Identifying Factors Associated With Variation in Telemental Health Delivery at Federally Qualified Health Centers

Abstract

Background:

COVID-19 accelerated federally qualified health centers’ use of telemental health. However, factors associated with telemental health service delivery remain unclear. We examined telemental health delivery by clinician type and mental health workforce composition across the U.S. to understand how staffing and other organizational characteristics are related to telemental health delivery at health centers.

Methods:

Using data from the 2021 Uniform Data System, we characterized the proportion of mental health care delivered via elemental (i.e., virtual visits) at health centers in the U.S. that received HRSA grant funding (n = 1,270) overall and by state and clinician type. Then, we conducted multivariate beta regression analyses to assess the proportion of mental health visits delivered via telemental health at health centers as a function of mental health workforce composition, while adjusting for health center size, patient mix, and state.

Results:

In 2021, health centers delivered 43% of their mental health visits via telemental health, with significant variation by state and clinician type. On average, the proportion of mental health visits delivered via telemental health was greater among psychiatrists (61%, P < .001) than psychologists (49%) and clinical social workers (45%). Factors significantly associated with the increased proportion of telemental health delivered by health centers included a higher supply of psychiatrists per patient, more grant dollars per patient, and a greater proportion of Asian patients served.

Conclusions:

Access to telemental health varies by state and across health centers based on mental health workforce composition. Future work is needed to examine funding and workforce strategies to increase telemental health service availability.

Telemental health (inclusive of video and telephonic virtual visits) was critical to delivering mental health care during the COVID-19 pandemic as in-person visit volume decreased and behavioral health needs increased.^{1 -3} In the U.S., federal and state regulatory and payment changes during the COVID-19 pandemic also facilitated a rapid increase in telemental health.^4,5 Pre-pandemic, less than 1% of behavioral health visits were delivered via telemental health compared to 39% of visits between March and August 2021. ⁶ Many outpatient health settings experienced this swift uptake of telemental health, with some expanding virtual services overnight.^7,8 This rapid change also occurred in federally qualified health centers (hereafter “health centers”) ⁸ —federally funded clinics that provide comprehensive primary care in medically underserved areas and to medically underserved populations. ⁹

Over the last decade, health centers have substantially increased their mental health workforce, and today, 99% of all health centers deliver at least some mental health care. ¹⁰ Emerging evidence also suggests that telemental health availability has increased access to mental health care at health centers as a means of pragmatically spreading limited psychiatric resources across multiple settings, allowing clinicians the ability to serve multiple locations from a single site.^8,11 However, uptake of telemental health during the pandemic varied significantly from one health center to the next, ¹¹ and factors associated with this variation are not well understood. Variation in health centers’ use of telemental health may be attributable to differences in both state Medicaid policy ¹² and organizational factors, including health center size, amount and types of grant funding received, patient and payor composition, the geographical setting of the health center, and the composition of the health center’s mental health workforce.^8,11,13

Health centers employ a diverse mental health workforce including psychiatrists, clinical psychologists, clinical social workers, and other licensed mental health professionals, ¹⁴ and the extent to which these different clinician types provide telemental health visits at health centers is unknown. While previous work suggests that telehealth delivery for primary care does not vary by clinician type (e.g., comparing advanced practice nurses, family nurse practitioners, or physician assistants to physicians), ¹⁵ those findings may not generalize to telemental health. For example, prior literature suggests some mental health professions (e.g., psychiatrists) may be more likely than others to deliver telemental health given established telehealth training and service provision opportunities. ¹⁶ Identifying variations in telemental health care delivery and in the workforce delivering telemental health care may assist with developing policies and deploying interventions to expand access to mental health care. Ultimately, increasing access to mental health care is critical in the U.S. as less than half of those who need mental health care report receiving it. ¹⁷ Telemental health is 1 way to potentially close treatment gaps and address the maldistribution of behavioral health workforce in the U.S. ¹⁸

Understanding the organizational factors of health centers that are associated with telemental health delivery could inform policies that support future growth of telemental health, an effective strategy for increasing access to mental health care. Thus, the rationale and objectives of this study were to examine the extent to which the proportion of total mental health visits delivered via telemental health varies by mental health clinician type, determine the degree of state variation in telemental health delivery overall and by clinician type, and identify factors—including mental health workforce composition—associated with the proportion of telemental health delivered by health centers. Compared to other types of mental health clinicians, we hypothesized that psychiatrists would provide a greater proportion of mental health visits via telemental health given their history of telemedicine training and payment. ¹⁹ Based on previous work, ²⁰ we also expected state and other organizational characteristics to be associated with the proportion of mental health visits delivered via telemental health at health centers.

