Abstract
Despite concern about access to mental health (MH) services for youth, little is known about the specialty treatment infrastructure serving this population. We used national data to examine which types of MH treatment facilities (hospital- and community-based) were most likely to offer youth services and which types of communities were most likely to have this infrastructure. Larger (p<0.001) and privately owned (p<0.001) facilities were more likely to offer youth services. Rural counties, counties in which a majority of residents were nonwhite, and/or counties with a higher percentage of uninsured residents were less likely to have a community-based MH treatment facility that served youth (p<0.001).
Keywords: Children and adolescents, access to services, mental health facilities
INTRODUCTION
Although one in five youth have suffered from a mental health (MH) disorder resulting in severe impairment at some point during their lifetime,1 less than half have ever received any MH treatment.2 Concern among policymakers, researchers, and practitioners has increasingly focused on poor availability of youth MH services as an impediment to care. For example, a White House report released in January 2013 emphasized the importance of “making sure students and young adults get treatment for mental health issues.”3 Developing strategies to achieve this goal will require a thorough understanding of the current MH treatment system. Yet, limited information exists concerning a key component of the system that provides MH care for our nation's youth – specialty MH treatment facilities.
Specialty MH treatment facilities provide services across the continuum of care in inpatient, residential, and/or outpatient settings, and they constitute key components of the broader MH treatment infrastructure that serves youth.4 Hospitals that offer inpatient MH treatment for youth provide services to those with very severe MH symptoms that require a highly restrictive treatment environment.5,6 Inpatient hospitalization may be required for those who experience symptoms that pose a danger to themselves (e.g., suicidal ideation, eating disorders) or to others. Inpatient hospitalization may also be needed for an individual to be monitored during an acute episode when their mental illness is deteriorating.7
Community-based MH treatment facilities also have a unique role within the broader youth-serving infrastructure by offering a range of evidence-based practices from which youth in ongoing recovery may benefit. More specifically, these facilities typically offer psychotropic medication management, individual psychotherapeutic approaches, family therapy, and other psychosocial interventions that can benefit youth with MH disorders that vary by type and severity.8,9 The availability of services beyond medication management is especially important for children and adolescents because practice guidelines for many common youth MH disorders (e.g., anxiety and disruptive behavioral disorders) recommend psychosocial interventions as first-line treatment in lieu of or concurrent with medication.10-12 The breadth of services offered also enables these entities to serve as an important complement to other components of the community-based MH care system for youth, such as primary care settings and schools.13 In fact, although data indicate that the number of mental health visits for children and adolescents has increased in recent years in primary care settings,14 most pediatricians believe it is their responsibility to assess and refer rather than provide mental health services for disorders other than attention-deficit/hyperactivity disorder.15,16 Furthermore, evidence-based mental health programs in schools generally do not target specific clinical syndromes for youth.17
In addition to the array of services offered by specialty community-based MH treatment facilities, another important function of these entities is that they help constitute the de-facto safety-net system for low-income youth with mental health disorders. More specifically, over 90% of MH treatment facilities that offer outpatient care accept Medicaid.18 In contrast, only 4 out of 10 office-based psychiatrists accept Medicaid,19 and only 3% to 8% of patient caseloads for psychiatrists in solo or group practice are covered by Medicaid.20 Access to services for low-income populations is especially paramount for the nearly half of youth that are insured through Medicaid and other public programs (39%) or that are uninsured (8%).21
In spite of the critical role of specialty MH facilities in the child MH services infrastructure, little is known about which of these facilities are most likely to offer youth services. Data indicate that approximately three-fourths of stand-alone outpatient MH treatment facilities provide youth services, while less than half of hospital-based MH treatment facilities serve this population.8 As the U.S. mental health care system evolves, it is important to understand whether there are systematic differences in the types of facilities that offer youth services. The implementation of the Mental Health Parity and Addiction Equity Act (MHPAEA) of 2008 and the Affordable Care Act (ACA) of 2010 will continue to improve the level of coverage for MH services among those with health insurance and expand insurance coverage to previously uninsured populations. Against the backdrop of these insurance expansions, the role of state governments in the organization and delivery of MH services has declined22 and MH treatment facilities and organizations are consolidating across the United States.23-25 Consequently, understanding whether there are systematic differences in the characteristics of youth-serving facilities (e.g., ownership status and size) would provide an important foundation to assess how ongoing trends in this system may affect the accessibility of youth MH services in the coming years.
