Effect of telemental health adoption on mental health services utilization in federally qualified health centers

Abstract

Introduction

Mental health issues disproportionately affect rural and low-income populations, where access to prevention and treatment is limited. Implementing telemental health services (TMHS) in Federally Qualified Health Centers (FQHCs) could improve accessibility to mental healthcare. This study assessed the effect of TMHS on mental health services utilization in FQHCs, comparing centers with and without TMHS and examined differences in TMHS effect between FQHCs that adopted TMHS before and during the pandemic.

Methods

The study utilized FQHC-level data from Uniform Data Systems annual performance reports (N = 9540 FQHC-year) and county-level demographic, health status, and provider characteristic data from 2016 to 2022. A two-way fixed effects model was applied to examine the effect of TMHS on mental health visit rates, supplemented by the Sun and Abraham difference-in-differences estimator. The analysis included both pre-COVID and during-COVID TMHS adopters.

Results

FQHCs offering TMHS showed 1.04 times increase in overall mental health visit rates compared to those without TMHS. Centers serving higher proportions of Medicaid-insured (1.07 times) and low-income (below 200% FPL; 1.05 times) populations showed even higher visit rates. During-COVID TMHS adopters showed a significant increase (1.05 times), particularly in urban areas, whereas rural adopters did not see significant changes.

Conclusion

FQHCs offering TMHS showed increased mental health visit rates, especially among low-income and Medicaid-insured populations. FQHCs that adopted TMHS during the pandemic experienced a marked rise in mental health visits, but rural-urban disparities persist, highlighting the ongoing challenges in expanding access to mental healthcare in rural settings.

Keywords

Telemental health telehealth equity mental health utilization disparities policy

Introduction

In 2020, 21% of US adults reported having experienced mental illnesses.¹ The Centers for Disease Control and Prevention (CDC) shows that from August 2020 to February 2021, the percentage of adults suffering from anxiety or depression increased from 36.4% to 41.5%, and those reporting an unmet mental healthcare need increased from 9.2% to 11.7%.² Moreover, studies show significant mental health disparities based on income levels³ and rurality.⁴ Low-income populations are 1.5 to 3 times more likely to experience depression or anxiety compared to high-income populations. Rural residents experience higher rates of depression and are less likely to access mental health services than urban residents.⁴ Despite effective prevention and treatment options to manage mental illnesses, the lack of accessible and affordable treatment resources remains a challenge.⁵ Moreover, mental health resources are not uniformly distributed; consequently, rural and geographically isolated areas are underserved.⁶ Relying on traditional modes of service delivery for mental healthcare may not meet the demand of people living in remote and underserved areas.

Telemental health services (TMHS) improve access and deliver cost-effective treatment while maintaining the quality of mental healthcare. Federally Qualified Health Centers (FQHCs) serve nearly 2.7 million patients seeking mental health services⁷ and a disproportionate share of low-income families⁸ who experience higher mental health problems and treatment disparities.³ FQHCs are primary care clinics in the USA that receive federal funding to provide healthcare services in underserved communities, in areas designated as health professional shortage areas and must meet requirements, including providing care on a sliding fee scale for low-income patients.⁹ Given the role of FQHCs as safety net primary care providers for low-income populations in urban and rural areas,¹⁰ provision of TMHS in FQHCs could improve access for remote and underserved populations.^6,11,12 However, lack of reimbursements, licensure requirements, privacy policies, and broadband gaps have historically made it challenging for FQHCs to offer TMHS.^13,14 With the declaration of COVID-19 public health emergency on January 31, 2020 (ended on May 11, 2023), regulatory relaxations and changes like Medicaid and Medicare reimbursements for telemental health services, allowing audio-only telehealth and providing telehealth services without geographic/distant site restriction specific to FQHCs supported adoption of TMHS.^13,15,16 However, most of these changes were temporary, and many of them are expected to expire by the end of 2024. Lawmakers are advocating for the continuation of these policies to maintain stable and ongoing telehealth services in FQHCs.

Our primary objective was to evaluate the effect of adopting TMHS on mental health service utilization in FQHCs to provide policy-relevant information on prolonging regulatory easing and boosting telehealth resources. Alongside, it is important to consider that the COVID-19 pandemic and subsequent PHE declaration led to state and federal telehealth policy relaxations that facilitated FQHCs to adopt and utilize telehealth to their full capabilities.¹⁷ Besides, the heightened burden of mental health disorders during the pandemic could have contributed to increased utilization of mental health services.¹⁸ To consider these pandemic-driven changes, our secondary objective was to examine if there are any differences in the effect of TMHS on mental health service utilization between FQHCs that adopted TMHS before the pandemic and those that adopted during the pandemic.

