Abstract
Objective
We examined factors associated with telehealth utilization during COVID-19 among adult Medicaid beneficiaries with behavioral health conditions.
Data Sources and Study Setting
NC Medicaid 2019–2021 beneficiary and claims data.
Study Design
This retrospective cohort study examined and compared behavioral health service use pre-COVID-19 (03/01/2019 to 02/28/2020) and during COVID-19 (04/01/2020 to 03/31/2021). Telehealth users included those with at least one behavioral health visit via telehealth during COVID-19. Descriptive statistics were calculated for overall sample and by telehealth status. Multilevel modified Poisson generalized estimating equation examined associations between telehealth use and patient- and area-level characteristics.
Data Collection/Extraction Methods
We identified individuals ages ≥ 21–64, diagnosed with a behavioral health condition, and had at least one behavioral-health specific visit before COVID-19.
Principal Findings
Almost two-thirds of the cohort received behavioral health services during COVID-19, with half of these beneficiaries using telehealth. Non-telehealth users had steeper declines in service use from pre- to during COVID-19 compared to telehealth users. Beneficiaries identifying as Black, multiracial or other were significantly less likely to use telehealth (ARR = 0.86; 95% CI: (0.83, 0.89)); (ARR = 0.92; 95% CI: (0.87, 0.96)) compared to White beneficiaries. Those eligible for Medicaid through the blind/disabled programs and who qualified for a state-specific specialized behavioral health plan were more likely to use telehealth (17% and 20%, respectively).
Conclusions
During the pandemic, telehealth facilitated continuity of care for beneficiaries with behavioral health conditions. Future research should aim to investigate how to reduce the digital divide and ensure equitable access to telehealth.
Keywords: Behavioral health, Mental health, Telehealth, COVID-19, Medicaid
Introduction
In the United States (U.S.), 21% of the adult population has a behavioral health condition [1]. During the COVID-19 pandemic, rates of behavioral health concerns such as depression, anxiety, and substance misuse surged [2-9]. In addition, financial stressors, job insecurity, and disruption in medical care exacerbated behavioral health concerns [3, 10]. Behavioral health related-emergency department (ED) visits also increased during the pandemic [11-13], and adults with Medicaid coverage were more likely to visit the ED for a behavioral health reason [14]. As the pandemic persisted, the unmet behavioral health need also grew, with 10 to 20% of adults with a behavioral health concern reporting that they were unable to receive the necessary treatment during this time [9, 15, 16]. Moreover, individuals of color and those with disabilities reported greater barriers to accessing behavioral health care compared to Whites and those without a disability [17, 18].
As a vast majority of outpatient behavioral health clinics closed concurrent with increasing rates of anxiety and depression symptoms, healthcare systems and agencies were faced with the challenge of quickly and seamlessly shifting to delivering behavioral health care remotely (also referred to as telehealth) [2]. Studies have shown that telehealth is an effective modality for delivering treatment for most behavioral health conditions [2, 19-22]. Patients have reported high satisfaction levels with telehealth due to its convenience and reduced travel time and missed work [23-27]. Additionally, telehealth utilization has been associated with fewer appointment no-shows among patients with behavioral health conditions [28]. Furthermore, research suggests that those with a behavioral health diagnosis are three times more likely to utilize telehealth than those without a diagnosis, suggesting that it is an accepted modality among this population [29].
The COVID-19 pandemic facilitated the rapid expansion of telehealth for behavioral health services. However, whether this service modality mitigated or exacerbated existing inequities in access to behavioral health care during the pandemic remains unclear. Among adults with a behavioral health condition, lower telehealth use has been associated with factors such as older age, less education, and serious mental illness [30, 31]. Sociodemographic differences also have been found when examining the use of audio-only versus video visits, with one study within an integrated psychology care clinic finding that older patients, Black patients, and those with Medicare or Medicaid coverage were more likely to use audio-only visits compared to video visits [32]. Although audio-only and video visits have shown comparable effectiveness, concerns about the lack of visual information (e.g., nonverbal cues, collateral information), provider satisfaction, and reimbursement rate exist [33-36]. Additionally, approximately 4 million NC residents do not have home broadband services (e.g., satellite, cable, fiber optic) and many Americans do not have the necessary equipment (e.g., computer, smartphone), both of which are required for telehealth visits that include video [37, 38].
Understanding how telehealth has affected access to behavioral health care for adults in North Carolina (NC) is critical for informing policy and advocating for the groups that have been historically marginalized. NC Medicaid provides an opportunity to broadly assess the use of telehealth for behavioral health services among a population who is economically disadvantaged and identify any sociodemographic disparities that may exist. The goal of our study was to examine factors associated with telehealth use during the COVID-19 pandemic among adult Medicaid beneficiaries with a behavioral health condition. Previous research has begun to shed light on the existing telehealth disparities, and we hope that our study will not only contribute to the current literature by examining both pre- and post-pandemic behavioral health service utilization using patient-specific billing claims, but it will also be more generalizable to our state. By understanding how telehealth influenced who received behavioral health treatment during the pandemic in NC, we hope to inform policy and interventions aimed at improving equitable access and supporting the continued use of telehealth.
Methods
Data Sources
For this study, NC Medicaid 2019–2021 beneficiary, institutional and professional claims data, which were provided by the NC Department of Health and Human Services (DHHS), were analyzed. Beneficiary-level information included date of birth, sex, race, ethnicity, waiver program participation, and enrollment dates with the accompanying programs by which they qualified for NC Medicaid. Claims for services rendered to beneficiaries were included in the institutional and professional data. NC DHHS and the [BLINDED] Institutional Review Board approved this study.
Study Population
Beneficiaries were included in our study if they met the following criteria: 1) age ≥ 21 years, 2) diagnosed with a behavioral health condition, 3) had at least one behavioral-health specific visit in the year prior to COVID-19 (March 1, 2019 to February 28, 2020), and 4) continuously enrolled during the study period (March 1, 2019 to March 31, 2021). A broad scope of behavioral health conditions was included based on clinical expertise to ensure a diverse sample with a range of behavioral health needs (see Supplemental Table 1 for specific conditions). Behavioral health conditions were identified by searching all claims in the pre-COVID-19 time period for at least one inpatient or two outpatient visits on separate days with a behavioral health diagnosis in any position on the claim (Supplemental Table 1). Behavioral health-specific service use was defined as having at least one claim for a behavioral health service (see Supplemental Table 2 for services). We excluded beneficiaries from our study if they were: 1) age > 64 years old, 2) dually enrolled in Medicare at any point during the study period, 3) missing county of residence, 4) institutionalized long-term (> 100 days), and 5) used any 1915(c) waiver services or intensive behavioral health services (e.g., residential, intensive outpatient/partial hospitalization) between March 2019 and March 2021.
Study Variables
Using member data files, beneficiary demographics were identified. Age was calculated for the first day of the study period (March 1, 2019). Sex (male, female), race (combined into White, Black, Other or multiracial or unreported), and ethnicity (Hispanic, non-Hispanic, or unreported) were self-reported. Other race or multiracial included those identifying as Asian, American Indian or Alaskan Native, Native Hawaiian or Pacific Islander, or multiple races, and was created due to small frequencies of these groups. Based on NC DHHS guidelines, rurality (rural/urban) was assigned based on county of residence (earliest non-missing value in 2019–2021) [39]. Medicaid eligibility was identified as the program on the first day of the study period. Eligibility for the state-specific specialized behavioral health plan (also known as Tailored Plan) was an indicator listed in the member files. Behavioral health conditions were identified in the pre-COVID-19 year during the inclusion assessment, with categories being non-mutually exclusive. We used validated algorithms for administrative claims data to calculate the Charlson comorbidity index [40, 41].
