Hematologists’ comfort and experiences with providing care to transgender youth

Up to 3.3% of youths identify as transgender or gender diverse (hereafter referred to as transgender [TG]).^1,2 TG youths experience higher rates of violence, victimization, and discrimination than their cisgender peers.^2-4 As a result, this group experiences numerous health disparities.^2,5 Some TG individuals experience gender dysphoria, or distress related to an incongruence between their gender identity and sex assigned at birth.⁶ Gender dysphoria is associated with many adverse health conditions, with suicide being a leading cause of mortality.^7,8 Untreated gender dysphoria is associated with a ninefold increase in suicide attempts compared with the general population.⁷ Gender-affirming hormone therapy (GAHT) is a key component of treatment for gender dysphoria and results in significantly decreased suicide risk and increased sense of well-being, including improvement in depression, anxiety, quality of life, and body image.^9-11 Despite some opinions to the contrary, data suggest that GAHT is a critical intervention for TG people.^9-11 

GAHT leads to development of secondary sexual characteristics that align with the individual’s gender identity, which improves gender dysphoria. Feminizing hormone therapy uses estradiol (often in combination with testosterone blockade using spironolactone, cyproterone, or gonadotropin-releasing hormone antagonists) as the primary medication for GAHT. Masculinizing therapy is primarily testosterone GAHT.^6,12 Exogenous estrogen leads to prothrombotic changes in the hemostatic system^13,14 and increased risk of thrombosis^15,16 when used by cisgender women for contraception or hormone replacement therapy. Use of exogenous testosterone is associated with erythrocytosis in cisgender men.¹⁷ Results of studies of hormone replacement therapy for cisgender men have been mixed as to whether testosterone results in increased risk of thrombosis.^18,19 

Among TG people, there are even less data about the risk of thrombosis associated with GAHT when managed as currently recommended in guidelines from the World Professional Association for Transgender Health and the Endocrine Society.^6,12 Estradiol GAHT was associated with increased risk of venous thromboembolism, stroke, and myocardial infarction.^18,19 Testosterone GAHT was associated with increased risk of myocardial infarction in 1 study,¹⁹ although others have not found similar risk.¹⁸ Two retrospective studies of TG youth did not find a significant number of thrombosis events in TG people taking estrogen or testosterone GAHT.^20,21 

Because of these uncertainties, TG youths with preexisting risk factors for thrombosis may be referred to hematologists for evaluation of thrombosis risk before GAHT.²² In a retrospective study, >50% of TG youths starting GAHT had a preexisting risk factor for thrombosis.²⁰ Given the lack of clear data and guidelines, clinicians must make decisions about care based on limited data and data extrapolated either from cisgender or adult TG populations. Clinicians must weigh the risks of thrombosis with GAHT against the many psychological benefits associated with GAHT. For individuals at high risk of thrombosis, hematologists may consider thromboprophylaxis.^20,23 

Because the number of people identifying as TG is increasing,²⁴ hematologists may be seeing more TG youths for evaluation of thrombotic risk before GAHT. Therefore, we interviewed a sample of hematologists to explore comfort with caring for TG youth, potential interventions to improve this comfort, and experiences with recommending and prescribing thromboprophylaxis with GAHT start for TG youth with baseline increased thrombosis risk.