There is also limited information about the geographic availability of youth-serving specialty MH treatment facilities and the characteristics of areas in which these facilities are most likely to be located. Although research has found that nearly two-fifths of counties lack youth-serving MH facilities that provide outpatient care,26 no comparable numbers are available for youth-serving hospital-based facilities. Concerns about bed shortages for youth in several regions of the country27 have emerged in the context of a 60% decline in the total number of inpatient psychiatric beds per capita in U.S. general and psychiatric hospitals between 1990 and 2008.28,29 Information about geographic gaps in the availability of youth-serving hospital-based MH treatment facilities will provide a point of reference to assess how the consolidation within this infrastructure affects the geographic availability of youth services in the coming years.
In addition to describing the extent to which geographic gaps exist in the specialty MH treatment infrastructure for youth, it is also important to know whether certain types of communities are more likely to experience these gaps. Prior research has documented reduced access to community-based MH safety-net resources in rural (versus urban/suburban) counties, and in counties with a higher percentage of racial/ethnic minorities.26,30 Yet, to understand the variation in the availability of community-based MH facilities that serve youth, it is important to examine additional correlates that include the age distribution of the population and the health insurance status of low-income populations in a community. An examination of a more robust set of correlates would inform which types of communities are least likely to have these crucial resources for the youth population and, consequently, which communities may experience the greatest supply-side constraints in the face of ongoing insurance expansions for MH services.
Using data from a national survey, this study addresses several gaps in the literature pertaining to youth service availability in hospital- and community-based MH treatment facilities. First, we examine the facility-level correlates associated with the provision of MH services for the youth population. Second, we provide descriptive information about the geographic accessibility of youth-serving MH treatment facilities. And finally, we examine the county-level correlates associated with the geographic availability of youth-serving MH treatment facilities. The findings from these analyses are discussed in light of current trends in the behavioral health care system.
METHODS
Data
Data come from the 2008 National Survey of Mental Health Treatment Facilities (NSMHTF), which was sponsored by the Substance Abuse and Mental Health Services Agency to provide information on facilities that met the following criteria: (1) a formal establishment by law, regulation, charter, license, or agreement; (2) an established organizational structure, including a staffing structure; (3) a primary goal for all or part of the facility of improving the MH of its clientele; (4) a clientele with psychiatric, psychological, or associated social adjustment impairments; and (5) provision of MH services.31
The sampling framework for the NSMHTF was developed from fourteen sources of information, including state mental health agencies and the National Council for Community Behavioral Healthcare.31 Six types of facilities were included in the survey: psychiatric hospitals, nonfederal general hospitals with a separate psychiatric inpatient unit, residential treatment centers for children with emotional disturbance, residential treatment centers for adults, freestanding outpatient facilities, and multi-setting (non-hospital based) MH facilities. Individual and small group practices, general hospitals without a psychiatric inpatient unit, military facilities not operated by the Veterans Administration, Indian Health Service facilities, and correctional facilities were excluded from the survey. The survey was mailed to the facility director, and completed by the director or a staff member of the facility's administrative office.
Measures from the Area Health Resources File (AHRF) and the Dartmouth Atlas Project were merged with the NSMHTF.32,33 The AHRF is a county-level dataset with information about health care resources and sociodemographic characteristics. The Dartmouth Atlas Project provides information about geographic units of analysis that were created to study geographic variations in health care utilization and spending. In this study, we use the Hospital Service Area (HSA), which was developed to capture the local health care markets for hospital care.34 The nation is divided into 3,436 HSAs.