Methods

Measures and data sources

We used FQHC-level Uniform Data Systems (UDS) annual performance data from 2016 to 2022.¹⁹ Our sample was an unbalanced panel including 9540 FQHC-year observations. Among other quality and clinical data, the UDS enquires about the types of telehealth services the clinic offers, including mental health services. We constructed our key predictor (treatment) variable, a binary variable indicating the availability of TMHS in a given FQHC. We focused on five utilization-related visit rate outcomes: (1) depression, (2) anxiety, (3) attention deficit and disruptive (ADD), (4) other mental disorders, and (5) all mental health (MH)—a composite of the above categories of mental health conditions (Supplementary Appendix 1). The visit rates (visits per patient) for each of these mental health conditions were calculated by dividing the total number of visits for a specific mental health diagnosis by the number of patients with that diagnosis. FQHC-level patient composition characteristics (e.g., age, sex, race, ethnicity, insurance, and poverty levels), and facility characteristics like rural-urban location, percentage of funding from community health center grants, and the number of delivery sites were included in the analysis as controls.

County socioeconomic covariates, including median income, unemployment, uninsured rates, rurality, and long commutes, were derived from Small Area Income and Poverty Estimates²⁰ and the American Community Survey.²¹ County population health indicators like the percentage of people with frequent mental distress and the average number of poor physical health days were derived from the Behavioral Risk Factor Surveillance System.²² Lastly, county provider characteristics like primary care provider ratio and mental health provider ratio) derived from Area Health Resource Files.²³ FQHC-level data was merged with county-level data by cross-walking the FQHC's ZIP code with the county using the Department of Housing and Urban Development ZIP Code Crosswalk Files.²⁴ As ZIP codes can cross county lines, we linked ZIP codes to the county in which the highest percent of the ZIP code was located.

Analytical methods

To assess our primary objective, we constructed a treatment variable “TMHS adopters” by classifying FQHCs offering TMHS as our treatment group (N = 5262) and those that did not offer it as the control group (N = 4278). To examine the effect of offering TMHS on mental health visit rate, we used a two-way fixed effects (TWFE) model with FQHC and year fixed effects (equation 1). We conducted a series of specification tests and evaluated empirical decisions before selecting TWFE as a suitable regression model (Supplementary Appendix 2).

{VisitRates}_{ict} = B_{0} + β_{1} T M H S_{i c t} + β_{2} {DemoVars}_{ict} + β_{3} {OrgVars}_{ict} + β_{4} {CountVars}_{ct} + δ_{i} + Y_{t} + є_{ict}

(Equation 1—TWFE Poisson estimator)

VisitRates refers to the rate of MH visits (outcomes 1–5) in each FQHC. Subscripts i, c, and t refer to individual FQHC, county, and time in years from 2016 to 2022, respectively. The treatment variable, TMHS, equals 1 if an FQHC offered TMHS in a given year and 0 if it did not. The analysis includes control variables for patient demographics (DemoVars), FQHC organizational characteristics (OrgVars), and county-level factors (CountVars). Lastly, δ and Y are FQHC and year-fixed effects, and Є is a time-varying error representing unobserved factors that change over time that effect Y. The results of regression estimates are reported as Incidence Rate Ratios (IRR).

Additionally, considering that there are significant mental health utilization disparities based on income and rurality, we conducted subgroup analysis for urban and rural FQHCs and those with federal poverty level (FPL) below 200 with the same model specifications specified in equation (1).^4,25 We also conducted subgroup analyses by insurance status, specifically for Medicaid and Uninsured categories which represent core FQHC populations.²⁶ For patient composition subgroups (FPL and insurance status), we selected the median as our standard cut-off point measurement as most groups showed skewed distribution, and the median is more resistant to outliers than the mean.

To assess our secondary objective, we constructed two treatment groups: (a) “pre-COVID TMHS adopters” defined as 1 if FQHC adopted TMHS before 2020 and “0” for those FQHCs that adopted on or after 2020 and those that did not adopt at all (treatment N = 3495; control N = 6064) and (b) “during-COVID TMHS adopters” defined as 1 if FQHCs adopted on or after 2020; and “0” for those that adopted before 2020 and those that did not adopt at all (treatment N = 2176; control N = 7383). We applied the same model described in equation (1) to estimate the effect of offering TMHS on mental health visit rates for both groups. Additionally, we looked at urban and rural FQHCs separately to examine if there are any differences in the treatment effects considering that there are substantial differences in telehealth adoption and its utilization based on rurality.²⁷

For all three treatments—TMHS adopters (not accounting for COVID effect), pre-COVID TMHS adopters, and during-COVID adopters—we applied the difference-in-differences (DiD) model using the Sun and Abraham estimator to evaluate the effect of providing TMHS on mental health visit rates over time.²⁸ This estimator is particularly useful when there is staggered treatment adoption which refers to a situation where the implementation of a treatment occurs at different times across different entities. FQHCs adopted TMHS in different years, and some did not adopt it at all.