Service use was assessed in the following time periods: pre-COVID-19 (March 1, 2019 to February 28, 2020) and during COVID-19 (April 1, 2020 to March 31, 2021). We did not include the month of March 2020 due to the rapidly changing policies and logistics that occurred during that time in the United States. Behavioral health services were reported overall and by subcategory and categorized as any use (one or more claims versus none) and number of visits among users (number of claims for that service, maximum of one per day, summed). Emergency department (ED) use during COVID-19 was assessed as a measure of healthcare utilization. To calculate total ED use, we summed the number of distinct days with an ED claim.
Telehealth use for the behavioral health-specific services was identified by the procedure modifier on those claims with CPT codes eligible for telehealth (Supplemental Table 2). Claims were considered video visits if they had a GT modifier; otherwise they were considered audio-only. We defined telehealth users as those with at least one behavioral health-specific service delivered via telehealth during COVID-19. The number of telehealth visits and number of visits eligible for telehealth (based on CPT codes) were calculated per person by summing a maximum of one per day. A sensitivity analysis defined telehealth users as those with two or more behavioral health-specific services delivered via telehealth.
Area-level characteristics were defined based on beneficiaries’ county of residence. The American Community Survey (ACS) 5-year estimates (2015–2019) was used to calculate the proportion of households in a county that did not have an internet subscription [42]. The 2018 Social Vulnerability Index (SVI) data on county-level rankings within NC was used for the themes: socioeconomic, minority status/ language, household composition, and housing type/transportation [43]. A higher value indicated higher vulnerability for all area-level variables. We mean-centered each variable and rescaled by ten, such that a one-unit change represented a 10-point change in ranking or percentage without internet subscription.
Statistical Analysis
Descriptive statistics were calculated for the overall sample and by type of user. Using χ2 and Kruskal–Wallis tests, differences by telehealth status were tested for categorical and continuous variables, respectively. Behavioral health service use was characterized by patient characteristics and type of service as well as pre- and during-COVID-19. A Wilcoxon signed-rank test was used to examine differences in number of visits by time period. Telehealth use variables were reported overall and by patient characteristics. Behavioral health service use categories were further characterized pre- and during COVID-19 stratified by telehealth status.
We examined whether the patient- and area-level characteristics described above were associated with using telehealth (users versus non-users). We first checked multicollinearity by calculating the variance inflation factor; the values were acceptable (< 6.5; all but one < 3.5)., We used a modified Poisson regression model with robust error variance to calculate relative risk of being a telehealth user [44]. County of residence was specified as a random intercept to account for geographic clustering.. A modified Poisson regression model was used to examine the association between telehealth conversion and ED use (summary count) and whether that association is moderated by key characteristics. Interaction terms between telehealth conversion and race, ethnicity, and rurality were tested in models adjusted for the patient- and area-level characteristics described above. Interaction terms with p < 0.05 were considered moderators; we reran the models stratified by these factors to isolate the association between telehealth conversion and ED use for each variable level. We used 2-tailed α = 0.05 to establish statistical significance and report 95% confidence intervals. All analyses were performed using SAS 9.4 (Cary, NC).
Results
Of the 54,996 adult beneficiaries with at least one behavioral health condition and one or more behavioral health-specific visits pre-COVID-19, the majority were female (73.3%), identified as White (59.3%) or Black (34.6%), identified as non-Hispanic (92.4%), and resided in an urban county in NC (72.8%) (Table 1). Most beneficiaries were eligible for Medicaid through the blind/disabled (51.0%) or income (47.0%) pathways. More than two-fifths of the cohort was eligible for a state-specific specialized behavioral health plan. The most common behavioral health conditions among the cohort included anxiety and obsessive-compulsive disorders (49.9%), depressive disorders (48.9%), and substance use disorders (31.8%).
Table 1.
Characteristics of the BH study cohort, adult enrollees by telehealth user status
| Non-telehealth user | |||||
|---|---|---|---|---|---|
| Variable | Overall | Telehealth user | In-person only | No service use | p-value |
| N | 54,996 | 18,796 | 14,649 | 21,551 | |
| Patient characteristics | |||||
| Age at beginning of study period, Median (Q1, Q3) | 38.0 (30.0, 50.0) | 38.0 (30.0, 48.0) | 40.0 (31.0, 52.0) | 37.0 (29.0, 49.0) | < .001 |
| Sex | < .001 | ||||
| Female | 40,300 (73.3%) | 14,104 (75.0%) | 10,347 (70.6%) | 15,849 (73.5%) | |
| Male | 14,696 (26.7%) | 4,692 (25.0%) | 4,302 (29.4%) | 5,702 (26.5%) | |
| Race | < .001 | ||||
| White | 32,624 (59.3%) | 11,456 (60.9%) | 8,496 (58.0%) | 12,672 (58.8%) | |
| Black | 19,049 (34.6%) | 6,191 (32.9%) | 5,319 (36.3%) | 7,539 (35.0%) | |
| Other or multiracial | 3,126 (5.7%) | 1,080 (5.7%) | 784 (5.4%) | 1,262 (5.9%) | |
| Unreported | 197 (0.4%) | 69 (0.4%) | 50 (0.3%) | 78 (0.4%) | |
| Hispanic/Latino ethnicity | < .001 | ||||
| Non-Hispanic | 50,830 (92.4%) | 17,361 (92.4%) | 13,464 (91.9%) | 20,005 (92.8%) | |
| Hispanic | 1,629 (3.0%) | 566 (3.0%) | 402 (2.7%) | 661 (3.1%) | |
| Unreported | 2,537 (4.6%) | 869 (4.6%) | 783 (5.3%) | 885 (4.1%) | |
| Urban/rural status | < .001 | ||||
| Rural | 14,973 (27.2%) | 5,235 (27.9%) | 3,716 (25.4%) | 6,022 (27.9%) | |
| Urban | 40,023 (72.8%) | 13,561 (72.1%) | 10,933 (74.6%) | 15,529 (72.1%) | |
| Deyo-Charlson Comorbidity Index, Median (Q1, Q3) | 1.0 (0.0, 2.0) | 1.0 (0.0, 2.0) | 1.0 (0.0, 2.0) | 0.0 (0.0, 2.0) | < .001 |
| Medicaid eligibility | < .001 | ||||
| Blind/disabled | 28,027 (51.0%) | 10,435 (55.5%) | 8,452 (57.7%) | 9,140 (42.4%) | |
| Pregnant Women & BCC | 775 (1.4%) | 195 (1.0%) | 120 (0.8%) | 460 (2.1%) | |
| Income Adult | 25,854 (47.0%) | 8,088 (43.0%) | 5,992 (40.9%) | 11,774 (54.6%) | |
| General Pediatrics | 157 (0.3%) | 56 (0.3%) | 40 (0.3%) | 61 (0.3%) | |
| Other | 183 (0.3%) | 22 (0.1%) | 45 (0.3%) | 116 (0.5%) | |
| Eligible for tailored plan | 24,307 (44.2%) | 9,639 (51.3%) | 7,315 (49.9%) | 7,353 (34.1%) | < .001 |
| Adult top diagnoses | |||||
| Anxiety and obsessive–compulsive disorder | 27,441 (49.9%) | 9,710 (51.7%) | 7,645 (52.2%) | 10,086 (46.8%) | < .001 |
| Depressive disorders | 26,897 (48.9%) | 9,903 (52.7%) | 7,085 (48.4%) | 9,909 (46.0%) | < .001 |
| Substance use disorders | 17,504 (31.8%) | 5,647 (30.0%) | 5,117 (34.9%) | 6,740 (31.3%) | < .001 |
| Opioid use disorders | 7,297 (13.3%) | 2,787 (14.8%) | 1,900 (13.0%) | 2,610 (12.1%) | < .001 |
| Bipolar and mood-related disorders | 13,333 (24.2%) | 5,712 (30.4%) | 3,742 (25.5%) | 3,879 (18.0%) | < .001 |
| PTSD and other trauma/stress-related disorders | 9,667 (17.6%) | 4,425 (23.5%) | 2,250 (15.4%) | 2,992 (13.9%) | < .001 |
| Other behavioral health condition | 17,341 (31.5%) | 7,203 (38.3%) | 4,740 (32.4%) | 5,398 (25.0%) | < .001 |
Behavioral Health Service Utilization
Overall, 60.8% of the cohort accessed behavioral health services during COVID-19. The median number of behavioral health service visits among users was higher during COVID-19 compared to before COVID-19 (median (first quartile (Q1), third quartile (Q3)) = 5 (2, 12) and 4 (2, 9), respectively) (Supplemental Table 3). When examining the median number of behavioral health service visits by behavioral health condition category and by time period, beneficiaries with opioid use disorder had a larger median number of visits during COVID-19 than prior to COVID-19 (7; Q1 = 2; Q3 = 16 and 6; Q1 = 2; Q3 = 14, respectively), whereas beneficiaries with bipolar and related disorders had a smaller median number of visits during COVID-19 than pre-COVID-19 (5; Q1 = 2; Q3 = 13 versus 6; Q1 = 2; Q3 = 12).
