Examining Health Care Mobility of Transgender Veterans Across the Veterans Health Administration

Abstract

Purpose: Transgender veterans are overrepresented in the Veterans Health Administration (VHA) compared with in the general population. Utilization of multiple different health care systems, or health care mobility, can affect care coordination and potentially affect outcomes, either positively or negatively. This study examines whether transgender veterans are more or less health care mobile than nontransgender veterans and compares the patterns of geographic mobility in these groups.

Methods: Using an established cohort (n = 5,414,109), we identified 2890 transgender veterans from VHA electronic health records from 2000 to 2012. We compared transgender and nontransgender veterans on sociodemographic, clinical, and health care system-level measures and conducted conditional logistic regression models of mobility.

Results: Transgender veterans were more likely to be younger, White, homeless, have depressive disorders, post-traumatic stress disorder (PTSD), and hepatitis C. Transgender veterans were more likely to have been health care mobile (9.9%) than nontransgender veterans (5.2%) (unadjusted odds ratio = 2.02, 95% confidence interval = 1.73–2.36). In a multivariable model, transgender status, being separated/divorced, receiving care in less-complex facilities, and diagnoses of depression, PTSD, or hepatitis C were associated with more mobility, whereas older age was associated with less mobility. For the top three health care systems utilized, a larger proportion of transgender veterans visited a second health care system in a different state (56.2%) than nontransgender veterans (37.5%).

Conclusions: Transgender veterans were more likely to be health care mobile and more likely to travel out of state for health care services. They were also more likely to have complex chronic health conditions that require multidisciplinary care.

Introduction

More than 1 million individuals in the United States identify as transgender,¹ meaning that their gender identity is different from the sex they were assigned at birth.² Transgender individuals may be overrepresented in the United States veteran population,³ with diagnoses related to transgender status being five times more prevalent in the Veterans Health Administration (VHA) than in the general population.⁴ Approximately 134,300 transgender individuals have veteran or retired National Guard or Reserve status from the U.S. military.⁵ Transgender veterans who utilize VHA services compared with nontransgender veterans are more likely to have lower incomes; to have experienced military sexual trauma, homelessness, or incarceration; and to have depression, serious mental illness, and post-traumatic stress disorder (PTSD).⁶ Suicide-related events alone are more than 20 times higher among transgender VHA veterans (4000/100,000) than the general VHA population (202/100,000).⁴

Transgender individuals face significant barriers to obtaining health care, including lack of affordability of services, fear of disrespect or discrimination, delays in receiving care, and needing to travel to receive care, and several of these challenges have been noted by transgender veterans receiving VHA care.^3,7,8 Transgender veterans have reported ambivalence about disclosing their gender identities when interacting with providers,⁹ many of whom have limited comfort with and training around gender identity constructs and transgender-specific health care (e.g., gender-affirming hormone therapy).¹⁰ This anticipation of and experience of barriers to care affects transgender veterans' decisions about care utilization at the VHA.¹¹ In response to these challenges, the VHA recently developed system-level training and resources around health care for transgender veterans with the aim of reducing the aforementioned barriers.¹² The implementation of these policies may help to explain that, despite many challenges to care, the majority of transgender veterans (nearly 8 out of 10) in a national study reported satisfaction with their medical care at the VHA.^8,13

Because of complex medical and mental health needs as well as unique barriers to accessing comprehensive care, transgender veterans' patterns of utilizing health care may be substantially different from nontransgender veterans, regardless of their overall satisfaction with care. One such pattern of utilization is the use of more than one health care system, referred to here as health care mobility.¹⁴ The characterization of health care mobility is limited, in both its dimensions (e.g., frequency, intensity) and its relationship to other types of geographic mobility (e.g., residential mobility). The reasons why an individual may be health care mobile are complex and varied. Studies suggest that among the general population, health care mobility is related to accessibility and availability of service, lack or presence of social support in a specific location, or perceptions of and satisfaction with care.^15–21 The relationship between residential and health care mobility has not been thoroughly characterized. For example, some individuals seek health care in different locations as a result of moving their residence,^16,17 whereas others travel for health care without changing their residence.^15,20 As there is limited characterization of health care mobility, the effect of health care mobility on outcomes is equivocal in the general and veteran population. Some studies suggest that health care mobility may have detrimental effects, such as on medication adherence or cost of care,^17,22,23 whereas other studies suggest that health care mobility may improve health by increasing access to comprehensive or patient-centered care.¹⁶