Analytic Sample
A response rate of 74% was achieved from the 13,068 facilities that were surveyed in the NSMHTF. Of the 1,680 psychiatric hospitals and non-federal general hospitals (i.e., hospital-based facilities) that responded to the survey, we excluded 11 facilities located in Washington D.C. or in a U.S. territory, 117 facilities owned by the Veterans Affairs (because of the unique population they serve and because none of these facilities offered youth services), and 19 facilities missing information on key measures (i.e., whether it served youth [n=9] and county-level sociodemographic characteristics [n=10]). These exclusions yielded for analysis a sample of 1,533 hospital-based facilities. Of the 6,365 stand-alone outpatient and multisetting facilities (i.e., community-based facilities) that responded to the survey, we excluded 29 facilities located in Washington D.C. or in a U.S. territory, 229 facilities owned by the Veterans Affairs, and 73 facilities missing information on key measures (i.e., whether it served youth [n=27] and county-level sociodemographic characteristics [n=46],). These exclusions yielded a sample of 6,034 community-based MH treatment facilities for analysis.
Measures
Facility-Level Measures
The first dependent variable was created using a question in which survey respondents were asked to indicate whether the facility provided services for those less than 18 years of age. Using the response to this question, we created a dichotomous indicator denoting facilities that provided any services for youth versus those that did not.
Facility-level independent variables included a categorical measure of ownership status (publicly owned; privately owned, for profit; and privately owned, not-for-profit), a measure of facility type as appropriate for each sample (hospital-based facilities: psychiatric hospital versus general hospital with a psychiatric inpatient unit; community-based facilities: stand-alone outpatient facility versus multi-setting facility), a dichotomous measure of religious affiliation (yes, no), and a categorical measure of patient volume at the facility. For hospital-based treatment facilities, the measure of facility volume was created from a survey item that asked for the number of clients receiving MH services in a 24-hour hospital inpatient care setting on April 30 of the survey year. For community-based facilities, this measure was created from a survey item that asked for the number of clients actively enrolled in an outpatient or partial care setting on April 30. The distribution of each measure was examined, and tertiles were used to classify facility patient volume as small, medium, or large. Among hospital-based facilities, small facilities included those with less than 15 clients; medium sized facilities included those with 15 to 39 clients; and large facilities included those with at least 40 clients in an inpatient setting on April 30. Among community-based facilities, small facilities included those with fewer than 130 clients; medium sized facilities included those with between 130 to 547 clients; and large facilities included those at least 548 clients enrolled in an outpatient or partial care setting on April 30.
County-Level Measures
Using the indicators denoting whether facilities offered youth services and the facility zip code, we aggregated these measures to the county-level to assess the number of each type of facility within the county. Based on this information, we created four dichotomous indicators for whether a county had: (1) any hospital-based MH treatment facility; (2) any hospital-based MH treatment facility that serves youth; (3) any community-based MH treatment facility; and (4) any community-based MH treatment facility that serves youth. We also created a measure of the number of youth-serving community-based MH treatment facilities in the county per 50,000 residents to examine in supplemental analyses.
County-level independent variables that assess local sociodemographic characteristics were created using the most recent year of data available in the AHRF preceding the facility survey year. Data from all U.S. counties were used to create categorical measures comprised of county-level quartiles of the percentage of residents that were: (1) living below the federal poverty level; (2) less than or equal to the age of 19; (3) enrolled in Medicaid; and (4) uninsured. Because the distribution of racial/ethnic composition is heavily skewed, a categorical measure was created to capture variation in the right tail of the distribution. The percentage of nonwhite residents (including Hispanics and racial minority groups) in the county was classified as: (1) less than 10% of residents (42.4% of all U.S. counties); (2) from 10% to less than 25% of residents (25.9% of all U.S. counties); (3) between 25% and 50% (22.1% of U.S. counties); at least 50% of residents (9.6% of U.S. counties). Finally, we created an indicator for rural counties versus urban or suburban counties (2000), using the US Census Bureau definition of non-Core-Based statistical areas.35
Analytic Strategy
Facility-level analysis
To provide contextual information about the specialty MH facilities in this sample, we describe the types of services offered in these settings -- including MH and substance abuse treatments, supportive practices, and emergency services. In addition, we present descriptive information about service availability for low-income populations and non-English speaking populations. We used χ2 tests to compare the likelihood that each type of service is offered in facilities that do and do not offer youth services.