Of the 34.2% (n = 18,796) of beneficiaries who used telehealth for a behavioral health service during COVID-19, the median number of behavioral health telehealth visits was 4 (Q1 = 2; Q3 = 11) (Table 2). Approximately three-quarters (n = 167,035) of visits eligible for telehealth used that modality, and one-sixth (n = 27,429) of telehealth visits were audio-only. Descriptively, the proportion of individuals using telehealth did not vary widely by sex, race, ethnicity, or rurality; greater variability was found by behavioral health condition, ranging from 32.3% of those with substance use disorders using telehealth to 42.8% of those with bipolar and mood-related disorders. Those who were eligible for Medicaid because of blindness or disability had the largest proportion of telehealth users (37.2%) compared to other eligibility programs (12.0–35.7%).
Table 2.
Telehealth utilization during COVID-19 of the adult BH cohort with any BH condition
| Overall Among telehealth users |
Among telehealth users | ||||
|---|---|---|---|---|---|
| Characteristic | Total N | Telehealth users N (%) | Number of telehealth visits Median (Q1, Q3) N (%) |
Proportion of eligi ble visits that were telehealth |
Proportion of TH visits that were audio-only N (%) |
| Behavioral health services, overall | 54,996 | 18,796 (34.2%) | 4 (2–11) | 167,035 (72.2%) | 27,429 (16.4%) |
| Sex | |||||
| Female | 40,300 | 14,104 (35.0%) | 4 (2–12) | 131,746 (73.2%) | 20,352 (15.4%) |
| Male | 14,696 | 4,692 (31.9%) | 3 (1–9) | 35,289 (68.8%) | 7,077 (20.1%) |
| Race | |||||
| White | 32,624 | 11,456 (35.1%) | 4 (2–11) | 98,838 (72.0%) | 15,841 (16.0%) |
| Black | 19,049 | 6,191 (32.5%) | 4 (1–11) | 56,517 (71.9%) | 9,554 (16.9%) |
| Other or multiracial | 3,126 | 1,080 (34.5%) | 4 (2–13) | 11,144 (75.4%) | 1,976 (17.7%) |
| Unreported | 197 | 69 (35.0%) | 3 (2–6) | 536 (79.8%) | 58 (10.8%) |
| Ethnicity | |||||
| Non-Hispanic | 50,830 | 17,361 (34.2%) | 4 (2–11) | 154,408 (72.2%) | 25,584 (16.6%) |
| Hispanic | 1,629 | 566 (34.7%) | 5 (2–12) | 5,622 (73.0%) | 744 (13.2%) |
| Unreported | 2,537 | 869 (34.3%) | 3 (1–9) | 7,005 (70.4%) | 1,101 (15.7%) |
| Urban/rural county of residence | |||||
| Rural | 14,973 | 5,235 (35.0%) | 4 (2–11) | 45,411 (75.5%) | 9,572 (21.1%) |
| Urban | 40,023 | 13,561 (33.9%) | 4 (2–11) | 121,624 (71.0%) | 17,857 (14.7%) |
| Anxiety and obsessive–compulsive disorder | 27,441 | 9,710 (35.4%) | 4 (2–11) | 84,874 (71.0%) | 12,568 (14.8%) |
| Depressive disorders | 26,897 | 9,903 (36.8%) | 4 (2–12) | 92,497 (71.6%) | 14,940 (16.2%) |
| Substance use disorders | 17,504 | 5,647 (32.3%) | 4 (1–12) | 53,022 (68.7%) | 7,377 (13.9%) |
| Opioid use disorders | 7,297 | 2,787 (38.2%) | 5 (2–14) | 27,319 (66.9%) | 3,154 (11.5%) |
| Bipolar and mood-related disorders | 13,333 | 5,712 (42.8%) | 4 (2–11) | 49,676 (70.2%) | 8,434 (17.0%) |
| Eligible for tailored plan | 24,307 | 9,639 (39.7%) | 4 (1–10) | 80,422 (69.6%) | 15,405 (19.2%) |
| Program eligibility group | |||||
| Blind/disabled | 28,027 | 10,435 (37.2%) | 4 (1–10) | 88,244 (71.1%) | 17,472 (19.8%) |
| Pregnant Women & BCC | 775 | 195 (25.2%) | 5 (2–14) | 1,912 (75.9%) | 187 (9.8%) |
| Income Adult | 25,854 | 8,088 (31.3%) | 5 (2–12) | 76,318 (73.4%) | 9,628 (12.6%) |
| General Pediatrics | 157 | 56 (35.7%) | 3 (1–13) | 451 (73.2%) | 127 (28.2%) |
| Other | 183 | 22 (12.0%) | 3 (1–5) | 110 (50.5%) | 15 (13.6%) |
Among telehealth users, those residing in rural areas had a slightly higher proportion of eligible visits that were telehealth compared to those in urban areas (75.5% versus 71.0%). Additionally, a larger percentage of rural beneficiaries had telehealth visits that were audio-only than urban beneficiaries (21.1% versus 14.7%).