To a limited degree, health care mobility has been quantified in veteran populations among those utilizing different health care systems both within and outside the Department of Veterans Affairs (VA)/VHA.^21,24–26 The VHA provides opportunities to characterize health care mobility among veterans, as it is a national, integrated health care delivery network comprising hospitals and community-based outpatient clinics across the United States. Each hospital is connected with one or more outpatient clinics; together, they are considered a “health care system.” In the VHA delivery network, there are 152 health care systems, comprising 170 medical centers and more than 1000 outpatient facilities, with systems specializing in specific health needs.²⁶ Using this expansive network of “health care systems,” health care mobility can thus be quantified within the VHA. Our objective was to determine whether transgender veterans are more health care mobile than nontransgender veterans within the VHA delivery network, and to compare the patterns of geographic mobility in these groups. Based on the unique health needs of transgender veterans, we hypothesize that transgender veterans are more likely to be health care mobile as compared with nontransgender veterans.

Methods

Data source

We conducted a secondary analysis by using retrospective cohort study data with 1 year of follow-up from the musculoskeletal diagnosis cohort (MSD cohort, 2000–2012). The MSD cohort is one of the largest cohorts in the VHA, comprising 5,414,109 veterans and accounting for more than 51% of all veterans in care during the observation period. The MSD cohort was constructed from the VHA electronic health record (EHR) and administrative data sources to identify and characterize veterans in care with a pain-related health diagnosis.²⁷ The institutional review board of VA Connecticut Healthcare System approved this study with a Health Insurance Portability and Accountability Act waiver and waiver of informed consent.

The VHA data do not contain standardized self-identified gender identity information. We identified 2890 transgender veterans from inpatient and outpatient EHR data for the years 2000–2012 by using previously validated methods based on gender identity disorder (GID)-related diagnoses as defined by the International Classification of Diseases, Ninth Revision (ICD-9) codes.^28,29 These codes include 302.3-Transvestic Fetishism (n = 259; 8.9%), 302.6-Gender Identity Disorder of Childhood (n = 843; 29.2%), and 302.85-Gender Identity Disorder in Adolescence or Adulthood (n = 1788; 61.8%).^4,6 This method has been replicated by the Centers for Medicare and Medicaid Services.³⁰

Design

We defined the index date for transgender veterans as the first outpatient or inpatient record with a GID diagnosis. Nontransgender veterans (N = 8670) were defined as veterans without GID diagnoses; we specifically use the term “nontransgender” as there is no collection of other gender identity in the VHA. Nontransgender veterans were randomly selected in a 3 to 1 ratio, matching on the source of the record (e.g., inpatient or outpatient) and receipt of care in the same year as the GID index date of the transgender veteran. For example, if a veteran had an outpatient GID diagnosis documented in the EHR in 2000, then three nontransgender veterans were randomly selected among those who entered the MSD cohort in 2000 with an outpatient visit. We used 1 year of follow-up data to capture health care mobility over time.

Variables of interest

Health care mobility was determined by counting the number of different health care systems (i.e., one hospital and its affiliated outpatient clinics) veterans used over the 1-year follow-up period. For example, a veteran who used the Pittsburgh, PA Veterans Affairs Medical Center (VAMC), and then the West Haven, CT VAMC, was defined as health care mobile, whereas a veteran who obtained health care only at Pittsburgh PA VAMC during the 1-year follow-up was not health care mobile. In addition, a patient who used care through Pittsburgh VAMC and at the Washington County VA Clinic (affiliated with Pittsburgh VAMC) would not be defined as health care mobile. The health care mobility variable was dichotomized as any mobility or none.