Methods

Data Source

Data were drawn from the 2021 Uniform Data System (UDS), which collects data for all health centers nationwide that receive Health Resources and Services Administration (HRSA) Community Health Center grants. ²¹ The UDS houses administrative data reported annually by health center grantees as a requirement of their federal funding under Section 330 of the Public Service Act. ¹⁰ Data are aggregated at the grantee (i.e., organizational) level and do not describe individual clinic site data, which limits observation of variation across the organization. However, this data source presents a unique opportunity to study all federally qualified health centers in the U.S.—a program that serves 32.5 million people each year. ²² While HRSA publishes publicly available UDS data files, this study used restricted data including information on health center staffing.

Measures

Our outcome of interest was the proportion of a health center’s total mental health visits delivered via a virtual visit (as compared to an in-person visit), the total of which could range from 0–100%. We excluded health centers that had no mental health visits in 2021 (n = 26), as inclusion of these health centers would result in our outcome being undefined.

Workforce Composition

UDS data include full-time equivalent (FTE) estimates for five types of mental health clinicians: psychiatrists, psychologists, social workers, other licensed mental health clinicians, and other unlicensed mental health personnel and support personnel (e.g., certified personnel who provide counseling, treatment, or support to mental health clinicians). Not all health centers employ all types of mental health clinicians, and we excluded these from our clinician-specific measures as needed. For example, a health center with no psychiatrists was excluded when calculating the proportion of mental health visits delivered via telemental health among psychiatrists.

Covariates

The UDS also contains health center organizational characteristics, including the number of clinic sites, urban/rural location, state, total annual grant dollars, and total number of patients served. It also includes patient characteristics such as race/ethnicity (i.e., percent of a health center’s patients who identify as Black, Asian, Hispanic, American Indian/Alaskan Native, Multiracial, and Unknown), age (i.e., percent under 18, 18-64, 65 years, and older), and insurance status (i.e., percent of a health center’s patients who are uninsured or enrolled in Medicaid, Medicare, private insurance, or other public insurance).

Data Analysis

First, we calculated descriptive statistics for the health centers in our sample. Then, we described the proportion of telemental health use at the grantee level and the proportion of telemental health delivery by mental health clinician type. We also generated maps to display state-level variation in these measures. Next, we modeled the proportion of mental health visits delivered via telemental health as a function of health center characteristics using beta regression with a logit link to identify factors associated with telemental health delivery at health centers. ²³ Beta regression is a method that can model continuous proportions which was appropriate for the dependent variable-proportion of telemental health delivery. Our model specification was as follows:

$$\begin{array}{l}\%Telementa{l}_{gs}={\alpha }_0+{\beta }_{1}Clinicia{n}_g+{\beta }_{2}HealthCente{r}_g\\ +{\beta }_{3}Patien{t}_g+{\beta }_{4}Stat{e}_s+{\epsilon }_{gs}\end{array}$$

Where % Telemental is the proportion of mental health visits delivered via telemental health at health center g in state s, Clinician is a vector indicating the supply of each of our five mental health clinician types at a health center (defined as FTEs of each clinician type per 1,000 mental health patients), HealthCenter is a vector of organizational characteristics including rural location, number of clinic sites, and grant dollars per patient served), Patient is a vector of patient characteristics defined at the health center level (i.e., percent of patients by age group, race/ethnicity, and insurance status), and State is a vector of state fixed effects.

The resulting odds ratios can be interpreted as the expected increase in the proportion of telemental health use. For example, an OR = 2.0 would mean that, on average, the proportion of telemental health use was 2 times higher than the comparison. Additionally, while values of 0 and 1 are natural in real-world proportion data, they are not permissible in a beta regression because the bounds of the beta distribution are log(x) and log(1 − x). Consequently, values of 0 and 1 would result in undefined values and we followed a recommended transformation to address this. ²³ We conducted analyses in R using the betareg package. Additionally, we checked the performance of these differing regression models using the performance R package, utilizing both the performance() and check_distribution() functions which further indicated that beta regression was the best statistical approach for the given data. ²⁴ This study was deemed exempt by the University of North Carolina at Chapel Hill Institutional Review Board.