Behavioral Health Service Utilization by Type of Service and User
Both telehealth and non-telehealth (in-person only or no services) users received fewer services during COVID-19 compared to pre-COVID-19 across almost all behavioral health service types, although the declines in service use were much steeper among those who did not access care through telehealth (Fig. 1). For all behavioral health service types except psychiatric ED use, a higher proportion of telehealth users received these services before and during COVID-19 compared to non-telehealth users. Of note, a greater proportion of telehealth users received psychotherapy during COVID-19 (72.0%) compared to pre-COVID-19 (66.9%). In contrast, a smaller percentage of non-telehealth users received psychotherapy during COVID-19 (10.2%) than prior to COVID-19 (38.1%). A higher percentage of non-telehealth users (49.0%) had a psychiatric ED visit before COVID-19 compared with telehealth users (34.3%). Conversely, during COVID-19, a slightly higher proportion of telehealth users accessed the ED for a psychiatric reason than non-telehealth users (25.7% versus 22.5%).
Fig. 1.
Behavioral health service use by type of user before and during COVID-19
Any BH service includes all services shown as well as collaborative care and neurostimulation, which were rare
Relative Risk of Using Telehealth and Emergency Department Use
We found significant differences across sociodemographic groups when examining the likelihood of using telehealth during COVID-19 in a multivariable adjusted model. Beneficiaries identifying as Black or other or multiracial were significantly less likely to use telehealth (14% and 8%, respectively) compared to White beneficiaries (Table 3). Females were significantly more likely to convert to telehealth than males (ARR = 1.14; 95% CI: (1.11, 1.17)). Compared to those beneficiaries who were eligible for Medicaid through the income adult program, those eligible through the blind/disabled programs were 17% more likely to become a telehealth user (ARR = 1.17; 95% CI: (1.13, 1.20)). Chronic medical conditions were associated with 3% lower likelihood of converting to telehealth (ARR = 0.97; 95% CI: (0.97, 0.98)) for each 1-point increase in the Charlson comorbidity index. No significant differences were found for ethnicity, rural status, or area-level characteristics; the exception was the county ranking for those identifying as minority status and language, where each 10-point increase in ranking was associated with a 2% increased likelihood of converting to telehealth.
Table 3.
Predictors of using telehealth during COVID-19 for the BH Adults cohort
| Predictor | Relative risk of converting to telehealth during COVID-19 | |||||
|---|---|---|---|---|---|---|
| Unadjusted RR | Adjusted RR | |||||
| RR | 95% CI | p value | ARR | 95% CI | p value | |
| Age, 5-year increments | 0.99 | (0.98, 0.99) | < .001 | 0.98 | (0.97, 0.98) | < .001 |
| Race | ||||||
| Black | 0.85 | (0.81, 0.89) | < .001 | 0.86 | (0.83, 0.89) | < .001 |
| Other or multiracial | 0.93 | (0.88, 0.99) | .03 | 0.92 | (0.87, 0.96) | < .001 |
| Unreported | 0.85 | (0.71, 1.02) | .08 | 0.95 | (0.82, 1.12) | .55 |
| White | Ref | Ref | ||||
| Ethnicity | ||||||
| Hispanic | 1.00 | (0.93, 1.08) | .93 | 0.98 | (0.91, 1.05) | .54 |
| Non-Hispanic | Ref | Ref | ||||
| Unreported | 0.95 | (0.90, 1.01) | .08 | 0.95 | (0.90, 1.00) | .07 |
| Sex | ||||||
| Female | 1.15 | (1.12, 1.18) | < .001 | 1.14 | (1.11, 1.17) | < .001 |
| Male | Ref | Ref | ||||
| Urban/rural county of residence | ||||||
| Rural | 1.00 | (0.92, 1.09) | .95 | 1.07 | (0.96, 1.20) | .19 |
| Urban | Ref | Ref | ||||
| Program eligibility group | ||||||
| Blind/disabled | 1.06 | (1.03, 1.10) | < .001 | 1.17 | (1.13, 1.20) | < .001 |
| Income Adult | Ref | Ref | ||||
| Pregnant Women & BCC | 0.81 | (0.71, 0.92) | < .001 | 0.81 | (0.72, 0.92) | < .001 |
| Other | 0.70 | (0.45, 1.07) | .10 | 0.70 | (0.48, 1.00) | .05 |
| Eligible for tailored plan | ||||||
| Eligible for tailored plan | 1.23 | (1.18, 1.29) | < .001 | 1.20 | (1.15, 1.25) | < .001 |
| Not eligible | Ref | Ref | ||||
| Charlson comorbidity index | 0.98 | (0.97, 0.98) | < .001 | 0.97 | (0.97, 0.98) | < .001 |
| Depressive disorders | 1.23 | (1.19, 1.26) | < .001 | 1.22 | (1.19, 1.26) | < .001 |
| Anxiety and obsessive–compulsive disorder | 1.16 | (1.12, 1.19) | < .001 | 1.05 | (1.02, 1.08) | < .001 |
| Substance use disorders | 1.02 | (0.95, 1.10) | .55 | 1.00 | (0.94, 1.06) | .92 |
| ADHD | 1.25 | (1.19, 1.32) | < .001 | 1.12 | (1.07, 1.17) | < .001 |
| Adjustment disorder | 1.11 | (1.05, 1.19) | < .001 | 1.15 | (1.09, 1.22) | < .001 |
| Bipolar and mood-related disorders | 1.40 | (1.36, 1.44) | < .001 | 1.25 | (1.21, 1.29) | < .001 |
| Disruptive behaviors including conduct disorder | 1.36 | (1.27, 1.46) | < .001 | 1.27 | (1.18, 1.37) | < .001 |
| Disruptive mood dysregulation disorder | 1.58 | (1.34, 1.87) | < .001 | 1.35 | (1.11, 1.64) | .002 |
| Dissociative disorder | 1.35 | (1.24, 1.46) | < .001 | 1.08 | (1.00, 1.17) | .04 |
| Feeding and eating disorders | 1.35 | (1.22, 1.49) | < .001 | 1.10 | (1.00, 1.21) | .04 |
| Gender dysphoria | 1.66 | (1.38, 2.00) | < .001 | 1.39 | (1.17, 1.65) | < .001 |
| Intellectual/developmental disability | 0.87 | (0.81, 0.93) | < .001 | 0.81 | (0.75, 0.87) | < .001 |
| Neurodevelopmental conditions | 0.98 | (0.86, 1.11) | .72 | 0.96 | (0.87, 1.06) | .44 |
| Personality disorders | 1.54 | (1.47, 1.60) | < .001 | 1.11 | (1.07, 1.16) | < .001 |
| PTSD and other trauma/stress-related disorders | 1.47 | (1.43, 1.51) | < .001 | 1.28 | (1.25, 1.31) | < .001 |
| Schizophrenia and other psychotic disorders | 1.24 | (1.17, 1.32) | < .001 | 1.15 | (1.08, 1.22) | < .001 |
| Somatic symptom disorders | 1.08 | (0.92, 1.26) | .37 | 1.00 | (0.86, 1.16) | .98 |
| County ranking for Socioeconomic theme, 10-point increments | 0.98 | (0.97, 1.00) | .007 | 0.99 | (0.96, 1.02) | .37 |
| County ranking for Household Composition theme, 10-point increments | 0.98 | (0.97, 1.00) | .03 | 1.00 | (0.97, 1.02) | .67 |
| County ranking for Minority Status/Language theme, 10-point increments | 0.99 | (0.98, 1.01) | .40 | 1.02 | (1.00, 1.04) | .04 |
| County ranking for Housing Type/Transportation theme, 10-point increments | 0.99 | (0.97, 1.00) | .046 | 0.99 | (0.97, 1.01) | .36 |
| County proportion of households without an internet subscription, 10-point | 0.96 | (0.91, 1.00) | .06 | 1.01 | (0.94, 1.08) | .81 |
Adjusted models include all covariates listed in the left column
Ranking variables are the percentile ranking on that construct with higher values indicating worsening disparity. Values rescaled to a 0–10 scale
When examining relative risk of converting to telehealth by behavioral health condition, beneficiaries with most conditions were more likely to become telehealth users (5–39%) compared to those without the respective condition. However, beneficiaries with intellectual and developmental disabilities were significantly less likely to convert to telehealth (ARR = 0.81; 95% CI: (0.75, 0.87)). Beneficiaries who qualified for a state-specific specialized behavioral health plan were 20% more likely to use telehealth (ARR = 1.20; 95% CI: (1.15, 1.25)) compared to beneficiaries who were not eligible for this plan.