We extracted sociodemographic and clinical characteristics associated with health care mobility.^14,21 Characteristics included age, race/ethnicity, marital status, and Operation Enduring Freedom/Operation Iraqi Freedom military service. We used gender recorded in the VHA EHR. It should be noted, however, that the VHA does not distinguish between assigned sex at birth and self-identified gender identity³¹; reported gender could, thus, indicate either sex assigned at birth or self-identified gender. We included housing stability, defined as the receipt of outpatient or inpatient VA homelessness services, or ICD-9 code V60.0 for homelessness.³² For clinical characteristics, we assessed pain by using the 0–10 pain intensity numeric rating scale (NRS),^33,34 recorded at baseline entry into the MSD cohort study. The ICD-9 codes were used to identify comorbid mental health (e.g., PTSD and depressive disorders) and medical conditions (e.g., hypertension, coronary artery disease, diabetes, chronic obstructive pulmonary disease (COPD), hepatitis C virus, and HIV). These characteristics were identified in the 12 months before or 6 months after entry into the MSD cohort.

Characteristics of the health care system (the facility at the index date) included urban/rural location^35,36 and facility complexity level. The complexity level is based on a weighted score of several criteria, including clinical services at a facility, the number of specialized clinical services (e.g., spinal cord injury, blind rehabilitation, cardiac surgery, interventional cardiac catheterization lab, neurosurgery, transplant, and radiation oncology), the volume of patients, and patient case mix (diagnosis-based measure that reflects the diversity and clinical complexity of all the patients in the system).³⁷ Facility complexity levels for all health care systems are 1a (most complex), 1b, 1c, 2, or 3 (least complex).

Analysis

Patient characteristics and health care system measures were summarized as frequencies and percentages. Bivariate analyses examining differences in the distribution of patient and health care system characteristics by transgender status were conducted by using chi-square tests adjusted for matching. We used conditional logistic regression among those without missing data to model the association of transgender status, patient characteristics, and health care system characteristics with a dichotomous outcome of health care mobility. We examined this association by conducting individual models to estimate the odds ratio for transgender status, controlling for one covariate at a time. This provides a measure of the influence of the covariate on the size of the odds ratio for transgender status. We estimated a full model by using backward elimination that included all characteristics from the individual models that were statistically significant at the p < 0.05 level, which was required for retention in the full model.

To describe the dimensions of health care mobility for transgender veterans, such as frequency and distance, we subset the sample to include only transgender veterans who used two health care systems. To illustrate frequency of mobility, we quantified the number of months during which transgender veterans utilized services in two health care systems. We quantified the percent of transgender veterans moving across states to use health care services, and then developed a Circos plot to visualize the patterns of mobility to and from these two health care systems.

We grouped the health care systems into the nine United States Census regions: New England (Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont), Mid Atlantic (New Jersey, New York, and Pennsylvania), East North Central (Illinois, Indiana, Michigan, Ohio, and Wisconsin), West North Central (Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, and South Dakota), South Atlantic (Delaware, Florida, Georgia, Maryland, North Carolina, South Carolina, Virginia, District of Columbia, and West Virginia), East South Central (Alabama, Kentucky, Mississippi, and Tennessee), West South Central (Arkansas, Louisiana, Oklahoma, and Texas), Mountain (Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, and Wyoming), and Pacific (Alaska, California, Hawaii, Oregon, and Washington).³⁸ All analyses were conducted by using SAS™ (SAS Institute Inc., Cary, NC).

Results

Transgender veterans were twice as likely to be health care mobile (9.9%) compared with nontransgender veterans (5.2%) over the study period (unadjusted odds ratio 2.02, 95% confidence interval [CI] 1.73–2.36). They were also more likely to be younger and White compared with nontransgender individuals (p < 0.001). The prevalence of homelessness (7.8 vs. 2.8%, p < 0.001), depression (49.2 vs. 17.3%, p < 0.001), PTSD (19.4 vs. 9.5% p < 0.001), and hepatitis C (4.0 vs. 2.9%, p = 0.002) was also higher among the transgender veterans. Transgender veterans were more likely to have missing pain NRS scores (33% vs. 24%, p < 0.001), but among those with NRS scores, they were more likely to report moderate or severe pain (54% vs. 44%, p < 0.001). Conversely, transgender veterans were less likely to have a diagnosis of hypertension, coronary artery disease, or diabetes (Table 1).

Table 1.