To implement the first research objective, we examined the bivariate association between each facility-level characteristic and youth service provision using adjusted Wald tests. Next, we estimated logistic regression models to calculate the model-adjusted difference in the likelihood that youth services were offered by each type of facility. These regression models included each facility-level characteristic, county-level sociodemographic characteristics, county-level insurance coverage, and state indicators to control for unobserved differences in the state fiscal and mental health policy environment. Regressions also included indicators for facilities with missing values on a given covariate. Findings from sensitivity analyses that limited each sample to facilities with complete data on model covariates were similar in direction and significance to those presented below.
Standard errors were clustered at the county-level and marginal effects were estimated using the “margins” command in Stata statistical software.36 We present the model-adjusted percentage point difference in the likelihood that facilities offer youth services for each category of a given measure relative to its reference group, holding the other covariates at their observed values.
County-level analysis
To describe the gaps in the geographic availability of this infrastructure, we calculated the percentage of counties and HSAs that had any youth-serving MH treatment facility as well as the percentage of counties and HSAs that had any MH treatment facility. To calculate these percentages, we aggregated facility-level information to each of these geographic units of analysis using the county code in the NSMHTF and the ZIP code to HSA crosswalk file from the Dartmouth Atlas of Health Care.33
To achieve the third research objective, we first examined the bivariate association between each county-level variable and the availability of any youth-serving MH facility in the county using adjusted Wald tests. Next, we estimated logistic regression models to examine the association between each county-level characteristic and the likelihood that a county had any youth-serving MH treatment facility. These regression models included county-level measures of sociodemographic characteristics, insurance coverage, total population, and state indicators. We used the “margins” command to estimate the model-adjusted difference in the percentage of counties that have any youth-serving MH treatment facility for each covariate value relative to its reference group (i.e., adjusted percentage point difference), holding the other covariates at their observed values.
In supplemental analyses, we also examined the association between each county-level characteristic and the number of facilities per 50,000 residents in a given county using a two part model. This model was estimated with a logistic regression in the first stage, and an ordinary least squares with a log link and gamma distribution in the second stage. We estimated the marginal effects using the combined results from the two-part model, which can be interpreted as the model-adjusted difference in the number of youth serving facilities per capita for each covariate.
The Emory Institutional Review Board (IRB) determined that this study did not require IRB review because it was not classified as human subjects research.
RESULTS
Types of services offered and populations served in specialty MH Treatment facilities
Descriptive information about the types of services offered by the specialty MH facilities in this sample is presented in Table 1. More than 95% percent of hospital-based facilities and nearly three-quarters of community-based facilities offered psychotropic medication management services (Table 1). A range of psychotherapeutic and psychosocial services were also commonly provided in specialty MH treatment facilities. In community-based MH treatment facilities, for example, most facilities offered cognitive behavior therapy (87.2%), interpersonal therapy (82.8%), group therapy (79.8%), couples counseling/family counseling (68.6%), and family psychoeducational services (63.1%). Compared to facilities that did not serve youth, facilities that offered youth services were more likely to offer every type of service that was examined (p<0.001).
Table 1.