We also used multivariable adjusted modeling and tested interactions between telehealth status and possible moderators to examine relative risk of ED visits during COVID-19 by telehealth user status. Only the interaction term between telehealth use and race was significant (p = 0.005); ethnicity (p = 0.433) and rurality (p = 0.517) interaction terms were not significant (Supplemental Table 4). Thus, the final model was stratified by race. Among beneficiaries identifying as Black, telehealth users were 12% less likely to utilize the ED (ARR = 0.88; 95% CI: (0.81, 0.96)) compared to non-telehealth users (Table 4). No significant differences were found in ED use between telehealth and non-telehealth users among beneficiaries identifying as White (ARR = 1.02; 95% CI: (0.96, 1.08)), other race, or multiracial (ARR = 0.99; 95% CI: (0.86, 1.14)).
Table 4.
Relative risk of ED visits during COVID-19 by telehealth status for the BH Adults cohort
| Comparison | Relative risk of ED use during COVID-19 | |||||
|---|---|---|---|---|---|---|
| Unadjusted RR | Adjusted RR | |||||
| RR | 95% CI | p value | ARR | 95% CI | p value | |
| Separate models by race | ||||||
| White: telehealth user vs non-telehealth user | 1.13 | (1.06, 1.20) | < .001 | 1.02 | (0.96, 1.08) | .46 |
| Black: telehealth user vs non-telehealth user | 0.93 | (0.84, 1.03) | .16 | 0.88 | (0.81, 0.96) | .006 |
| Other or multiracial: telehealth user vs non-telehealth user | 1.06 | (0.86, 1.30) | .59 | 0.99 | (0.86, 1.14) | .88 |
Each line is a separate model for that subgroup. COVID-related ED visits are excluded from the counts
Discussion
In our study, we found that about 60% of adult beneficiaries with at least one behavioral health condition and one or more behavioral health-specific visits during pre-COVID-19 received behavioral health treatment during the pandemic. Furthermore, among adult beneficiaries receiving behavioral health during COVID-19, they had a greater median number of behavioral health visits, particularly psychotherapy, compared to pre-COVID-19. This finding suggests a majority of beneficiaries were able to maintain continuity of care during COVID-19. Between August 2020 through March 2021, biweekly analysis of data from the U.S. Census Bureau, Household Pulse Survey found that about 70–80% of adults in NC with anxiety and depressive symptoms and who needed care during the pandemic, received it [45]. Of note, this percentage declined to nearly 60% during the end of March 2021, consistent with our finding. Possible explanations for our lower treatment rate during COVID-19 compared to the findings from Household Pulse Survey may be due to our analysis of a different group of adults (Medicaid beneficiaries with a mental health diagnosis), individuals no longer needing care during this time, or those experiencing more pandemic-related stressors. Nonetheless, our finding adds to the evidence that a majority of beneficiaries were able to maintain access to care during COVID-19.
Approximately one-half of those who received behavioral health care during the pandemic utilized telehealth, suggesting that this service modality helped support access to behavioral health treatment at a time of immense uncertainty and concerns about virus transmission. A prior study using a large electronic health record database found that 39% of outpatient behavioral health treatment visits were delivered through telehealth during the first six months of the pandemic, and this rate remained steady during the same six months in 2021 [16]. Non-telehealth users had steeper declines in behavioral health service utilization, including psychotherapy, evaluation and management services (e.g., medication management), and psychiatric ED visits during COVID-19 compared to telehealth users. However, almost half of non-telehealth users utilized the ED for a behavioral health reason visit prior to COVID-19 compared to a third of telehealth users. It is unclear what may have happened for these individuals, who were experiencing acute behavioral health needs prior to the pandemic, and then stopped receiving treatment. Future research should aim to better understand factors affecting access to and use of behavioral health care, especially for those with tenuous connections to healthcare.
No significant differences were found in the likelihood of becoming a telehealth user by rural status. A previous study found that a greater proportion of individuals residing in rural areas used telehealth for behavioral health concerns than those in urban areas (55% versus 35%), which may be due to a shortage of providers in those areas.16 We did find that a larger percentage of rural beneficiaries had telehealth visits that were audio-only compared to urban beneficiaries; however, some visits coded as audio-only may have been non-telehealth visits due to an issue with the modifier used. Concerns with the accuracy of audio-only visit data has made it challenging to understand and compare the utilization and quality between audio-only and video visits [46]. Furthermore, there are provider concerns about the quality and reimbursement rate of this audio-only visits [33-36].
Future research should aim to better understand how telehealth is being delivered to individuals residing in rural areas. Policymakers could consider increasing access to video visits through investments in broadband infrastructure, digital technology, and digital literacy, particularly in rural areas, as well as increasing the reimbursement rate for audio-only visits as it is currently only 80% of the rate for video visits [36]. Despite federally qualified health centers (FQHC) in California attempting to expand video visits as the pandemic persisted, they continued to depend on audio-only visits in 2022, with two in five behavioral health visits being delivered through audio-only [47]. This finding suggest that barriers remain when accessing the necessary technology required for video visits [47].
When examining the likelihood of using telehealth, results from our study indicated that beneficiaries identifying as Black, multiracial, or other were significantly less likely to use telehealth compared to White beneficiaries. Numerous studies have shown that Black and Hispanic individuals have lower behavioral health treatment utilization rates and often receive less quantity and quality of treatment compared to their White and non-Hispanic counterparts, compounding inequities in access to behavioral health care via telehealth [48-53]. In addition, racial and ethnic disparities exist in access to broadband internet and smart devices [46, 54]. Even prior to the pandemic, a greater proportion of Black and Hispanic Americans did not have access to home internet compared to White and Non-Hispanic Americans [46]. This trend has continued throughout the pandemic, with a 2021 survey finding that Black and Hispanic adults in the U.S. are less likely than White and Non-Hispanic adults to have access to internet at home or own a computer [54]. Additionally, results from the U.S. Census Bureau, Household Pulse Survey found that a higher proportion of individuals of color had audio-only visits between July and October 2021 compared to White individuals [55]. Our study also found that among Black beneficiaries, telehealth users were less likely to utilize the ED compared to non-telehealth users. This suggests that telehealth may help mitigate the use of the ED for Black beneficiaries.