Level of Health Care Mobility and Sociodemographic and Clinical Characteristics by Transgender Status

Characteristic	Transgender veteran (n = 2890)		Nontransgender veteran (n = 8670)		pa
Mobility (no. of outpatient health care systems)
Zero or one	2604	90.1	8220	94.8
Two	265	9.2	425	4.9	<0.001
Three or more	21	0.7	25	0.3
Gender
Male	1997	69.1	8125	93.7	<0.001
Female	893	30.9	545	6.3
Age
18–25	117	4.1	297	3.4	<0.001
26–40	522	18.1	960	11.1
41–50	848	29.3	1288	14.9
51–60	1013	35.1	2139	24.7
61–70	300	10.4	1941	22.4
71 and over	90	3.1	2045	23.6
Race
White	2441	84.5	6286	72.5	<0.001
Black	183	6.3	1395	16.1
Hispanic	80	2.8	421	4.9
Other	105	3.6	238	2.8
Unknown	81	2.8	330	3.8
Marital status
Married	710	24.6	4645	53.6	<0.001
Not married	732	25.3	1011	11.7
Separated/divorced	1322	45.7	2306	26.6
Widowed	125	4.3	672	7.8
Unknown	1	0.1	36	0.4
OEF/OIF	144	5.0	676	7.8	<0.001
Homeless	227	7.8	238	2.8	<0.001
Health care system location
Urban	2731	94.5	8161	94.1	<0.033
Rural	76	2.6	300	3.5
Missing	83	2.9	209	2.4
Clinical characteristics
Hypertension	959	33.2	4175	48.2	<0.001
Coronary artery disease	237	8.2	1397	16.1	<0.001
Diabetes	413	14.3	1663	19.2	<0.001
COPD	205	7.1	737	8.5	<0.016
Depression	1421	49.2	1501	17.3	<0.001
PTSD	562	19.4	824	9.5	<0.001
Hepatitis C	117	4.0	248	2.9	0.002
Pain
None	615	21.3	2696	31.1	<0.001
Mild	279	9.6	1006	11.6
Moderate	527	18.2	1490	17.2
Severe	507	17.5	1401	16.2
Missing	962	33.3	2077	24.0
Health care system complexity
1a	1252	43.3	3226	37.2	<0.001
1b	345	11.9	1266	14.6
1c	619	21.4	1928	22.2
2	357	12.4	1247	14.4
3	317	11.0	994	11.5
No complexity level	0	0	9	0.1

^aBivariate analyses using chi-square tests adjusted for matching.

COPD, chronic obstructive pulmonary disease; OEF/OIF, Operation Enduring Freedom/Operation Iraqi Freedom military service; PTSD, post-traumatic stress disorder.

In the multivariable model (Table 2), transgender status was associated with higher health care mobility (adjusted odds ratio [AOR] 1.43, 95% CI 1.20–1.71, p < 0.001). Sociodemographic covariates associated with higher odds of health care mobility included being separated/divorced (Ref: married, AOR 1.29, 95% CI 1.07–1.55, p = 0.008), whereas older age (age group 61–70, AOR 0.58, 95% CI 0.44–0.77; age group 71 and over, AOR 0.44, 95% CI 0.32–0.61, p < 0.001; Ref: versus age group 18–25) was associated with lower odds of mobility. Clinical covariates associated with health care mobility included depressive disorder (AOR 1.35, 95% CI 1.13–1.61, p = 0.001), PTSD (AOR 1.34, 95% CI 1.08–1.64, p = 0.007), and hepatitis C diagnoses (AOR 1.52, 95% CI 1.08–2.14, p = 0.016). Lastly, those receiving care in a less complex health care system had greater odds of mobility (complexity level 3, or lowest level of complexity, Ref: versus complexity level 1, highest level, AOR 2.69, 95% CI 2.16–3.35, p < 0.001).

Table 2.