From 29 June 2022 through 6 November 2023, 15 hematologists who care for adolescents and young adults were recruited from pediatric and adult institutions within the footprint of the Midwest Transgender Research Collaborative (Cincinnati Children’s Medical Center, Cincinnati, OH), Nationwide Children’s Hospital (Columbus, OH), Riley Hospital for Children (Indianapolis, IN), and Children’s Hospital of Pittsburgh (Pittsburgh, PA). The inclusion criteria were (1) being a pediatric and/or an adult medical hematologist, (2) providing care to any patients aged ≤22 years, and (3) not being in medical training. Subject recruitment was performed from a random sample of hematologists from this geographic area. This list was generated from names of physicians practicing hematology or hematology/oncology listed on institutional websites. Initial contact was made via email to the business addresses of potential participants. Individual semistructured interviews were conducted via a virtual video meeting (Microsoft Teams) by 2 trained and experienced interviewers (13 interviews by T.L.K.M.; 2 by a trained clinical research coordinator). Interviews were recorded for transcription. The interview content was informed by the theory of planned behavior,²⁵ prior work on clinician behaviors by our team,^26-30 and extant literature.^18,20,31,32 The theory of planned behavior is a well-established model of health-related behavior²⁵ that has been used successfully to examine other prevention-focused physician behaviors.^27-29,33-36 The sample size of 15 was chosen to meet sample size recommendations for qualitative interview studies seeking to achieve maximum variation.³⁷ Data saturation was reached after 15 interviews for the constructs of the theory of planned behavior, which will be reported separately. The study was approved by the Cincinnati Children’s Medical Center institutional review board with waiver of written consent. Participants provided informed consent and received $100 in compensation for their time.

Interviews obtained demographic information for the participants, including medical specialty, region of training, number of years since medical school graduation, number of adolescent and young adult patients seen per week, and number of TG adolescents and young adults in their practices currently. Because clinical care of TG people is affected by clinician education^38-40 and transphobia,^41-43 participants were asked about (1) any prior training they had received in the care of gender diverse people and level at which such training was received and (2) comfort caring for TG patients; familiarity with the gender transition process; use of gender terms; and comfort asking patients about gender identity, patient’s name, and pronouns. To improve comfort with caring for TG people, participants were asked to provide potential interventions to increase the comfort of hematologists with caring for TG individuals. Finally, physician experiences providing care to TG patients at higher risk for thrombosis were elicited, including patient risk factors for thrombosis that prompted recommendations to start thromboprophylaxis and decision-making about prescribing thromboprophylaxis with GAHT start.

The recorded interviews were transcribed by independent transcriptionists and then refined by the primary interviewer (T.L.K.M.) and a trained clinical research coordinator. Field notes taken during interviews were incorporated into transcripts. All transcripts were coded by both authors (T.L.K.M.: cisgender female adolescent medicine physician; E.S.M.: cisgender male pediatric hematologist). After familiarization with the data, initial topics for coding were derived from the data. Data were then charted to the initial topics using NVivo (version 12). Topics and coding were iteratively refined by both authors. Discrepancies in coding were resolved via discussion between the coauthors. Descriptive statistics were generated in GraphPad Prism (version 10.1.2).

Among the 15 hematologists, the mean age was 47.5 years (standard deviation [SD], 8.7; range: 35-62) (Table 1). Fourteen participants (80%) were White, and 1 was Asian (7%). Two participants were Latina/o (13%). Nine participants (60%) were cisgender female. Participants had a mean of 22 years (SD, 9.5; range, 8-29) since graduation from medical school. Twelve participants (80%) were pediatric hematologists, 2 (13%) were adult hematologists, and 1 (7%) was both a pediatric and an adult hematologist (dual trained). Ten participants (67%) trained in the midwestern United States, 2 (13%) in the southeastern United States, and 1 (7%) each in the eastern United States, southwestern United States, and internationally, respectively.

Participants saw a mean of 6.1 patients aged 13 to 17 years (hereafter adolescents) per week (SD, 5.2; range, 0-17.5). Participants saw a mean of 3.3 patients aged 18 to 22 years (hereafter young adults) per week (SD, 2.9; range, 0-10). The majority of participants had provided care to TG adolescents (n = 12) and TG young adults (n = 14). Participants estimated having a mean of 3 TG adolescents (SD, 3.5; range, 0-10) and 2 TG young adults (SD, 2.6; range, 0-8) currently in their practice.