When comparing beneficiaries identifying as Hispanic and non-Hispanic, we found no significant differences in telehealth use. However, when examining the county ranking for those identifying as minority status and language, results indicated that a 10-point increase in ranking was associated with a 2% increased likelihood of converting to telehealth. Additional analyses conducted by our analytic team found that adult beneficiaries with a behavioral health diagnosis and no service use pre-COVID were more likely to identify as White, Hispanic, and residing in a rural area then compared to services users in our cohort [56]. Given that the number of beneficiaries identifying as Hispanic was quite small in our study (3.0%), it may be that our sample excluded Hispanic beneficiaries who never connected to care as inclusion criteria required receipt of at least one behavioral health service and a behavioral health diagnosis prior to pandemic. It will be critical that policymakers and community leaders identify ways to enable access to tools necessary for telehealth, including broadband internet and smart devices, to ensure that all beneficiaries have the option to use telehealth given that this modality will likely continue to be offered, and is associated with a decreased likelihood of utilizing the ED in select populations.
Overall, our study found that individuals with greater physical and behavioral health needs were more likely to become telehealth users. This modality has several benefits including reducing the need to travel to receive services, which can be a barrier for individuals with more needs. However, in contrast, we found that beneficiaries with intellectual and developmental disabilities were significantly less likely to convert to telehealth. It may be that the services most beneficial to beneficiaries with intellectual and developmental disabilities are better suited to be delivered in person rather than via telehealth or these beneficiaries encountered issues accessing the technology required for telehealth.
As part of the Public Health Emergency (PHE) that was put in place in response to the pandemic, many states expanded Medicaid coverage for behavioral health telehealth visits. States determined which telehealth services would be reimbursed and many have maintained these policies as the pandemic has continued, especially flexibilities around behavioral health care. The Centers for Medicare and Medicaid Services (CMS) have approved continued coverage of telehealth flexibilities for behavioral health visits until the end of 2024 [57-59]. It is unclear at this time if these flexibilities will be become permanent or additional stipulations will be implemented in the future such as requiring a behavioral health visit in person every six months, which were proposed as part of the 2022 Consolidated Appropriations Act but has been delayed in the 2023 Consolidated Appropriations Act [57, 60].
Continued research examining the specific factors influencing access to and receipt of behavioral health services via telehealth or in person, particularly among historically marginalized groups and those with more complex medical and behavioral needs, is needed to reduce disparities, ensure continuity of care, and inform policy regarding telehealth care in the future. If beneficiaries will be required to attend an in-person visit for behavioral health care in the future, it will be important to understand how this might influence one’s ability to stay connected to care, especially if there are barriers to getting into the office (e.g., transportation, child care). This information may inform the implementation of a state-specific specialized behavioral health plan for those beneficiaries with more significant behavioral health needs, which is planned to launch in NC in the future [61].
Our study has several limitations. Since we were unable to access data about clinical need, our understanding of non-telehealth users is limited as this sample included both beneficiaries who only utilized in-person behavioral health treatment as well as those who did not access any behavioral health care. There may have been differences between these groups, such as symptom severity and social determinants of health that may have influenced one’s ability to seek out and receive treatment, which we were unable to assess. Additionally, for the beneficiaries who did not receive any services during COVID-19, we do not have information on whether this was clinically indicated or they needed treatment but experienced barriers to accessing it. The primary goal of the study was to examine overall telehealth utilization for behavioral health services, and as such, we did not examine or compare outcomes associated with audio-only versus video visits. Furthermore, the modifier used to identify telehealth also could be used for other COVID-related changes (e.g., changes in frequency requirements). As a result, some visits coded as audio-only may have been non-telehealth visits, and therefore our analysis may be an overestimate of telehealth use, especially audio-only. We examined the possible extent of this and found it to be minimal (at most, 7% of telehealth users were misclassified), but future research should explore this further. Unfortunately, we were unable to investigate factors that may influence treatment access and engagement, such as person-level data about access to broadband internet, smart devices, and behavioral health providers as well as digital literacy and personal preferences.
Telehealth helped adult Medicaid beneficiaries with behavioral health conditions maintain access to care during the COVID-19 pandemic. Racial differences were found when examining the likelihood of telehealth use, with Black and other or multiracial beneficiaries being less likely to become telehealth users, which is consistent with prior research. Beneficiaries with more complex medical and behavioral health needs were more likely to use telehealth. Future studies should aim to better understand the factors contributing to these differences and identify strategies and interventions to ensure equitable access to telehealth.
Supplementary Material
Funding
This project was funded by the Kate B. Reynolds Charitable Trust (#2834758). Dr. Rebecca Whitaker, Ms. Yolande Pokam Tchuisseu, and Mrs. Repka are currently funded by North Carolina Medicaid. Drs. Maslow and French receive grant support from several sources: Substance Abuse Mental Health Services Administration, North Carolina Medicaid, North Carolina Department of Health and Human Services, and Health Resources and Services Administration. Dr. Maslow also receives support from Healthy Blue and received research funding from Pfizer and the National Institutes of Health, and the Arc of North Carolina. Dr. Cholera was supported by K12-HD105253.
Footnotes
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s40615-023-01730-2.
Ethics Approval This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of Duke University on 08/06/2021 (# Pro00108541). The study was determined to be exempt.
Conflicts of Interest/Competing Interests The authors have no relevant financial or non-financial interests to disclose.
Data Availability
The data are only available to those who have received approval from NC Medicaid to access it.