In this approach, FQHCs that offered TMHS are grouped into cohorts based on the year they started providing the service. For each cohort, the average treatment effect on the treated (ATT) is estimated by comparing the visit rates of FQHCs that adopted TMHS at a specific time to those that never adopted TMHS (the counterfactual scenario). This is done in terms of the timing of treatment, focusing on relative differences between when the service was introduced and the observed outcomes (equation (2)). By analyzing outcomes relative to the time of adoption, this method avoids contamination from spillover effects from earlier periods, providing unbiased and consistent estimates of how offering TMHS affects visit rates for each cohort.

{Visitrates}_{i c t} = β_{0} + \sum_{k = - K}^{K} β T M H S_{k} D_{i t}^{k} + β_{1} {DemoVars}_{i c t} + β_{2} {OrgVars}_{i c t} + β_{3} {CountVarsi}_{c t} + δ_{i} + Y_{t} + є_{ict}

(Equation 2—Sun and Abraham difference-in-differences estimator using ordinary least squares regression)

$V i s i t r a t e s_{i c t}$ refers to mental health visit rates for FQHC i in cohort c at year t. $\sum_{k = - K}^{K} β T M H S_{k} D_{i t}^{k}$ captures the cohort-specific ATT. It shows how average mental health visit rates change relative to when an FQHC first started offering TMHS. $D_{{it}}^{k}$ indicates whether the FQHC is in the kth time period (years) relative to TMHS adoption. k represents leads and lags relative to the year of TMHS (e.g., k = −1: one year before adoption (lead), k = 0: adoption year, k = 1: one year after adoption (lag)), and K defines the maximum range leads and lags included in the analysis. All the other terms in the equation have similar definitions as explained in equation (1). We used Stata18.0 for all statistical analyses.

Results

Descriptive characteristics for the treatment “TMHS adopters” are summarized in Table 1. FQHCs that offered TMHS experienced 1.04 times increase in all MH visit rates compared to those without TMHS (Table 2). This increase was consistent across specific conditions such as depression, anxiety, and ADD. Urban FQHCs with TMHS showed higher visit rates for all MH (1.06 times), depression (1.07 times), and anxiety (1.07 times) when compared to urban centers without such services. In contrast, rural FQHCs offering TMHS did not show significant changes in visit rates, except for ADD, where visit rates were 1.06 times higher than rural centers without TMHS. In comparison to other FQHCs, FQHCs that offered TMHS serving a larger Medicaid population showed 1.07 times increase in all MH visit rates, consistent across different MH categories. Centers serving fewer uninsured individuals also noted a positive effect of TMHS, with increases in all MH (1.05 times) and depression-specific (1.06 times) visit rates. Additionally, FQHCs serving populations with incomes below 200% FPL exhibited higher visit rates (1.06 times) for depression, anxiety, and ADD.

Table 1.

Mean and standard deviation with significant tests for mental health visits, FQHC, and county-level characteristics among FQHCs that offered telemental health service (treatment group: n = 5262) and FQHC that do not offer telemental health service (control group: n = 4278) for all years (2016–2022 combined).

Variable	Treatment group		Control group		Significance
Variable	Mean	SD	Mean	SD	Significance
Mental health visits (rate^a)
All mental health visits	3.4	1.6	2.8	1.5	***
Depression and other mood disorders	3.6	1.8	2.9	1.6	***
Anxiety disorders, including post-traumatic stress disorder (PTSD)	3.4	1.8	2.8	1.5	***
Attention deficit and disruptive (ADD) behavior disorders	3.6	2.0	3.0	2.0	***
Other mental disorders, excluding drug or alcohol dependence	2.9	1.7	2.5	1.9	***
Depression—clinical measure
% of patients screened and follow-up plan documented for depression	67.5	20.8	65.2	22.8	***
FQHC level covariates
Age (in percentage)
0–18y	25.9	13.3	27.1	13.4	***
19–44y	34.3	7.9	35.5	9.0	***
45–64y	26.6	7.1	27.2	7.9	**
Above 64y (ref)	11.7	6.9	10.0	6.3	***
Sex (in percentage)
Female (ref: male)	56.4	5.9	56.5	6.7	ns
Insurance (in percentage)
Medicaid	43.6	18.1	43.8	19.6	ns
Uninsured	22.2	16.4	25.7	19.1	***
Other insurance (ref)	33.6	17.4	29.7	17.7	***
Poverty levels(in percentage)
FPL ≤ 200	59.7	25.0	64.6	25.3	***
Other FPL (ref)	40.3	24.9	35.4	25.2	***
Race (in percentage)
Non-White	40.0	27.8	43.0	29.6	***
White (ref)	59.4	27.9	57.0	29.6	***
Ethnicity
Hispanic (ref: non-Hispanic)	26.8	26.3	28.0	27.9	**
Rurality (in percentage)^b
Rural	44.6	–	40.5	–	***
Urban	55.4	–	59.5	–	***
CHC funding (in percentage)	87.1	25.5	85.0	28.8	***
# of Delivery Sites per FQHC	10.6	11.7	7.5	8.5	***
County-level covariates
% Unemployment	5.6	2.6	5.6	1.9	ns
Median household income	59957.4	17149.2	54838.0	15411.4	***
% Uninsured	11.2	5.0	13.1	5.7	***
% Rural population	28.3	31.3	24.9	30.0	***
% Population with long commute	32.9	13.3	33.1	12.4	ns
Frequent mental distress %	13.5	2.4	12.0	1.9	***
Average number of physically unhealthy days	4.1	0.7	3.9	0.6	***
Primary care provider ratio	1645.9	1221.1	1649.1	1288.7	ns
Mental health provider ratio	744.7	1676.6	921.5	1729.5	**
Number of FQHCs	4278		5262		-

Mean for mental health visits, FQHC- and county-level characteristics among FQHCs for each year are provided in Supplementary Appendix 3. ns: non-significant.