Hematologists varied in terms of whether they had received education in caring for TG youth. Three hematologists (20%) reported having had no training/education in providing care to TG youth. Three participants (20%) reported having primarily engaged in self-learning. Eleven participants (73%) reported receiving some form of education or training for caring for TG people. Seven participants (46.7%) received education at professional conferences, 7 (46.7%) reported learning at local lectures or grand rounds, 5 (33.3%) reported reading journal articles, 2 (13.3%) consulted with colleagues, 2 (13.3%) attended online/hybrid learning sessions, and 1 (7%) listened to a podcast. Most participants reported that they received their education about caring for TG youth as faculty (n = 9; 60%):“That has pretty much been solely as an attending. In fellowship, you’re. . . I don’t want to say forced to learn certain things, but you almost don’t have time to learn other things. And transgender care was unfortunately not something that was a focus.”

Only 2 participants (13%) reported education during medical training; 1 participant (7%) each had education as a fellow or as a resident and fellow:“During my fellowship, we had a clinic at [Y]. . . we definitely did see younger adults undergoing [gender] care, and. . . the most common question that was asked to us was XYZ had thrombosis while on estrogen. What do I do moving forward and what kind of anticoagulation would I recommend? So, those were my questions as a hematologist in hematology training and attending.”

No hematologists reported being uncomfortable with caring for TG youth. Seven participants (46.7%) reported being somewhat comfortable caring for TG youth:“I would go with somewhat. I would probably screw things up from time to time.”

Another participant noted that the lack of data on thrombosis risk with GAHT led to decreased comfort caring for TG youth:“I guess somewhat, [be]cause there’s so little data. The difficulty comes in not knowing what to say or recommend.”

Eight participants reported being very comfortable caring for TG youth. One participant noted that the various state limitations on caring for minor-aged TG patients decreased their comfort to some degree:“Although I must say, with all the recent legal and all of the political issues behind it, I’m not sure what our youth go through now, but like before all of this was like. . . before it became so politicized, definitely felt more comfortable with the medical aspects of it.”

Hematologists reported more variability in familiarity with the gender transition process. One participant (7%) reported being unfamiliar with the process. Most participants (10; 66.7%), reported being somewhat familiar with the process. One hematologist reported some familiarity with the gender transition process, which had been learned from their patients:“That is something that actually it would be kind of cool to learn more about. But what I end up doing is just asking the patient, you know. So, where are you, what’s the plan, you know, give me the details.”

Three participants (20%) reported being very familiar with the transition process.

Variability in understanding of gender terms was also present. One participant (7%) reported lack of familiarity with gender terms:“I’ve been Googling to learn. . . My patient said that. . . their preferred pronouns are ‘they’ and ‘them,’ and I said, ‘I know, I haven’t been doing it.’ And they said that I’m trying.”

Seven participants reported being somewhat familiar with gender terms, and 7 reported being very familiar with gender terms.

With respect to confidence in asking about gender, 1 participant (6.7%) reported lack of confidence, 6 (40%) reported being somewhat confident, and 7 (46.7%) reported being very confident in being able to ask youths about their gender. Two participants (13.3%) noted that the presence of parents in the exam room impacted confidence in asking about gender:“I mean, I have no problems asking the person, but sometimes the family in the room may have problems with that. And so, it can be a delicate issue depending on. . . Sometimes I’ve gone into a room where I know in advance that this is a transgender youth, but the parents don’t want to use the transgender terminology. They want to stick with the original name, the original sex, don’t mention this, and then everything can change at 18 [years old]. ...I’m comfortable asking, but if there are parental and family issues in the way, then I have to tread lightly.”

Overall, participants reported greater confidence in being able to ask patients about names than about pronouns. Only 1 participant (6.7%) reported being somewhat confident asking about names, whereas 13 (86.7%) reported being very confident asking about names:“Oh, very confident. I mean, that’s usually how the conversation starts. . . that’s usually the first 3 sentences: . . . what do you want me to call you?”

In contrast, 1 participant reported lack of confidence, 4 (26.7%) reported being somewhat confident, and 10 (66.7%) reported being very confident asking about pronouns:“I advertise my comfort on my badge. I have the badge buddy with my gender preferred pronouns. . . So, I feel like I advertise my allyship, and I advertise my own preferred pronoun so that helps to break the ice.”