Association Between Being a Transgender Veteran and Health Care Mobility Adjusted for Individual Characteristics and in the Combined Full Model

Outcome is mobility: 1 = Mobile 0 = Not mobile	Individual modelsa			Full modelb
	Odds ratio	95% CI	p-Valuec	Odds ratio	95% CI	p-Valuec
Transgender veteran	2.02	1.73–2.36	<0.001	1.43	1.20–1.71	<0.001
Transgender and age (Ref = 18–25)	1.62	1.38–1.91	<0.001	Age (Ref = 18–25)
	NA
26–40	1.07	0.70–1.64	0.743	1.17	0.75–1.82	0.478
41–50	1.17	0.92–1.48	0.214	1.29	1.01–1.66	0.045
51–60	0.94	0.77–1.15	0.553	0.95	0.77–1.17	0.617
61–70	0.54	0.41–0.71	<0.001	0.58	0.44–0.77	<0.001
71 and over	0.40	0.29–0.54	<0.001	0.44	0.32–0.61	<0.001
Transgender and race (Ref = White)	2.00	1.71–2.35	<0.001
	NA
Black	1.01	0.80–1.27	0.933
Hispanic	0.59	0.36–0.95	0.030
Other	1.07	0.70–1.64	0.758
Transgender and marital status (Ref = Married)	1.82	1.55–2.15	<0.001	Marital status (Ref = Married)
	NA
Not married	1.28	1.01–1.61	0.038	1.05	0.83–1.34	0.680
Separated/divorced	1.48	1.24–1.78	<0.001	1.29	1.07–1.55	0.008
Widowed	1.00	0.71–1.40	0.991	1.31	0.92–1.86	0.131
Transgender and OEF/OIF	2.05	1.75–2.39	<0.001
	1.47	1.09–2.00	0.012
Transgender and homeless	1.96	1.68–2.30	<0.001
	1.62	1.20–2.17	0.002
Transgender and urban health care system	1.98	1.69–2.32	<0.001
	0.49	0.35–0.69	<0.001
Clinical characteristics
Transgender and hypertension	1.98	1.69–2.31	<0.001
	0.85	0.73–1.00	0.048
Transgender and coronary artery disease	1.99	1.70–2.32	<0.001
	0.79	0.62–1.00	0.051
Transgender and diabetes	2.02	1.73–2.37	<0.001
	1.01	0.83–1.23	0.938
Transgender and COPD	2.03	1.74–2.37	<0.001
	1.10	0.85–1.43	0.468
Transgender and depression	1.72	1.46–2.03	<0.001	Depression
	1.63	1.38–1.92	<0.001	1.35	1.13–1.61	0.001
Transgender and PTSD	1.90	1.63–2.23	<0.001	PTSD
	1.68	1.38–2.05	<0.001	1.34	1.08–1.64	0.007
Transgender and Hepatitis C	2.01	1.72–2.35	<0.001	Hepatitis C
	1.74	1.25–2.42	0.001	1.52	1.08–2.14	0.016
Transgender and pain (Ref = None)	1.91	1.57–2.31	<0.001
	NA
Mild	1.26	0.93–1.70	0.134
Moderate	1.50	1.18–1.92	0.001
Severe	1.85	1.47–2.34	<0.001
Transgender and health care system complexity (Ref = 1a)	2.08	1.78–2.44	<0.001	Complexity (Ref = 1a)
	NA
1b	1.23	0.96–1.57	0.104	1.22	0.95–1.57	0.116
1c	0.92	0.73–1.15	0.456	0.93	0.74–1.17	0.518
2	1.61	1.28–2.03	<0.001	1.66	1.31–2.09	<0.001
3	2.62	2.11–3.25	<0.001	2.69	2.16–3.35	<0.001

^aIndividual models examine the transgender odds ratio, adjusting for one other factor. This provides a measure of the influence of the covariate on the size of the odds ratio for transgender status.

^bFull model: Variable was included in backward elimination model if the odds ratio was ≥1.30 or ≤0.75 and p < 0.05. Pain was excluded due to missing data (26.3%).

^cConditional logistic regression models, among those without missing data.

CI, confidence interval; NA, not applicable.