^aRate: Number of visits by diagnosis per number of patients with diagnosis. ^bChi-square test statistic; the rest all are t-test statistics.

*p < 0.05, **p < 0.01, ***p < 0.001.

Table 2.

Two-way fixed effects Poisson regression estimates of the association between telemental health service and mental health visit rate in all FQHCs and other subgroups of population served by FQHCs.

	All MH visit rates IRR (SE)	Depression visit rate IRR (SE)	Anxiety visit rate IRR (SE)	ADD visit rate IRR (SE)	Other MH visit rate IRR (SE)
All FQHCs^a	1.04*** (0.01)	1.05*** (0.01)	1.04*** (0.01)	1.05** (0.02)	1.03 (0.01)
N	8929	8923	8921	8702	8905
Subgroups of FQHCs
Location
Rural FQHCs	1.03 (0.01)	1.03 (0.02)	1.01 (0.02)	1.06* (0.02)	1.02 (0.02)
N	3635	3631	3630	3526	3616
Urban FQHCs	1.06***	1.07***	1.07***	1.03	1.04*
	(0.01)	(0.02)	(0.01)	(0.03)	(0.02)
N	5237	5235	5234	5119	5232
Insurance status
FQHCs serving higher than the median percentage of Medicaid population	1.07*** (0.02)	1.07*** (0.02)	1.08*** (0.02)	1.07* (0.03)	1.03* (0.02)
N	3785	3779	3777	3747	3747
FQHCs serving lower than the median percentage of Medicaid population	1.02 (0.01)	1.02 (0.02)	1.00 (0.02)	1.03 (0.02)	1.01 (0.02)
N	4364	4358	4357	4186	4344
FQHCs serving higher than the median percentage of Uninsured population	1.04* (0.01)	1.03* (0.01)	1.05** (0.02)	1.05* (0.02)	1.01 (0.02)
N	4388	4382	4379	4222	4368
FQHCs serving lower than the median percentage of Uninsured population	1.05*** (0.01)	1.06** (0.02)	1.04 (0.02)	1.05 (0.03)	1.03 (0.02)
N	4337	4337	4337	4278	4333
Income level
FQHCs serving higher than the median percentage of population with FPL below 200	1.05*** (0.01)	1.06** (0.02)	1.06*** (0.02)	1.06** (0.02)	1.03 (0.02)
N	4330	4327	4326	4181	4318
FQHCs with lower than the median percentage of population with FPL below 200	1.04** (0.01)	1.04** (0.02)	1.02 (0.02)	1.03 (0.03)	1.03 (0.02)
N	4365	4362	4361	4283	4353

N refers to the sample size. Estimates are incident rate ratios (IRRs) representing the effect of TMHS on mental health visit rates in FQHCs; robust standard errors (SE) clustered at FQHC level in parentheses, MH refers to mental health, and the “all MH visit rate” refers to combined mental health visit rates of all mental health service categories.

Table with all covariates is provided in Supplementary Appendix 4.

*p < 0.05, **p < 0.01, ***p < 0.001. Bold values represent significant estimates.

Descriptive characteristics of pre-COVID TMHS adopters and during-COVID TMHS adopters are summarized in Table 3. For pre-COVID TMHS adopters, no significant effects on mental health visit rates were observed, these results were consistent across both rural and urban subgroups (Table 4). In contrast, during-COVID TMHS adopters showed significantly higher MH visit rates, with increases noted at 1.05 times for all MH visits combined and depression, and 1.06 times for anxiety and ADD visit rates. Urban FQHCs that adopted TMHS during the pandemic showed even greater increases, with visit rates being 1.09 times higher for all MH combined, depression, and anxiety, 1.07 for ADD, and 1.06 for other MH conditions. However, no significant effects were observed in rural counterparts during the same period. Further analysis using the same estimator but incorporating both pre- and during-COVID TMHS adopters as treatment covariates in a unified model, yielded similar results, validating the significant effect of TMHS on visit rates observed primarily among urban during-COVID adopters.

Table 3.

Mean and standard deviation with significant tests for mental health visits, FQHC, and county-level characteristics among FQHCs that offered TMHS pre-COVID (treatment = 3495; control = 6064) and FQHCs that offered TMHS during COVID (treatment = 2176; control = 7383).