Results

Health Center and Staffing Description

In 2021, the average health center provided 94,136 visits to 23,040 unique patients across 10.6 delivery sites (see Table 1). However, there was significant variation in size, location, and patient population between health centers. Of the 1,270 grantees in the sample, 42.7% were in rural areas. On average, 58.8% of health centers patients were White, 18.9% Black, 4.0% Asian, 2.5% American Indian/Alaska Native, 2.7% Multiracial, and 13.1% Unknown race. Regarding ethnicity, health centers were, on average, 26% Hispanic. More than 21% of the patient sample was uninsured and organizations served a large share of Medicaid-enrolled patients (43%). Workforce staffing varied across health centers. On average, health centers employ 0.69 FTE of psychiatrists, 0.66 FTE of psychologists, 3.78 FTEs of mental health social workers, 3.74 FTEs of other licensed mental health clinicians, and 2.36 FTEs of other non-licensed mental health professionals.

Table 1.

Sample Description of Community Health Centers, 2021 (n = 1,270).

Characteristic	N (%) or mean (SD)	Range (if applicable)
Setting
Rural	542 (42.7%)
Urban	728 (57.3%)
Number of CHC sites	10.59 (12.11)	1.00-154.00
Race patient population
White	58.81% (27.51%)	0.89%-99.48%
Black	18.89% (22.92%)	0.00%-96.47%
Asian	4.04% (10.51%)	0.00%-97.58%
American Indian/Alaska Native	2.45% (9.26%)	0.00%-96.59%
Multiracial	2.69% (5.1%)	0.00%-68.50%
Unknown race	13.10% (14.00%)	0.00%-90.08%
Ethnicity of patient population
Hispanic	26.64% (25.65%)	0.00%-98.73%
Age (years) of patient population
Under 18	25.31% (11.42%)	0.00%-76.11%
18-64	62.29% (10.32%)	22.32%-97.67%
65 and older	12.40% (6.82%)	1.00%-45.90%
Insurance status of patient population
Uninsured	21.40% (15.29%)	0.46%-91.26%
Medicare	12.07% (7.48%)	0.18%-48.73%
Medicaid	43.24% (17.38%)	1.13%-89.96%
Other public	0.65% (2.21%)	0.00%-31.34%
Private	22.64% (12.44%)	0.00%-88.38%
Size of patient population
Number of total patients	23 039.99 (27 846.50)	749.00-247 428.00
Number of mental health patients	1981.38 (3512.93)	0.00-62 134.00
Total mental health visits	11 368.09 (25 667.53)	0.00-344 378.00
Total telemental health visits	6 166.9 (17 081.84)	0.00-215 571.00
Mental health staffing
Psychiatrist FTE	0.69 (2.32)	0.00-40.11
Psychologist FTE	0.66 (2.02)	0.00-22.10
Social Work FTE	3.78 (6.65)	0.00-83.58
Other licensed mental health clinician FTE	3.74 (11.03)	0.00-194.62
Other mental health staff FTE	2.36 (8.41)	0.00-133.80
Total grant dollars (reported in millions)	$10.05 ($15.25)	$1.01-$403.26

Abbreviations: CHC, Community Health Center; FTE, full time equivalent.

Telemental Health Delivery and Mental Health Workforce Composition

The average health center delivered 11,368 mental health visits in 2021 (12.1% of all visits), of which 43% were delivered via telemental health. However, the proportion of mental health visits delivered via telemental health at health centers ranged from 0% to 100%. Telemental health delivery also varies by clinician type and state (see Figure 1). For example, telemental health accounted for 61% of psychiatrists’ visits nationally, compared to 49% for psychologists, and 44% for social workers. Yet telemental health accounted for 100% of psychiatrists’ visits in North Dakota, but only 1% of psychiatrists’ visits in Mississippi. At the state level, inclusive of all clinician types, state rates of telemental health ranged from 10.8% to 73.9%.

Figure 1.

Proportion of mental health care delivered via telemental health within community health centers and by behavioral health clinician type.