Types of services provided at specialty mental health treatment facilities
| Hospital-based Facilities+ | Community-based Facilities‡ | |||||
|---|---|---|---|---|---|---|
| Total | Offer Youth Services | Total | Offer Youth Services | |||
| (N=1,533) % |
No± (n=940) % |
Yes± (n=593) % |
(N=6,034) % |
No± (n=1,467) % |
Yes± (n=4,567) % |
|
| Mental health treatments offereda,b | ||||||
| Psychotropic medication management | 95.6 | 95.3 | 96.1 | 74.4 | 70.1 | 75.8 *** |
| Cognitive behavior therapy | 79.0 | 73.4 | 87.8 *** | 87.2 | 73.3 | 91.6 *** |
| Interpersonal therapy | 81.1 | 78.1 | 85.8 *** | 82.8 | 67.5 | 87.7 *** |
| Behavior modification | 52.4 | 45.5 | 63.3 *** | 60.5 | 44.9 | 65.4 *** |
| Couples counseling or family therapy | 54.5 | 44.1 | 71.0 *** | 68.6 | 36.6 | 78.8 *** |
| Group therapy | 94.9 | 94.3 | 96.0 | 79.8 | 73.8 | 81.7 *** |
| Supportive Practicesc,d | ||||||
| Family psychoeducation | 65.2 | 58.7 | 74.3 *** | 63.1 | 46.1 | 68.7 *** |
| Therapeutic foster care | 1.3 | 0.8 | 2.1 | 9.5 | 0.6 | 12 4 *** |
| Multisystemic therapy | 14.9 | 12.2 | 18.6 *** | 21.3 | 8.6 | 25.5 *** |
| Functional family therapy | 19.3 | 12.2 | 29.1 *** | 26.8 | 6.5 | 33.5 *** |
| Any substance abuse services offered | 49.5 | 40.4 | 63.9 *** | 53.5 | 50.4 | 54.5 ** |
| Emergency services offered | ||||||
| Emergency walk-in service available e,f | 61.4 | 55.2 | 71.2 *** | 38.2 | 29.1 | 41.1 *** |
| Offsite acute intervention team g,h | 19.7 | 17.3 | 23.5 *** | 29.5 | 24.8 | 31.0 *** |
Notes:
Hospital-based facilities include psychiatric hospitals and nonfederal general hospitals with a separate psychiatric inpatient unit
community-based facilities include stand-alone outpatient and multisetting (non-hospital based) facilities.
Statistical tests were conducted using chi-square tests.
* p<0.05
p<0.01
p<0.001
Total does not add to 1,533 hospital-based facilities due to missing values: n=2
Total does not add to 6,034 community-based facilities due to missing values: n=22
Total does not add to 1,533 hospital-based facilities due to missing values: n=403
Total does not add to 6,034 community-based facilities due to missing values: n= 1,005
Total does not add to 1,533 hospital-based facilities due to missing values: n=3
Total does not add to 6,034 community-based facilities due to missing values: n=4.
Total does not add to 1,533 hospital-based facilities due to missing values: n=17
Total does not add to 6,034 community-based facilities due to missing values: n=98.
When examining service availability for vulnerable populations in specialty MH treatment facilities (Table 2), more than nine-tenths of hospital- and community-based facilities accepted Medicaid, and approximately seven-tenths offered services at no charge to those who cannot pay. Moreover, nearly half of hospital-based (49.1%) and community-based facilities (45.9%) offered services in Spanish. Compared to facilities that did not serve youth, youth-serving facilities were more likely to accept Medicaid (p<0.01) and more likely to provide services in Spanish (p<0.01).
Table 2.
Service availability for low-income and non-English speaking populations at specialty mental health treatment facilities
| Hospital-based Facilities | Community-based Facilities | |||||
|---|---|---|---|---|---|---|
| Total | Offer Youth Services | Total | Offer Youth Services | |||
| (N=1,533) % |
No± (n=940) % |
Yes± (n=593) % |
(N=6,034) % |
No± (n=1,467) % |
Yes± (n=4,567) % |
|
| Payments accepted | ||||||
| Medicaid a,b | 92.1 | 89.7 | 96.1 *** | 91.0 | 88.8 | 91.7 *** |
| Sliding scale fee c,d | 33.6 | 30.6 | 38.2 ** | 77.5 | 68.8 | 80.3 *** |
| Services offered at no charge to those who cannot pay e,f | 69.6 | 70.2 | 68.5 | 72.4 | 75.6 | 71.4 ** |
| Private Insurance Accepted a,b | 95.2 | 93.8 | 97.4 *** | 79.6 | 65.6 | 84.1 *** |
| Services offered in other language g,h | ||||||
| Spanish | 49.1 | 46.1 | 53.9 ** | 45.9 | 39.1 | 48.1 *** |
| Other | 25.0 | 25.5 | 24.2 | 13.8 | 14.9 | 13.4 |
Notes:
statistical tests were conducted using chi-square tests.