Participants recommended several interventions to improve the comfort of hematologists with caring for TG youth. Eight participants (53.3%) suggested educational interventions including topics such as gender-related terms, gender transition process, outcomes of GAHT, navigating visits in which parents are less supportive of gender transition, and identifying mental health concerns:“I would love to hear a review of the psychiatric data. . . is there any way to pinpoint the kids who are really at risk for suicide?... That’s a legit[imate] concern because some of these patients we are anticoagulating, right? So, we’re sending [them] home with a potentially life-threatening medication.”

Four participants (26.7%) described in-clinic interventions that would improve comfort caring for TG youth, including having mental health resources available, having a reference for gender terms, developing a script for asking gender-related questions, and having staff gather information about gender and pronouns during intake:“When I see them in clinic, at least here, we have the nurses kind of doing the intake. So, it would be nice when they present or they’re ready for me to see the patient, ‘Dr [Y], we have this patient, their preferred name is, and they would like to be addressed as he, she, her, him, or their.’ I think it will be helpful as a team with the interaction to make clear how the patient preferred to be addressed prior to the visit.”

Three participants (20%) reported that identifying a youth as TG before the clinical visit would improve their comfort:“I think it helps kind of as part of the clinical summary (either for providers, nursing staff, or anyone working in the team) to kind of make that introduction [about the patient’s gender] because I think it opens up trust when you meet the patient the first time.”

Two participants (13.3%) reported that having access to clinicians specializing in gender care would improve comfort:“I have a lot to learn from behavioral medicine and [adolescent medicine] and social work on how to talk to people and how to ask questions that get answered honestly.”

Two participants (13.3%) each reported that education during medical training and having more personal experience caring for TG youth would improve their comfort caring for these patients:“More practice. I think I’d still be afraid I would offend somebody or make a mistake, would be the main thing.”

One participant each reported that having centrally available resources from the different specialties involved in gender care, having more data on the risk of thrombosis with estrogen GAHT, and learning from TG patients themselves would improve their comfort:“Just patients speaking on their own behalf about their experience and ways that providers have effectively treated them and navigated the complicated pieces of the [electronic medical record that] says ‘F’ [female] or ‘M’ [male] and not often allowing for additional details that way.”

Four hematologists (26.7%) had recommended and prescribed thromboprophylaxis to TG youths starting GAHT. All 4 hematologists had recommended thromboprophylaxis before estrogen GAHT to a total of 4 adolescents. Of those who had recommended thromboprophylaxis to adolescents, 3 hematologists (20%) had prescribed it to a total of 3 patients. One patient was prescribed aspirin, and 2 patients were prescribed rivaroxaban. One patient declined thromboprophylaxis because of concerns about bleeding related to sports participation. Participant-reported combinations of risk factors in the adolescents contributing to the recommendations for thromboprophylaxis included (1) a first-degree relative with thrombosis, morbid obesity, and smoking, (2) a personal history of thrombosis and obesity, and (3) a personal history of thrombosis and a family history of thrombosis. No patients prescribed thromboprophylaxis had known adverse events, and all were reported to be taking the medication at the time of the interviews. Two hematologists (13%) had recommended and prescribed thromboprophylaxis before estrogen GAHT to a total of 4 young adults. All of the young adults were prescribed rivaroxaban. Participant-reported combinations of risk factors contributing to the recommendations for thromboprophylaxis in the young adults included (1) history of self-management of mail-order estrogen GAHT, factor V Leiden homozygosity, personal history of thrombosis, morbid obesity, and smoking; (2) multiple first-degree relatives with thrombosis; and (3) significant protein C deficiency. No patients prescribed thromboprophylaxis had known adverse events. Three patients were reported to be taking the medication at the time of the interviews.