References
- 1.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/sites/default/files/reports/rpt35325/NSDUHFFRPDFWHTMLFiles2020/2020NSDUHFFR1PDFW102121.pdf. Accessed 14 Nov 2022. [Google Scholar]
- 2.Czeisler MÉ, Howard ME, Rajaratnam SMW. Mental Health During the COVID-19 Pandemic: Challenges, Populations at Risk, Implications, and Opportunities. Am J Health Promot. 2021;35(2):301–11. 10.1177/0890117120983982b. [DOI] [PubMed] [Google Scholar]
- 3.Coley RL, Carey N, Baum CF, Hawkins SS. Covid-19–related stressors and mental health disorders among us adults. Public Health Rep. 2022;137(6):1217–26. 10.1177/00333549221120451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Daly M, Sutin AR, Robinson E. Depression reported by US adults in 2017–2018 and March and April 2020. J Affect Disord. 2021;278:131–5. 10.1016/j.jad.2020.09.065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ettman CK, Abdalla SM, Cohen GH, Sampson L, Vivier PM, Galea S. Prevalence of depression symptoms in us adults defore and during the covid-19 pandemic. JAMA Netw Open. 2020;3(9):e2019686. 10.1001/jamanetworkopen.2020.19686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Jewell JS, Farewell CV, Welton-Mitchell C, Lee-Winn A, Walls J, Leiferman JA. Mental Health During the COVID-19 Pandemic in the United States: Online Survey. JMIR Form Res. 2020;4(10):e22043. 10.2196/22043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Lee H, Singh GK. Monthly trends in self-reported health status and depression by race/ethnicity and socioeconomic status during the COVID-19 Pandemic, United States, April 2020 - May 2021. Ann Epidemiol. 2021;63:52–62. 10.1016/j.annepidem.2021.07.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Jia H, Guerin RJ, Barile JP, et al. National and state trends in anxiety and depression severity scores among adults during the COVID-19 pandemic — United States, 2020–2021. Morb Mortal Wkly Rep. 2021;70:1427–32. 10.15585/mmwr.mm7040e3externalicon. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Vahratian A, Blumberg SJ, Terlizzi EP, Schiller JS. 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. Morbid Mortal Wkly Rep. 2021;70(13):490–4. 10.15585/mmwr.mm7013e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wilson JM, Lee J, Fitzgerald HN, Oosterhoff B, Sevi B, Shook NJ. Job insecurity and financial concern during the covid-19 pandemic are associated with worse mental health. J Occup Environ Med. 2020;62(9):686–91. 10.1097/JOM.0000000000001962. [DOI] [PubMed] [Google Scholar]
- 11.Holland KM, Jones C, Vivolo-Kantor AM, et al. trends in us emergency department visits for mental health, overdose, and violence outcomes before and during the covid-19 pandemic. JAMA Psychiat. 2021;78(4):372–9. 10.1001/jamapsychiatry.2020.4402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Anderson KN, Radhakrishnan L, Lane RI, et al. Changes and inequities in adult mental health-related emergency department visits during the covid-19 pandemic in the us. JAMA Psychiat. 2022;79(5):475–85. 10.1001/jamapsychiatry.2022.0164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Stroever S, Brett C, Michael K, Petrini J. Emergency department utilization for mental health conditions before and after the COVID-19 outbreak. Am J Emerg Med. 2021;47:164–8. 10.1016/j.ajem.2021.03.084. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Penner F, Rajesh A, Kinney KL, et al. Racial and demographic disparities in emergency department utilization for mental health concerns before and during the COVID-19 pandemic. Psychiatry Res. 2022;310:114442. 10.1016/j.psychres.2022.114442. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Nagata JM, Ganson KT, Bonin SL, et al. Prevalence and sociodemographic correlates of unmet need for mental health counseling among adults during the covid-19 pandemic. Psychiatr Serv. 2022;73(2):206–9. 10.1176/appi.ps.202100111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Lo J, Rae M, Amin K, Cox C, Panchal N, Miller BF. Telehealth has played an outsized role meeting mental health needs during the COVID-19 pandemic, KFF, March 15, 2022. https://www.kff.org/coronavirus-covid-19/issue-brief/telehealth-has-played-an-outsized-role-meeting-mental-health-needs-during-the-covid-19-pandemic/. Accessed 14 Nov 2022. [Google Scholar]
- 17.Thomeer MB, Moody MD, Yahirun J. Racial and Ethnic Disparities in Mental Health and Mental Health Care During The COVID-19 Pandemic [published online ahead of print, 2022 Mar 22]. J Racial Ethn Health Disparities. 2022;1–16. 10.1007/s40615-022-01284-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Czeisler MÉ, Board A, Thierry JM, et al. Mental health and substance use among adults with disabilities during the Covid-19 pandemic - United States, February-March 2021. MMWR Morb Mortal Wkly Rep. 2021;70(34):1142–9. 10.15585/mmwr.mm7034a3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Luna P, Lee M, Vergara Greeno R, et al. Telehealth care before and during COVID-19: trends and quality in a large health system. JAMIA Open. 2022;5(4):ooac079. 10.1093/jamiaopen/ooac079. (Published 2022 Oct 4). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hailey D, Roine R, Ohinmaa A. The effectiveness of telemental health applications: a review. Can J Psychiatry. 2008;53(11):769–78. 10.1177/070674370805301109. [DOI] [PubMed] [Google Scholar]
- 21.Langarizadeh M, Tabatabaei MS, Tavakol K, Naghipour M, Rostami A, Moghbeli F. Telemental health care, an effective alternative to conventional mental care: a systematic review. Acta Inform Med. 2017;25(4):240–6. 10.5455/aim.2017.25.240-246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hilty DM, Ferrer DC, Parish MB, Johnston B, Callahan EJ, Yellowlees PM. The effectiveness of telemental health: a 2013 review. Telemed J E Health. 2013;19(6):444–54. 10.1089/tmj.2013.0075. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bulkes NZ, Davis K, Kay B, Riemann BC. Comparing efficacy of telehealth to in-person mental health care in intensive-treatment-seeking adults. J Psychiatr Res. 2022;145:347–52. 10.1016/j.jpsychires.2021.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Mann DM, Chen J, Chunara R, Testa PA, Nov O. COVID-19 transforms health care through telemedicine: evidence from the field. J Am Med Inform Assoc. 2020;27(7):1132–5. 10.1093/jamia/ocaa072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Powell RE, Henstenburg JM, Cooper G, Hollander JE, Rising KL. Patient perceptions of telehealth primary care video visits. Ann Fam Med. 2017;15(3):225–9. 10.1370/afm.2095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Hommel KA, Hente E, Herzer M, Ingerski LM, Denson LA. Telehealth behavioral treatment for medication nonadherence: a pilot and feasibility study. Eur J Gastroenterol Hepatol. 2013;25(4):469–73. 10.1097/MEG.0b013e32835c2a1b. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ballester JMS, Scott MF, Owei L, Neylan C, Hanson CW, Morris JB. Patient preference for time-saving telehealth postoperative visits after routine surgery in an urban setting. Surgery. 2018;163(4):672–9. 10.1016/.surg.2017.08.015. [DOI] [PubMed] [Google Scholar]
- 28.Eyllon M, Barnes JB, Daukas K, Fair M, Nordberg SS. The impact of the covid-19-related transition to telehealth on visit adherence in mental health care: an interrupted time series study. Adm Policy Ment Health. 2022;49(3):453–62. 10.1007/s10488-021-01175-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Andersen JA, Rowland B, Gloster E, McElfish PA. Telehealth utilization during covid-19 among people with diagnosed mental health conditions. Telemed J E Health. 2022;28(5):743–6. 10.1089/tmj.2021.0356. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Sizer MA, Bhatta D, Acharya B, Paudel KP. Determinants of telehealth service use among mental health patients: a case of rural louisiana. Int J Environ Res Public Health. 2022;19(11):6930. 10.3390/ijerph19116930. (Published 2022 Jun 6). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Zhu JM, Myers R, McConnell KJ, Levander X, Lin SC. Trends in outpatient mental health services use before and during the covid-19 pandemic. Health Affairs (Project Hope). 2022;41(4):573–80. 10.1377/hlthaff.2021.01297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Tobin ET, Hadwiger A, DiChiara A, Entz A, Miller-Matero LR. Demographic predictors of telehealth use for integrated psychological services in primary care during the covid-19 pandemic. J Racial Ethn Health Disparities. 2022. 10.1007/s40615-022-01334-2. (Published online ahead of print, 2022 Jul 6). [DOI] [PubMed] [Google Scholar]
- 33.Chen PV, Helm A, Caloudas SG, et al. Evidence of phone vs video-conferencing for mental health treatments: a review of the literature. Curr Psychiatry Rep. 2022;24(10):529–39. 10.1007/s11920-022-01359-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Chen PV, Helm A, Fletcher T, et al. Seeing the value of video: a qualitative study on patient preference for using video in a veteran affairs telemental health program evaluation. Telemed Rep. 2021;2(1):156–62. 10.1089/tmr.2021.0005. (Published 2021 May 31). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Fletcher TL, Amspoker AB, Wassef M, et al. Increasing access to care for trauma-exposed rural veterans: A mixed methods outcome evaluation of a web-based skills training program with telehealth-delivered coaching. J Rural Health. 2022;38(4):740–7. 10.1111/jrh.12628. [DOI] [PubMed] [Google Scholar]
- 36.North Carolina Medicaid. SPECIAL BULLETIN COVID-19 #59: Telehealth Clinical Policy Modifications - Outpatient Behavioral Health Services. Published on April 23, 2020. https://medicaid.ncdhhs.gov/blog/2020/04/23/special-bulletin-covid-19-59-telehealth-clinical-policy-modifications-outpatient. Accessed on January 20, 2023.