*p<0.05
p<0.01
p<0.001
Total does not add to 1,533 hospital-based facilities due to missing values: n=21
Total does not add to 6,034 community-based facilities due to missing values: n=92
Total does not add to 1,533 hospital-based facilities due to missing values: n=79
Total does not add to 6,034 community-based facilities due to missing values: n=99
Total does not add to 1,533 hospital-based facilities due to missing values: n=55
Total does not add to 6,034 community-based facilities due to missing values: n=133
Total does not add to 1,533 hospital-based facilities due to missing values: n=9
Total does not add to 6,034 community-based facilities due to missing values: n=39.
Facility-level correlates associated with youth service provision
When examining which hospital-based facilities were most likely to offer youth services, results indicated that facility type, ownership status, and size were each correlated with youth service availability. More than half of psychiatric hospitals offered youth services, compared only to three-tenths of general hospitals with an inpatient psychiatric unit (Table 3). This difference remained sizeable and significant in the adjusted comparison (adjusted difference = −18.3 percentage points; p<0.001). The adjusted comparisons also indicated that larger (versus smaller) hospital-based facilities and privately owned not-for profit and for-profit facilities (versus publically owned facilities) were more likely to offer youth services (p<0.001). Additional tests indicated that the availability of youth services did not significantly differ between private for-profit and private not-for-profit facilities.
Table 3.
Facility-level characteristics associated with youth service availability in specialty mental health treatment facilities
| Hospital-based Facilities+ (N=1,533) | Community-based Facilities‡ (N=6,034) | |||||
|---|---|---|---|---|---|---|
| # of facilities | % that Serve Youth§ (Unadjusted) | Percentage Point Difference± (Adjusted) | # of facilities | % that Serve Youth§ (Unadjusted) | Percentage Point Difference± (Adjusted) | |
| Facility type | ||||||
| Psychiatric hospital | 611 | 52.5% (Ref) | (Ref) | - | - | - |
| General hospital with separate psychiatric inpatient unit | 922 | 29.5%*** | −18.3%*** | - | - | - |
| Freestanding outpatient facility | - | - | - | 4,992 | 76.4% (Ref) | (Ref) |
| Multisetting, non-hospital facility | - | - | - | 1,042 | 72.4%** | −3.5%* |
| Ownership type a,b | ||||||
| Public | 478 | 33.9% (Ref) | (Ref) | 1,561 | 69.0% (Ref) | (Ref) |
| Private, for profit | 279 | 48.8%*** | 14.6%*** | 315 | 83.2%*** | 23.1%*** |
| Private, not-for-profit | 770 | 38.1% | 14.1%*** | 4,151 | 77.7%*** | 14.7%*** |
| Facility Patient Volume | ||||||
| Small (< 33rd percentile) | 333 | 17.7% (Ref) | (Ref) | 1,364 | 65.6% (Ref) | (Ref) |
| Medium (33rd – 66th percentile) | 384 | 34.9%*** | 12.9%*** | 1,370 | 78.5%*** | 15.2%*** |
| Large (> 66th percentile) | 387 | 56.6%*** | 23.1%*** | 1,409 | 81.1%*** | 20.7%*** |
| Missing information on volume | 429 | 42.2%*** | 18.0%*** | 1,891 | 76.9%*** | 12.0%*** |
| Religious affiliation c,d | ||||||
| No religious affiliation | 1,269 | 39.0% (Ref) | (Ref) | 5,586 | 74.7% (Ref) | (Ref) |
| Has religious affiliation | 249 | 36.6% | −4.8% | 386 | 90.7%*** | 22.2%*** |
Notes:
p<0.05
p<0.01
p<0.001
Statistical tests were conducted using adjusted Wald tests.