No participants had prescribed thromboprophylaxis to a patient starting testosterone. However, hematologists discussed that data about thrombosis risk associated with testosterone were largely lacking:“None of the patients I’ve seen starting testosterone had really risk factors that were all that bad, number 1. Number 2: the data I’ve seen, my interpretation of it, is that testosterone might be a risk, but the data is not as clear as it is for estrogen. So, I tend to be a little less worried about those patients. Whether or not that’s right or wrong I think is yet to be definitively determined.”

An additional 5 hematologists (66.7%) had evaluated TG adolescents and young adults for possible thromboprophylaxis but did not recommend it or prescribe it:“So, we did discuss thromboprophylaxis, as well as taking into account family history. . . Did they have a personal history themselves of any blood clots? . . . There was for the most part lack of. . . strong family history of blood clots and there was no personal history, which we discussed as something, you know, to take into account. But we chose to not move forward with thromboprophylaxis in those settings.”

Hematologists noted that their ability to confidently make recommendations about thromboprophylaxis were limited by the available data and need to extrapolate from studies of cisgender women:“Well, here’s what we know about estrogens and thrombosis and what I would recommend for a biological female. But you’re a biological male getting ready to go on estrogen, and we don’t really have any data about what estrogen does to biological males and whether or not it’s really thrombogenic. You’re [assigned] female [at birth], it would be pretty straightforward: we’d recommend prophylaxis. But for you, we don’t really know.”

Participants also discussed that they convey this lack of data to patients and families as they engage in shared decision making about thromboprophylaxis:“These conversations can be hard because we don’t have any great data to support the care surrounding these decisions. And, so, I’m always very open with the fact that my experience and what little literature we have and their personal preferences, anxiety around their history of blood clot, anxiety that they might have around blood clots in the future, the risks that estrogen brings all collectively have to go into the decision. And then I talk about what lack of data we have for this specific situation. We have a little bit of data surrounding exogenous estrogen in cisgender females that take exogenous estrogen for period control, menstrual suppression, birth control, contraception. And then we have some data surrounding cisgender women who take exogenous estrogen in perimenopausal, postmenopausal situations. So, 2 situations where cisgender females are already producing or had produced estrogen. But that data doesn’t specifically apply to transgender females.”

This study examined hematologists’ experiences and comfort with caring for TG adolescents and young adults. We found that almost all of the hematologists interviewed had experience with caring for TG youth, and many had experience assessing thrombosis risk in youth starting GAHT. Given the increasing prevalence of youth who identify as TG,²⁴ the number of referrals for hematologic evaluation before starting hormones would be expected to increase. Thus, more hematologists are likely to be called upon to provide assessment for, and recommendations about, thromboprophylaxis to youth at higher baseline risk of thrombosis.

Although there were low levels of education and training reported by this cohort, almost all participants expressed at least some comfort with caring for TG people. Lack of training may be contributing to the observed variability in familiarity with the gender transition process and comfort with obtaining pronouns and names. Prior studies demonstrate that clinician attitudes and behaviors toward TG patients affects provision of care and the health-seeking behaviors of TG patients.^41-43 For example, TG youths reported avoiding medical care after experiences in which they were referred to by their nonaffirmed gender and/or their nonaffirmed pronouns, and such youths reported experiencing a range of negative emotions after such encounters, including fear for their safety.⁴⁴ Gaps in clinician knowledge about care of TG people³⁸ and failure to use affirmed names and pronouns create barriers to care.^38,39 Similar to our cohort, other studies demonstrate that clinicians reported little professional training in this area^41,45 but recognized the need for more training⁴⁶ and sensitivity in communicating with TG patients.^45,46 Thus, to improve clinical care of TG youth, interventions designed to improve comfort in caring for this population are needed.