Factors Significantly Associated With Telemental Health Delivery at Health Centers

Only a few health center characteristics in our multivariable beta regression model were significantly associated with telemental health delivery at health centers (see Table 2). Of greatest interest in this study, we found that the proportion of mental health delivered via telemental health increased on average by 1.15% for each 1 FTE of psychiatrists per 1,000 mental health patients at a health center (b = 0.14, OR = 1.15, CI = 1.04-1.28). In fact, psychiatrists were the only mental health clinician type associated with a significantly higher proportion of telemental health delivery at health centers. We also found that levels of grant funding were positively associated with telemental health delivery at health centers. Specifically, every additional $1,000 grant dollars per patient was associated with a 13% increase in the proportion of mental health visits delivered via telemental health at health centers (b = 0.12, OR = 1.13, CI = 1.01-1.27). Additionally, each 1 percentage point increase in the proportion of a health center’s patients who identify as Asian was associated with a 1% increase in telemental health delivery (b = 0.012, OR = 1.01, CI = 1.00-1.02, P < .01). Finally, state was also a significant predictor of telemental health delivery, with health centers in 38 states using significantly less telemental health than those in California. For example, health centers in West Virginia had a 79% lower proportion of telemental health delivery on average compared to health centers in California (b = -1.58, OR = 0.21, CI = 0.12-0.35).

Table 2.

Community Health Center Characteristics and Mental Health Staffing Composition Associated with the Proportion of Telemental Health Care Delivered at Health Centers.

Variable	b	SE	OR	95% CI	P
Mental health staffing composition per patient population
Psychiatrist FTE per 1k MH patients	0.14	0.05	1.15	1.04-1.28	.007
Psychologist FTE per 1k MH patients	0.05	0.03	1.05	0.99-1.12	.11
Social work FTE per 1k MH patients	−0.01	0.01	0.99	0.98-1.00	.11
Other licensed MH clinician FTE per 1k MH patients	0.02	0.01	1.02	0.99-1.04	.20
Other MH FTE per 1k MH patients	−0.01	0.01	0.99	0.97-1.00	.13
Community health center characteristics
Urban location (compared to rural)	0.14	0.10	1.15	0.95-1.39	.17
Number of sites	0.01	0.00	1.01	1.00-1.02	.05
Race/ethnicity of patient population (compared to White)
Percent Hispanic	0.004	0.002	1.00	1.00-1.01	.06
Percent Black	0.0002	0.002	1.00	1.00-1.01	.91
Percent Asian	0.01	0.004	1.01	1.00-1.02	.003
Percent AI/AN	−0.001	0.005	1.00	0.99-1.01	.83
Percent multiracial	0.003	0.01	1.00	0.99-1.02	.64
Percent unknown	0.001	0.003	1.00	1.00-1.01	.83
Age of patient population (compared to 18-64 years)
Percent under 18 years	0.0004	0.004	1.00	0.99-1.01	.97
Percent 65 years and older	−0.02	0.01	0.98	0.96-1.00	.13
Insurance status of patient population (compared to private insurance)
Percent uninsured	−0.01	0.004	0.99	0.99-1.00	.15
Percent medicare	0.01	0.01	1.01	0.98-1.03	.64
Percent Medicaid	−0.01	0.005	0.99	0.98-1.00	.13
Percent public	−0.02	0.02	0.98	0.95-1.01	.13
Total grant dollars (mil) per 1k patients	0.12	0.06	1.13	1.01-1.27	.04
State (California as reference group)
Alaska	−1.68	0.31	0.19	0.10-0.34	<.0001
Alabama	−1.61	0.34	0.20	0.10-0.39	<.0001
Arkansas	−1.65	0.39	0.19	0.09-0.41	<.0001
Arizona	−0.84	0.27	0.43	0.25-0.74	.002
Colorado	−1.39	0.29	0.25	0.14-0.44	<.0001
Connecticut	−0.30	0.31	0.74	0.40-1.38	.35
District of Columbia	−0.17	0.45	0.84	0.35-2.02	.71
Delaware	−0.10	0.67	0.90	0.24-3.35	.88
Florida	−1.30	0.24	0.27	0.17-0.43	<.0001
Georgia	−1.85	0.28	0.16	0.09-0.27	<.0001
Hawaii	−0.61	0.37	0.54	0.26-1.13	.10
Iowa	−1.32	0.37	0.27	0.14-0.52	.0001
Idaho	−1.52	0.33	0.22	0.11-0.42	<.0001
Illinois	−0.56	0.21	0.57	0.38-0.86	.007
Indiana	−0.93	0.26	0.40	0.24-0.66	.0003
Kansas	−1.51	0.32	0.22	0.12-0.41	<.0001
Kentucky	−0.97	0.29	0.38	0.21-0.68	.001
Louisiana	−1.00	0.25	0.37	0.22-0.60	.0001
Massachusetts	−0.04	0.24	0.96	0.60-1.52	.85
Maryland	−0.51	0.34	0.60	0.31-1.16	.13
Maine	−0.38	0.32	0.68	0.37-1.28	.24
Michigan	−0.88	0.23	0.42	0.26-0.65	.0001
Minnesota	−0.38	0.33	0.69	0.36-1.31	.25
Missouri	−1.37	0.27	0.26	0.15-0.43	<.0001
Mississippi	−2.13	0.34	0.12	0.06-0.23	<.0001
Montana	−1.19	0.35	0.30	0.15-0.59	.0006
North Carolina	−1.19	0.26	0.30	0.18-0.50	<.0001
North Dakota	−1.33	0.59	0.27	0.08-0.85	.03
Nebraska	−1.46	0.49	0.23	0.09-0.61	.003
New Hampshire	−0.57	0.45	0.57	0.23-1.38	.21
New Jersey	−1.23	0.28	0.29	0.17-0.50	<.0001
New Mexico	−0.54	0.32	0.58	0.31-1.10	.095
Nevada	−1.15	0.45	0.32	0.13-0.76	.01
New York	−0.63	0.20	0.53	0.36-0.78	.001
Ohio	−0.89	0.22	0.41	0.27-0.63	<.0001
Oklahoma	−1.65	0.29	0.19	0.11-0.34	<.0001
Oregon	−0.70	0.25	0.50	0.30-0.81	.005
Pennsylvania	−0.69	0.23	0.50	0.32-0.79	.003
Rhode Island	0.36	0.42	1.43	0.63-3.22	.39
South Carolina	−1.82	0.31	0.16	0.09-0.30	<.0001
South Dakota	−0.02	0.60	0.98	0.30-3.15	.97
Tennessee	−1.34	0.28	0.26	0.15-0.45	<.0001
Texas	−1.32	0.22	0.27	0.17-0.41	<.0001
Utah	−0.77	0.37	0.46	0.22-0.96	.040
Virginia	−0.81	0.28	0.45	0.26-0.77	.004
Vermont	−0.75	0.38	0.47	0.23-0.99	.046
Washington	−1.04	0.25	0.35	0.21-0.58	<.0001
Wisconsin	−0.83	0.31	0.44	0.24-0.80	.008
West Virginia	−1.61	0.27	0.20	0.12-0.34	<.0001
Wyoming	−1.39	0.60	0.25	0.08-0.80	.02