Variable	Pre-COVID TMHS adopters		During-COVID TMHS adopters		Significance
Variable	Mean	SD	Mean	SD	Significance
Mental health visits (rate^a)
All mental health visits	3.4	1.6	3.3	1.6	**
Depression and other mood disorders	3.4	1.7	3.6	1.8	***
Anxiety disorders, including post-traumatic stress disorder (PTSD)	3.3	1.6	3.5	2.0	***
Attention deficit and disruptive (ADD) behavior disorders	3.5	2.1	3.2	1.9	***
Other mental disorders, excluding drug or alcohol dependence	2.9	1.7	2.9	1.8	ns
FQHC level covariates
Age (in percentage)
0–18	27.1	12.7	24.0	14.1	***
19–44	34.0	7.8	34.9	8.5	***
45–64	26.1	6.9	27.4	7.4	***
Above 64 (ref)	11.8	7.1	11.6	6.7	ns
Sex (in percentage)
Female (ref: male)	56.5	5.3	56.3	6.8	ns
Insurance (in percentage)
Medicaid	41.6	17.9	46.2	18.6	***
Uninsured	22.8	16.2	21.7	17.1	*
Other insurance (ref)	35.1	17.6	31.5	17.1	***
Poverty levels (in percentage)
FPL ≤ 200	58.2	25.2	61.6	24.7	***
Other FPL (ref)	41.7	25.2	38.4	24.7
Race (in percentage)
Non-White	38.2	27.5	44.1	28.2	***
White (ref)	61.7	27.5	55.8	28.2	***
Ethnicity
Hispanic (ref: non-Hispanic)	25.0	25.4	29.1	27.4	***
# of delivery sites per FQHC	11.9	12.9	8.6	9.7	***
Rurality (in percentage)^b
Rural	51.7	–	34.4	–	***
Urban	48.4	–	65.6	–	***
CHC funding (in percentage)	88.0	24.0	85.8	27.4	***
County-level covariates
% Unemployment	5.6	2.5	5.6	2.7	ns
Median household income	56923.5	15418.7	63861.9	18534.6	***
% Uninsured	12.0	5.1	10.0	4.5	***
% Rural population	32.4	32.8	22.9	28.8	***
% Population with long commute	31.2	13.2	35.4	13.1	***
Frequent mental distress %	13.1	2.4	13.9	2.4	***
Average # of physically unhealthy days	4.0	0.7	4.1	0.7	ns
Primary care provider ratio	1684.1	1252.7	1619.4	1243.6	ns
Mental health provider ratio	820.1	1899.0	656.7	1193.3	**
Number of FQHCs	3495		2176

ns: non-significant.

^aRate: Number of visits by diagnosis per number of patients with diagnosis. ^bChi-square test statistic; the rest all are t-test statistics.

*p < 0.05, **p < 0.01, ***p < 0.001.

Table 4.

Two-way fixed effects Poisson regression estimates of the association between telemental health service and mental health visit rate in FQHCs with treatment subgroups pre-COVID and during-COVID telemental health service (TMHS) adapters.

I. Association between pre-COVID TMHS adopters and mental health visit rate^a
	All MH visit rate IRR (SE)	Depression visit rate IRR (SE)	Anxiety visit rate IRR (SE)	ADD visit rate IRR (SE)	Other MH visit rate IRR (SE)
Pre-COVID all TMHS adopters	1.00 (0.02)	1.00 (0.02)	1.00 (0.02)	1.04 (0.03)	0.98 (0.03)
N	8940	8933	8930	8710	8915
Pre-COVID rural TMHS adopters	1.04 (0.03)	1.02 (0.03)	1.04 (0.03)	1.05 (0.03)	1.02 (0.05)
N	3646	3641	3639	3534	3626
Pre-COVID urban TMHS adopters	0.98 (0.03)	0.99 (0.03)	0.98 (0.03)	1.04 (0.04)	0.96 (0.05)
N	5237	5235	5234	5119	5232
II. Association between during-COVID TMHS adopters and mental health visit rate^a
During-COVID all TMHS adopters	1.05** (0.02)	1.05** (0.02)	1.06** (0.02)	1.06** (0.02)	1.03 (0.02)
N	8940	8933	8930	8710	8915
During-COVID rural TMHS adopters	1.01 (0.02)	1.01 (0.02)	1.00 (0.02)	1.03 (0.03)	1.01 (0.03)
N	3646	3641	3639	3534	3626
During-COVID urban TMHS adopters	1.09*** (0.02)	1.09*** (0.02)	1.09*** (0.03)	1.07* (0.03)	1.06* (0.03)
N	5237	5235	5234	5119	5232
III. Association between pre- and during-COVID TMHS adopters and mental health visit rate^a,b
Pre-COVID all TMHS adopters	1.02	1.03	1.03	1.07**	1.00
	(0.02)	(0.02)	(0.02)	(0.03)	(0.03)
During-COVID all TMHS adopters	1.06***	1.06***	1.06***	1.08***	1.03
	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)
N	8940	8933	8930	8710	8915
Pre-COVID rural TMHS adopters	1.04	1.03	1.04	1.06	1.03
	(0.03)	(0.03)	(0.03)	(0.04)	(0.05)
During-COVID rural TMHS adopters	1.02	1.02	1.01	1.04	1.02
	(0.02)	(0.02)	(0.02)	(0.03)	(0.03)
N	3646	3641	3639	3534	3626
Pre-COVID urban TMHS adopters	1.02	1.03	1.02	1.08*	0.98
	(0.03)	(0.03)	(0.03)	(0.04)	(0.05)
During-COVID urban TMHS adopters	1.09***	1.10***	1.09***	1.10**	1.06*
	(0.02)	(0.02)	(0.02)	(0.03)	(0.02)
N	5237	5235	5234	5119	5232

