication abortion: A 518 sample protocol for increasing access during a pandemic and beyond