Note. CHC = Community Health Center; FTE = Full Time Equivalent; MH = Mental Health.

Discussion

The COVID-19 pandemic spurred an increase in telemental health use across the U.S., and telemental health remains a significant mechanism for delivering care.^3,25 While it may increase access to mental health care, ¹¹ telemental health availability varies significantly from 1 community to the next. ²⁶ We found that 43% of mental health visits at health centers were delivered via telemental health in 2021, but also documented significant differences in the proportion of telemental health delivery by clinician type, workforce composition, grant funding levels, and state.

The proportion of mental health services delivered via telemental health at health centers reflects a growing understanding that telemental health will remain a significant fixture in the care delivery system. Studies describing telemental health use during the COVID-19 pandemic identify telemental health service delivery ranged from 20 to 60% of all mental health care.^4,11,25 As telemental health delivery continues post-pandemic, with many pandemic-era regulations and payment issues having been resolved (e.g., payment parity, allowing for audio-only reimbursement), future work is needed to understand how and for whom this model expands access to mental health care. Given that health centers are safety net settings with an increasing proportion of mental health clinicians, supporting the continued use of telemental health is critical for addressing the national mental health crisis and being responsive to patient preferences for how to most effectively access mental health care. ²⁷

Telemental health delivery at the health centers assessed in our study varied by mental health clinician type. As hypothesized, psychiatrists provided a greater proportion of telemental health care than did other clinician types. This trend may be due to the history of telepsychiatry or the relationships health centers develop with psychiatrists to deliver care 1 day/week, thus promoting remote care. Telepsychiatry models within health centers have often been used to expand access to mental health care, particularly in rural and underserved areas. ¹⁶ At the same time, health centers had much fewer psychiatrist FTE than other types of clinicians such as clinical social workers and clinical mental health therapists. Training for telehealth is growing in graduate-level behavioral health programs, yet future work is needed to understand how training impacts telemental health delivery in the post-pandemic era.