N refers to the sample size. Estimates are incident rate ratios (IRRs); robust standard errors (SE) clustered at FQHC level in parentheses MH refers to mental health.

^aTable with all covariates is provided in Supplementary Appendices 5–13. ^bBoth the treatments (pre- and during-COVID telemental health service (TMHS) adopters) are assessed in the same model as covariates to each other.

*p < 0.05, **p < 0.01, ***p < 0.001. Bold values represent significant estimates.

The Sun and Abraham DiD estimator (equation (2)) findings, illustrated in Figure 1, showed a significant increase in all MH visit rates for the first four years following TMHS adoption in FQHCs, with a slight but still positively significant decrease in the fifth year compared to FQHCs that did not adopt TMHS (plot I). Both rural and urban TMHS adopters (depicted in plots II and III, respectively) demonstrated similar trends, with urban adopters experiencing a slightly larger effect. Notably, urban adopters showed non-significant effects of TMHS in the years prior to adoption, followed by a significant increase in the years after adopting TMHS. Consistent with the results from the TWFE Poisson estimator (equation (1)), pre-COVID adopters, including their urban and rural subgroups (shown in plots IV–VI), did not exhibit a significant effect of TMHS on all MH visit rates overtime. During-COVID adopters experienced a significant increase in overall mental health visit rates in the first year following adoption, with a slight reduction in the magnitude of this effect during the second year (plot VII). This effect was more pronounced in urban during-COVID adopters (plot IX), unlike rural during-COVID that showed no effect. Similar patterns were observed for depression, anxiety, ADD, and other mental health visit rates, provided in Figures 1–4 in Supplementary Appendix 14.

Figure 1.

Sun and Abraham difference-in-differences estimator event study plots of the effects of telemental health service (TMHS) on all mental health (MH) visit rate. All MH visit rates refer to combined mental health visit rates of all mental health service categories. Plots contain estimates and 95% confidence intervals of the effects of TMHS from the estimation of equation (2). The reference time (0) was the year when TMHS was adopted by a given FQHC. Years before or after relative to TMHS adoption are expressed as leads (negative values) and lags (positive values) on the x-axis of the graph. All the plots (I–IX) use the same specifications as equation 2. Sun and Abraham estimators for other mental health visit outcomes (depression, anxiety, ADD, and other mental health visit rates) are provided in Figures 1–4 in Supplementary Appendix 14, respectively.

Discussion

In this study, we examined the effect of offering TMHS on mental health visit rates in FQHCs and differences in the effect of TMHS on visit rates between FQHCs that adopted it before the pandemic and those that adopted it during the pandemic. We found that FQHCs providing TMHS showed significantly higher rates of mental health visits overall, particularly for depression and anxiety, compared to FQHCs without such services. These findings highlight the critical role of TMHS in supporting underserved and vulnerable populations with mental health conditions. Our results are consistent with previous studies that noted a positive effect of telehealth availability on mental health visit rates at select FQHCs.^29,30

The Sun and Abraham event study estimator that evaluated the effect of TMHS overtime showed a significant effect of THMS only among during-COVID adopters, no effect was observed in the pre-COVID adopters. The pandemic-related relaxations of telehealth policies specific to FQHCs like provision of reimbursements for TMHS, allowing audio-only telehealth and providing telehealth services without geographic/distant site restriction that supported them to offer TMHS, and increased funding to FQHCs to support them in providing telehealth services.³¹ Prior research highlights how these changes reduced barriers, fostering broader adoption and utilization of telehealth.³² Since all FQHCs were subjected to the same policy relaxations, one would expect an increase in mental health visit rates across both pre-COVID and during-COVID adopters after 2020. However, we saw a positive effect only in during-COVID adopters.