Exploring the dimensions of health care mobility among veterans who used only two health care systems, transgender veterans were more likely than nontransgender veterans to travel to the two health care systems over a 1-year follow-up period (Table 3). Within the sample of those who used only two health care systems, transgender veterans used 73% of the health care systems in the VHA, and nontransgender veterans used 79% of the health care systems in the VHA. Compared with 3.8% of nontransgender veterans, 11.0% of transgender veterans traveled to the two health care systems in 4 to 5 months out of the year; 6.0% of transgender veterans compared with 3.3% of nontransgender veterans traveled to both health care systems 6 or more months out of the year. The health care systems used most frequently by transgender veterans and nontransgender veterans, respectively, were identified. For the top three facilities used by transgender veterans, 56.2% of the transgender veterans also used another health care system that was out-of-state. For the top three health care systems used by nontransgender veterans, 37.5% also used another health care system that was out-of-state.

Table 3.

Number of Months That Two Health Care Systems Were Used Over 1 Year of Follow-up Among Transgender and Nontransgender Veterans (n = 690)

No. of months that two health care systems were useda	Transgender veteranb(n = 265)		Nontransgender veteranb(n = 425)
	n	%	n	%
Zero	85	32.1	184	43.3
One	83	31.3	113	26.6
Two	33	12.4	71	16.7
Three	19	7.2	27	6.3
Four to five	29	11.0	16	3.8
Six or more	16	6.0	14	3.3

^aEach month the number of health care systems visited is counted (maximum of two). Then, the sum for the 12-month period is totaled to give the number of months.

Note that individuals had a zero count because they visited two facilities but not within the same month.

^bn is based on the number of individuals reporting two health care systems in Table 1.

A Circos plot was then used to illustrate mobility between U.S. regions among transgender veterans (Fig. 1). Regions are indicated along the perimeter. The colored ribbons connecting regions reflect travel between two regions, with ribbon width indicating the number of veterans traveling along that route, and ribbon color indicating the region of origin. The plot illustrates more intra-regional travel than inter-regional travel in general, but in some regions, there is more travel outside of the region than within. For example, a wide light blue line from the South Atlantic region reconnects with the South Atlantic region, illustrating intra-regional travel; however, the thinner light blue lines from the South Atlantic region that connect to other regions, illustrating inter-regional travel, account for more of the total travel from the South Atlantic region. The Circos plot thus indicates that there are more transgender veterans traveling to facilities outside of the South Atlantic than within the South Atlantic.

FIG. 1.

Circos plot illustrating health care mobility among transgender veterans between the U.S. regions. ENC, East North Central (Illinois, Indiana, Michigan, Ohio, and Wisconsin); ESC, East South Central (Alabama, Kentucky, Mississippi, and Tennessee); MA, Mid Atlantic (New Jersey, New York, and Pennsylvania); MT, Mountain (Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah, and Wyoming); NE, New England (Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont); PAC, Pacific (Alaska, California, Hawaii, Oregon, and Washington); SA, South Atlantic (Delaware, Florida, Georgia, Maryland, North Carolina, South Carolina, Virginia, District of Columbia, and West Virginia); WNC, West North Central (Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, and South Dakota); WSC, West South Central (Arkansas, Louisiana, Oklahoma, and Texas).

Discussion

In this study, transgender veterans were more likely to be health care mobile than nontransgender veterans, with 40% greater odds of mobility compared with their nontransgender peers after adjusting for clinical, demographic, and facility-level characteristics. Transgender veterans were also more likely to travel between health care systems in two states, and traveled more frequently between two facilities in the follow-up period than their nontransgender peers. Transgender people have unique health care needs (e.g., hormone therapy), for which many providers lack training and expertise,³⁹ and have demonstrated medical and mental health disparities compared with nontransgender veterans.⁶ Their increased health care mobility may contribute to these health disparities, and this mobility may have important ramifications regarding continuity of care, treatment decisions, and, ultimately, health care outcomes.^15–17

Our main finding that transgender veterans had higher health care mobility is likely multifactorial. Past research in nonveteran and nontransgender individuals describes several motivations for health care mobility, including, but not limited to, seeking specialty care services, accessibility and availability of care, and quality of care.^15–17 For transgender veterans, motivations for mobility are also likely influenced by their experiences and needs as transgender individuals.¹¹ Moreover, it is unclear whether the mobility was patient- or provider-driven (i.e., a patient seeking specific care or specific providers on their own, or providers directing patients to different health care systems via formal care referrals, or a combination of both). For example, as health care mobility was measured in the year after GID diagnosis, it is possible that the increased mobility is related to establishing gender-affirming care for transgender individuals. In addition, our data suggest that transgender veterans were more likely to have complex chronic health conditions, such as hepatitis C, pain, and PTSD, that require multidisciplinary care and may require services at different facilities with greater frequency.