Study findings follow previously reported trends in geographic variation in telemental health service delivery, suggesting that there are state effects impacting the use of this model. In a study conducted early in the pandemic (April 2020-June 2020), Cole ²⁰ found the highest rates of telehealth delivery occurred in Connecticut, Rhode Island, and Massachusetts, with the lowest rates in South Dakota, Kansas, and Georgia. ²⁰ Previous work suggests that telemental health availability is associated with state Medicaid payment for audio-only telehealth services. ²⁸ Varied adoption by states is also related to differences in the uptake of Medicaid reimbursement policies for live video and audio-only services over the course of the pandemic—that is, some states acted more quickly to align payment and regulation to support telehealth. ¹² For instance, although states were given broad flexibility in determining how to reimburse for virtual visits during the public health emergency ²⁹ (i.e., suspended licensing requirements, reimbursement policies, and location stipulation), these changes may impacted state variability.^12,25,29 Relatedly, states with a larger rural population may have limited broadband access, restricting telemental health delivery. However, we did not observe significant differences between rural and urban health centers in the final model—suggesting that the distribution of rural patients and health centers may be related to telemental health delivery. State variation in telemental health delivery is likely related to the interplay of several factors, and further work is needed to understand which state-based strategies increase telemental health.

Our findings are also consistent with evidence that organizations lacking technical expertise ³⁰ or sufficient telecommunications infrastructure ³¹ are more likely to struggle with telehealth implementation. Our finding that grant funding levels are positively associated with telemental health delivery suggests that organizations with more financial resources are better able to implement and expand telemental health service delivery and leverage its benefits. Funding for health centers is critical to maintaining safety net clinics and may be a factor in supporting access to telemental health care. Health centers heavily rely on revenue from third-party payers and federal funding to sustain operations in light of tight profit margins, ³² suggesting that increased federal funding for these centers could help ensure the continued availability of telemental health. As such, health centers are significantly impacted by inadequate reimbursement rates for behavioral health services and by limits on the workforce able to be reimbursed for behavioral health service delivery. Increased funding and adequate reimbursement for behavioral health care would likely improve access to telemental health care, particularly for lower-income populations without a source of usual care.

Beyond investments in health centers, investments in telehealth technologies and equipment, as well as broader digital infrastructure, are important ways to enable health centers to extend the number and types of behavioral health clinical services available. ³³ Delivery of telemental health services relies on a physical infrastructure to ensure connectivity and digital accessibility to telehealth technologies are available to all communities, including rural and remote areas. ³⁴ Specifically, broadband internet access has recently been identified as a social determinant of health, an important consideration given health centers’ patient population and related social, economic, and overall health needs. ³⁴ Future work might consider introducing technology funding opportunities specifically designed for health centers to expand access to telemental health care.

Limitations

Our findings may not generalize to health settings outside of health centers, particularly for the impact of workforce composition and health center organizational factors. There was significant heterogeneity between health centers in this study, which we were only partially able to explain. As such, continued work is needed to explore this variation. Additionally, the administrative data used in this study may be limited by reporting differences among health centers and, because the study was limited to 1 year during the pandemic, future work is needed to understand if our findings persist over time. Data were not presented on substance use disorder (SUD) virtual visits or care in this study. Although UDS does present data on SUD care at health centers, it was unclear what proportion of the behavioral health workforce included were only serving mental health care or were treating both mental and SUD needs. As mental health and SUD are intricately linked, understanding the use of telemedicine for SUD and the workforce serving this group in health centers is greatly needed. While prior work ²⁷ has assessed patient satisfaction with virtual visits delivered in health centers, this information was not available within the existing data for this study and is an area for future exploration.

Conclusion

This study intended to explore the variation of telemental health service delivery by workforce composition in U.S. health centers. Although health centers are increasingly delivering mental health care, we found that rates of telemental health varied widely between health centers and identified significant variation by clinician type, grant funding levels, and state. These findings suggest that multiple factors at the organizational, state, and federal levels influence the availability and provision of telemental health at health centers, and that multiple concurrent policy strategies are needed to ensure equitable access to telemental health care.