Hematologists provided suggestions for interventions at various levels that could increase comfort with caring for TG individuals. Although participants reported taking part in some self-directed and formal education, they also reported the need for more education. Education about identifying mental health concerns and which youths might be at risk for suicidal ideation was noted. In addition to risk of suicide or self-harm by other means, hematologists reported concern about suicide risk related to ingestion of thromboprophylactic medication. Interventions, such as a half-day educational session, resulted in improved comfort and knowledge with transgender medicine in other settings.⁴⁷ Another intervention to improve comfort with caring for TG youth is optimizing the electronic medical record (EMR) to identify a TG youth before a clinical visit in order to facilitate communication about gender identity to different members of the team. Incorporating questions about gender into screeners completed by adolescents was acceptable to both adolescents and parents,⁴⁸ and embedding questions about gender into clinician documentation improved recording of gender in the EMR.⁴⁹ Computer algorithms have been developed to use existing documentation in the EMR to identify TG people.^50-52 Incorporating elements of these existing interventions across sites of medical care may improve the comfort of hematologists, and likely other subspecialists, caring for TG youth.

This is one of the first studies to explore the experiences of hematologists with recommending and prescribing thromboprophylaxis to TG adolescents and young adults. In this cohort, 4 hematologists (27%) had recommended thromboprophylaxis before GAHT to 8 discreet patients. When discussing indications for thromboprophylaxis, having multiple thrombosis risk factors was typically cited as the reason for the recommendation. This is consistent with guideline recommendations in at least 1 other situation: long-distance travel.⁵³ A direct oral anticoagulant was the most commonly recommended agent for thromboprophylaxis; however, aspirin was also used. This is consistent with numerous trials supporting the use of direct oral anticoagulants for thromboprophylaxis in other clinical situations.^54-56 

Given the lack of data on thrombosis risk with long-term use of GAHT, hematologists are providing recommendations to patients based on the limited studies of GAHT use and extrapolation of data from use of estrogen and testosterone in cisgender people. Clinicians in this study discussed the challenges associated with the lack of available data or guidelines, and such gaps in knowledge likely contribute to variations in practice. In a prior study, our team found differences among pediatric hematologists with respect to recommending thromboprophylaxis to youth starting GAHT, continuing prophylaxis for youth with history of thrombosis before GAHT, and prescribed prophylactic agent.²⁰ Thromboprophylaxis was also recommended to 2 youths with personal or family history of thrombosis who were starting testosterone GAHT.²⁰ Therefore, studies assessing risk of thrombosis in the setting of GAHT are critically needed to inform clinical practice and development of practice guidelines, ultimately resulting in optimization of care of TG youth.

This study is subject to several limitations. All hematologists were recruited from the midwestern United States, which may limit generalizability to other regions of the United States or to other countries. However, the goal of qualitative research is not to generate generalizable data but rather to develop a nuanced understanding of a topic area. Similarly, although the sample size is small, it is aligned with sample size recommendations for qualitative research.

In conclusion, the majority of hematologists in this sample had experience providing clinical care to TG youth. Although many participants voiced comfort and confidence with various aspects of gender care, several potential interventions were recommended to further improve comfort caring for this population. Targeted educational offerings, optimizing EMRs to consistently identify TG patients, and developing clinical tools may improve the comfort and confidence of hematologists caring for TG adolescents and young adults, thereby improving the health of these youths.

The authors thank Andrea Meisman for her assistance with conducting interviews.

This work was supported by the National Institutes of Health (grant R01HL161153) and Cincinnati Children's Hospital Medical Center Gap Funding.

Contribution: E.S.M. conceived and designed the research, analyzed and interpreted the data, cowrote and revised the manuscript, and gave final approval of the version to be submitted; and T.L.K.M. conceived and designed the research, conducted research interviews, analyzed and interpreted the data, cowrote and revised the manuscript, and gave final approval of the version to be submitted.

Conflict-of-interest disclosure: E.S.M. has served on advisory boards for Novo Nordisk for matters unrelated to this work. T.L.K.M. received funding for an investigator-initiated research project from Gilead Sciences, Inc, that is unrelated to this work.

Correspondence: Tanya L. Kowalczyk Mullins, Division of Adolescent and Transition Medicine, Department of Pediatrics, Cincinnati Children's Hospital Medical Center, 3333 Burnet Ave, MLC 4000, Cincinnati, OH 45229; email: [email protected].