- 37.Carolina Public Press. No internet, no telehealth: Rural North Carolina residents struggle to connect with doctors virtually. NC Health News, October 22, 2022. https://www.northcarolinahealthnews.org/2022/10/22/no-internet-no-telehealth/. Accessed on January 20, 2023. [Google Scholar]
- 38.Amin K, Rae M, Ramirez G, Cox C. How might internet connectivity affect health care access? Peterson-KFF Health System Tracker, December 14, 2020. https://www.healthsystemtracker.org/chart-collection/how-might-internet-connectivity-affect-health-care-access/#Share%20of%20population%20with%20no%20internet%20access%20at%20home%202019. Accessed 16 Dec 2022. [Google Scholar]
- 39.North Carolina Department of Health and Human Services. North Carolina Metropolitan (Urban) and Non-Metropolitan (Rural) Counties. Retrieved from https://files.nc.gov/ncdhhs/documents/files/%23J223%20AttachmentTwo_Urban%20Rural%20Counties.xlsx.
- 40.Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130–9. 10.1097/01.mlr.0000182534.19832.83. [DOI] [PubMed] [Google Scholar]
- 41.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–83. 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 42.Manson S, Schroeder J, Van Riper D, Kugler T, & Ruggles S. IPUMS National Historical Geographic Information System: Version 16.0. Minneapolis, MN: IPUMS. 2021; 10.18128/D050.V16.0. [DOI] [Google Scholar]
- 43.Centers for Disease Control and Prevention/ Agency for Toxic Substances and Disease Registry/ Geospatial Research, Analysis, and Services Program. CDC/ATSDR Social Vulnerability Index 2018 Database US.
- 44.Zou G. A Modified Poisson Regression Approach to Prospective Studies with Binary Data. Am J Epidemiol. 2004;159(7):702–6. 10.1093/aje/kwh090. [DOI] [PubMed] [Google Scholar]
- 45.Kaiser Family Foundation. Unmet Need for Counseling or Therapy Among Adults Reporting Symptoms of Anxiety and/or Depressive Disorder During the COVID-19 Pandemic. https://www.kff.org/other/state-indicator/unmet-need-for-counseling-or-therapy-among-adults-reporting-symptoms-of-anxiety-and-or-depressive-disorder-during-the-covid-19-pandemic/. Accessed on January 22, 2023.
- 46.Hailu R, Uscher-Pines L, Ganguli I, Huskamp HA, Mehrotra A. Audio-only telemedicine visits: flaws in the underlying data make it hard to assess their use and impact". Health Aff Forefront. 2022. 10.1377/forefront.20220713.689213. [DOI] [Google Scholar]
- 47.Uscher-Pines L, McCullough CM, Sousa JL, et al. Changes in in-person, audio-only, and video visits in california’s federally qualified health centers, 2019–2022. JAMA. 2023;329(14):1219–21. 10.1001/jama.2023.1307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Lasser KE, Himmelstein DU, Woolhandler SJ, McCormick D, Bor DH. Do minorities in the united states receive fewer mental health services than whites? Int J Health Serv. 2002;32(3):567–78. 10.2190/uexw-rarl-u46v-fu4p. [DOI] [PubMed] [Google Scholar]
- 49.Cook BL, Zuvekas SH, Carson N, Wayne GF, Vesper A, McGuire TG. Assessing racial/ethnic disparities in treatment across episodes of mental health care. Health Serv Res. 2014;49(1):206–29. 10.1111/1475-6773.12095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Garland AF, Lau AS, Yeh M, McCabe KM, Hough RL, Landsverk JA. Racial and ethnic differences in utilization of mental health services among high-risk youths. Am J Psychiatry. 2005;162(7):1336–43. 10.1176/appi.ajp.162.7.1336. [DOI] [PubMed] [Google Scholar]
- 51.Kataoka S, Zhang L, Wells K. Unmet need for mental health care among U.S. children: variation by ethnicity and insurance status. Am J Psychiatry. 2002;159(9):1548–55. 10.1176/appi.ajp.159.9.1548. [DOI] [PubMed] [Google Scholar]
- 52.Alegria M, Vallas M, Pumariega AJ. Racial and ethnic disparities in pediatric mental health. Child Adolesc Psychiatr Clin N Am. 2010;19(4):759–74. 10.1016/jxhc.2010.07.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Atske S, Perrin A. Home broadband adoption, computer ownership vary by race, ethnicity in the US. July 16, 2021. https://www.pewresearch.org/fact-tank/2021/07/16/home-broadband-adoption-computer-ownership-vary-by-race-ethnicity-in-the-u-s/. Accessed 16 Dec 2022. [Google Scholar]
- 54.Karimi M, Lee EC, Couture SJ, Gonzales AB, Grigorescu V, Smith SR, et al. National trends in telehealth use in 2021: disparities in utilization and audio vs. Video services. (research report no. Hp-2022–04). Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. February 2022. [Google Scholar]
- 55.Hough RL, Hazen AL, Soriano FI, Wood P, McCabe K, Yeh M. Mental health services for Latino adolescents with psychiatric disorders. Psychiatr Serv. 2002;53(12):1556–62. 10.1176/appi.ps.53.12.1556. [DOI] [PubMed] [Google Scholar]
- 56.Swietek K, Jones KA, Prvu Bettger J, French A, Maslow G, Norman KS, Lake AD, Carvalho M, Cholera R, Freed S, Tchuissum YP, Repka S, & Whitaker RG. What Explains Inequalities in Telehealth Utilization Among North Carolina Medicaid Beneficiaries? (under review). [DOI] [PubMed] [Google Scholar]
- 57.Consolidated Appropriations Act, 2023, H.R.2671, 117th Congress, 2021–2022. https://www.congress.gov/bill/117th-congress/house-bill/2617. [Google Scholar]
- 58.Association of American Medical Colleges. Biden signs FY 23 omnibus with new GME slots, other patient care policies. https://www.aamc.org/advocacy-policy/washington-highlights/biden-signs-fy-23-omnibus-new-gme-slots-other-patient-care-policies. Accessed 6 Jan 2023.
- 59.North Carolina Medicaid. Behavioral Health Clinical Coverage Policies. Outpatient behavioral health services provided by direct-enrolled providers. https://medicaid.ncdhhs.gov/media/10885/open. Accessed 22 Jan 2023.
- 60.Consolidated Appropriations Act, 2022, H.R.2471, 117th Congress, 2021–2022. https://www.congress.gov/bill/117th-congress/house-bill/2471. [Google Scholar]
- 61.NC Medicaid Ombudsman. Tailored plans. https://ncmedicaidombudsman.org/tailored-plans/. Accessed 5 Apr 2023.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data are only available to those who have received approval from NC Medicaid to access it.