The health care mobility pattern of transgender veterans, including greater frequency of travel and traveling out of state, is also notable. These travel patterns may reflect transgender veterans' individual decisions, or the influence of their social network on where to seek expertise and health care services within the VHA.⁴⁰ Future work can seek to examine social network structure among transgender veterans. In addition, previous research in veteran and nonveteran populations suggests that use of multiple systems likely results in increased costs of care for the patient and unnecessary laboratory tests.^23,41–43 Future studies can explore how distance and frequency of travel between more than one health care system influences satisfaction with care, costs and quality of care, and health outcomes.

Future work

It is important to note that the study period predates several significant initiatives implemented by the VA/VHA to expand clinical resources for transgender individuals and education for health care providers.^12,44,45 In 2011, the VHA implemented a national directive to provide health care to transgender veterans in a manner that is consistent with their self-identified gender identity, and to offer coverage for medically necessary health care for transgender veterans⁴⁵; in 2015, the VHA implemented a nationwide electronic consultation service for providers to use when needing additional guidance for transgender-related care (e.g., hormone therapy).⁴⁴ Although the study period includes the 2011 national directive, the study predates the overall adoption of policies aimed at improving transgender health care at the VHA. The findings from this study, thus, provide a baseline for future studies to assess trends in health care mobility, and the impact of these programs over time on health care mobility among transgender veterans.

Future studies should also investigate how health care mobility affects health outcomes in the transgender veteran population, as well as how transgender veterans seeking care in other locations relates to seeking transgender-specific care and/or care for chronic health conditions. Further, it is important to examine not only transgender veterans' motivation for seeking and utilizing health care in specific locations, but also providers' knowledge of and influence in this mobility to better understand determinants of health outcomes and identify potential points of intervention.

Limitations

There are several limitations to the study. Because VHA data do not contain self-identified gender identity data, the reliance on ICD codes—though established to be fairly valid in identifying transgender patients—likely underestimates the number of transgender veterans.²⁸ In addition, this lack of self-reported gender identity data likely misclassifies some individuals in the nontransgender category, such as nonbinary or gender nonconforming veterans who have not received an ICD code and may identify as neither transgender nor cisgender. Further, the lack of two separate fields for current gender identity and sex assigned at birth undermines the validity of reported gender in the VHA EHR.⁴⁶ The cohort used for this study may also limit generalizability to other transgender or nontransgender veterans, because the MSD cohort was created to focus specifically on musculoskeletal disorders; however, the percentage of transgender veterans in this sample is similar to other studies conducted in the VHA with transgender individuals.⁴⁷ Lastly, our results should be interpreted in the context of the study period, before important VA/VHA initiatives for transgender individuals, and should not be used to infer the current status of transgender veterans.

Conclusion

Ensuring high-quality continuity of care across all VHA facilities is important for optimal patient care and system-level effectiveness and efficiency. To our knowledge, this is the first study to find disparities in health care mobility between transgender and nontransgender patient populations, raising several areas of further inquiry to better understand the factors driving these differences. Although the VA/VHA has taken steps to implement clinical education, training, and resources to improve transgender care, these initiatives are recently implemented. The effects of these measures—if any—on mobility are largely unknown, offering a compelling opportunity for future research.

Disclaimer

The views expressed in this article are those of the authors and do not represent the views of the U.S. Department of Veterans Affairs or the National Institutes of Health.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

At the time of this article, K.H.W. was funded by CTSA Grant Number KL2 TR001862 from the National Center for Advancing Translational Sciences, a component of the National Institutes of Health. A.P. was funded by the Robert Wood Johnson Foundation Future of Nursing Scholars Program. This material is based on work supported by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, and Health Services Research and Development # CRE12-012, and Pain Research, Informatics, Multi-morbidities, and Education (PRIME) Center.