A possible explanation is that FQHCs adopting TMHS during the pandemic could set up a robust telehealth infrastructure, benefitting from increased funding to support their efforts to establish telehealth services compared to those that were already offering it.³³ Additionally, during-COVID adopters served a significantly higher population of low-income and Medicaid-insured populations compared to pre-COVID adopters. This demographic difference might explain the greater effect of TMHS in these centers, as reflected in our subgroup analysis which suggests that FQHCs serving higher populations of Medicaid and low-income individuals experience more significant benefits from TMHS. Enhanced Medicaid reimbursement policies, such as coverage for behavioral health services and parity for audio-only telehealth, likely facilitated this increased uptake and effectiveness of TMHS during the pandemic.^13,34

The positive effect of TMHS in during-COVID adopters was more pronounced in urban FQHCs, and contrastingly no significant effect in their rural counterparts, highlighting urban-rural disparities in TMHS provision. Previous studies identified telehealth as a key strategy to overcome issues like the mental health professional shortages and distance barriers to a mental health center in rural areas.^35,36 While telehealth has been identified as a crucial strategy, these populations still face challenges such as limited broadband access, low digital literacy, and reduced acceptance of telehealth, which can deter TMHS utilization.^35,36 Telehealth interventions should address barriers in rural FQHCs, such as improving broadband connectivity, enhancing digital literacy, and increasing the cultural and practical acceptability of telehealth among rural communities. Such tailored strategies are vital for boosting the adoption of mental health services through telehealth in rural areas and reducing access disparities.

Limitations

First, the study analyzes data only from FQHCs and may not represent other healthcare settings. Second, the study included a period (2020–2022) when the pandemic heightened the burden of mental health disorders. Therefore, while interpreting the results, it is important to consider that the greater need for mental health services itself might have led to increased visits and not the presence of TMHS itself. Third, FQHCs that offer TMHS might have many in-house mental health services compared to those that do not offer. This could cause an important issue of confounding that is driving both the presence of TMHS and higher mental health visits. Lastly, to address endogeneity concerns, we lagged a few explanatory variables by one year, which led to the loss of the explanatory power of their effects on mental health utilization.

Conclusion

The study suggests that TMHS availability significantly increased mental health visit rates for conditions like depression and anxiety in FQHCs. Our findings reveal that FQHCs integrating TMHS during the pandemic saw a notable increase in visit rates. FQHCs serving higher populations of low-income and Medicaid individuals showed a greater effect of TMHS on visit rates, particularly those adopting TMHS during COVID, specifically indicating that they may have benefited from telehealth policy relaxations and associated expansion of telehealth services. The study also revealed significant rural-urban disparities: rural FQHCs adopting TMHS did not experience increased visit rates, unlike urban counterparts, despite a critical need for telehealth in rural areas to address mental healthcare gaps.

Supplemental Material

sj-docx-1-acr-10.1177_1357633X241307417 - Supplemental material for Effect of telemental health adoption on mental health services utilization in federally qualified health centers

Supplemental material, sj-docx-1-acr-10.1177_1357633X241307417 for Effect of telemental health adoption on mental health services utilization in federally qualified health centers by Khyathi Gadag, Kanika Arora and Whitney E Zahnd in Journal of Telemedicine and Telecare

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical approval and informed consent statements

Not applicable.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Khyathi Gadag

Data availability statement

The datasets generated during and/or analyzed during the current study are available in the archived resources of the Uniform Data System (UDS) Training and Technical Assistance, Health Resources and Services Administration (HRSA)¹⁹ [].

Supplemental material

Supplemental material for this article is available online.

References

Substance Abuse and Mental Health Services Administration. Key substance use and mental health indicators in the United States: Results from the 2020 National Survey on Drug Use and Health (HHS Publication No. PEP21-07-01-003, NSDUH Series H-56). Rockville, MD: Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration, 2021. https://www.samhsa.gov/data/ .

Vahratian

. Symptoms of anxiety or depressive disorder and use of mental health care among adults during the COVID-19 pandemic—United States, August 2020–February 2021. MMWR Morb Mortal Wkly Rep 2021; 70: 490–494.

Sareen

Afifi

McMillan

, et al. Relationship between household income and mental disorders: Findings from a population-based longitudinal study. Arch Gen Psychiatry 2011; 68: 419–427.

Carpenter-Song

Snell-Rood

. The changing context of rural America: A call to examine the impact of social change on mental health and mental health care. PS (Wash DC) 2017; 68: 503–506.

Demyttenaere

Bruffaerts

Posada-Villa

, et al. Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. J Am Med Assoc 2004; 291: 2581–2590.

Chakrabarti

. Usefulness of telepsychiatry: A critical evaluation of videoconferencing-based approaches. World J Psychiatry 2015; 5: 286–304.

Health Center Program: Impact and Growth | Bureau of Primary Health Care. https://bphc.hrsa.gov/about-health-centers/health-center-program-impact-growth (accessed 8 August 2023).

Federally Qualified Health Centers? Importance in the Safety Net Continues for Many as Affordable Care Act Implementation Moves Ahead, https://apps.urban.org/features/hrms/briefs/federally-qualified-health-centers-importance-aca.html (accessed 8 August 2023).

Federally Qualified Health Center, https://www.hhs.gov/guidance/sites/default/files/hhs-guidance-documents/FQHC-Text-Only-Factsheet.pdf (2019).

10.

Federally Qualified Health Centers (FQHCs) and the Health Center Program Overview—Rural Health Information Hub, https://www.ruralhealthinfo.org/topics/federally-qualified-health-centers (accessed 8 August 2023).

11.

Shore

. Telepsychiatry: Videoconferencing in the delivery of psychiatric care. AJP 2013; 170: 256–262.

12.

Bekerian

. Telemental health and diverse populations amid COVID-19. Encyclopedia 2023; 3: 267–278.

13.

Lombardi

de Saxe Zerden

Greeno

. Federally qualified health centers use of telehealth to deliver integrated behavioral health care during COVID-19. Community Ment Health J 2024; 60: 215–223.

14.

Shaver

. The state of telehealth before and after the COVID-19 pandemic. Prim Care 2022; 49: 517–530.

15.

King

. New House, Senate bills aim to make telehealth expansion permanent in Medicare, Medicaid. FierceHealthcare, https://www.fiercehealthcare.com/payer/new-house-senate-bills-aims-to-make-telehealth-expansion-permanent-medicare-medicaid (2021, accessed 11 December 2021).

16.

Bridges

Steggerda

Guzman

, et al. Telebehavioral health at a federally qualified health center pre- and peri-COVID-19. Psychol Serv 2024; 21: 34–41.

17.

Schofield

. Regulatory and legislative issues on telehealth. Nutr Clin Pract 2021; 36: 729–738.

18.

Adepoju

Valdez

. Trends in mental health utilization before and during the COVID-19 pandemic: Federally qualified health centers as a case study. Popul Health Manag 2023; 26: 143–148.

19.

Archived Resources | Bureau of Primary Health Care, https://bphc.hrsa.gov/data-reporting/uds-training-and-technical-assistance/archived-resources (accessed 8 August 2023).

20.

United States Census Bureau. SAIPE State and County Estimates for 2021. Census.gov, https://www.census.gov/data/datasets/2021/demo/saipe/2021-state-and-county.html (accessed 29 August 2023).

21.

United States Census Bureau. American Community Survey Data Tables. Census.gov, https://www.census.gov/programs-surveys/acs/data/data-tables.html (accessed 29 August 2023).

22.

CDC—BRFSS Annual Survey Data, https://www.cdc.gov/brfss/annual_data/annual_data.htm (2022, accessed 29 August 2023).

23.

Area Health Resources Files, https://data.hrsa.gov/data/download (accessed 9 June 2024).

24.

HUD USPS ZIP Code Crosswalk Files | HUD USER, https://www.huduser.gov/portal/datasets/usps_crosswalk.html (accessed 29 August 2023).

25.

Morales

Barksdale

Beckel-Mitchener

. A call to action to address rural mental health disparities. J Clin Transl Sci 2020; 4: 463–467.

26.

Nath

Costigan

Hsia

. Changes in demographics of patients seen at federally qualified health centers, 2005-2014. JAMA Intern Med 2016; 176: 712–714.

27.

Chen

Amaize

Barath

. Evaluating telehealth adoption and related barriers among hospitals located in rural and urban areas. J Rural Health 2021; 37: 801–811.

28.

Sun

Abraham

. Estimating dynamic treatment effects in event studies with heterogeneous treatment effects. J Econom 2021; 225: 175–199.

29.

Cole

Lee

Davoust

, et al. Comparison of visit rates before vs after telehealth expansion among patients with mental health diagnoses treated at federally qualified health centers. JAMA Network Open 2022; 5: e2242059.

30.

Uscher-Pines

Raja

Qureshi

, et al. Use of tele-mental health in conjunction with in-person care: A qualitative exploration of implementation models. PS (Wash DC) 2020; 71: 419–426.

31.

Hudnall

Zhao

, et al. Pandemic-triggered adoption of telehealth in underserved communities: descriptive study of pre- and postshutdown trends. J Med Internet Res 2022; 24: e38602.

32.

Anderson

Bouchacourt

Sussman

, et al. Telehealth adoption during the COVID-19 pandemic: A social media textual and network analysis. Digit Health 2022; 8: 20552076221090041.

33.

Demeke

. Telehealth practice among health centers during the COVID-19 pandemic—United States, July 11–17, 2020. MMWR Morb Mortal Wkly Rep 2020; 69: 1902–1905.

34.

Weber

Miller

Shroff

, et al. Recent telehealth utilization at a large federally qualified health center system: Evidence of disparities even within telehealth modalities. Telemed J E Health. 2023; 29: 1601–1612.

35.

Mehrotra

Huskamp

Souza

, et al. Rapid growth in mental health telemedicine use among rural medicare beneficiaries, wide variation across states. Health Aff 2017; 36: 909–917.

36.

Myers

. Using telehealth to remediate rural mental health and healthcare disparities. Issues Ment Health Nurs 2019; 40: 233–239.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.33 MB