Adjusted percentage point difference based on results from logistic regression models that control for all covariates in the table, county socio-demographic characteristics, county-level insurance coverage, and state indicators; N=15 hospital-based facilities and N=14 community-based facilities were excluded from each respective regression model due to perfect prediction (e.g., all facilities in a specific state offered youth services).
Total does not add to 1,533 hospital-based facilities due to missing values: n=6
Total does not add to 6,034 community-based facilities due to missing values: n=7.
Total does not add to 1,533 hospital-based facilities due to missing values: n=15
Total does not add to 6,034 community-based facilities due to missing values: n=62.
In supplemental analyses, we estimated stratified regressions for psychiatric hospitals and general hospitals with an inpatient psychiatric unit. These analyses revealed that the association between ownership status and youth service availability differed by hospital type. Although privately owned psychiatric hospitals were more likely to offer youth services than publicly owned psychiatric hospitals, private for-profit ownership (versus public ownership) was negatively associated with youth service availability (a 10.4 percentage point decrease, p<0.05) among general hospitals with a psychiatric inpatient unit.
Turning next to community-based facilities, approximately three-fourths of stand-alone outpatient facilities and multisettings facilities offered youth services. As with hospital-based facilities, privately (versus publicly) owned and larger (versus smaller) community-based facilities were more likely to offer youth services (p<0.001; Table 2). More specifically, regression results indicated that private for-profit ownership (versus public ownership) was associated with a 23.1 percentage point increase (p<0.001) and private not-for-profit ownership (versus public ownership) was associated with a 14.7 percentage point increase (p<0.001) in the likelihood that facilities serve youth. Community-based facilities with a religious affiliation were also more likely to offer youth services than those without a religious affiliation (model-adjusted difference = 22.2 percentage points; p<0.001).
Geographic availability of youth-serving MH treatment facilities
The percentage of counties (25.8%) and HSAs (28.9%) with any hospital-based MH facility was more than twice as high as the percentage of counties (12.4%) and HSAs (12.4%) that had at least one youth-serving hospital-based facility (Table 4). However, the difference between the geographic availability of all MH facilities and youth-serving MH facilities was not as pronounced for community-based facilities. Nearly two-thirds of counties (65.7%) and HSAs (65.8%) had any community-based MH facility, compared to 62.3% of counties and 61.3% of HSAs that had at least one youth-serving community-based facility.
Table 4.
Geographic availability of specialty mental health treatment facilities in the United States
| County (N=3,141) |
Hospital Service Area (N=3,436) |
|||
|---|---|---|---|---|
| Any youth serving facility, % | Any facility, % | Any youth serving facility, % | Any facility, % | |
| Hospital-based facility | 12.4 | 25.8 | 12.4 | 28.9 |
| Psychiatric hospital | 7.3 | 12.8 | 7.0 | 12.6 |
| General hospital with inpatient psychiatric unit | 6.9 | 20.1 | 6.8 | 22.4 |
| Community-based facility | 62.3 | 65.7 | 61.3 | 65.8 |
| Freestanding outpatient facility | 55.5 | 59.5 | 55.1 | 60.2 |
| Multi-setting, non-hospital facility | 15.3 | 18.7 | 15.2 | 19.0 |
Notes: Data are from the National Survey of Mental Health Treatment Facilities (2008), the Area Health Resource Files, and the Dartmouth Atlas of Health Care
County-level correlates associated with the geographic availability of youth-serving facilities
The most important correlate associated with county-level availability of a youth-serving hospital based facility was whether the county is in an urban or suburban (versus rural) area (Table 5). Just over one-fifth of urban/suburban counties had one of these resources, compared to only 1% of rural counties (p<0.001); this difference remained sizeable and significant in the adjusted comparison. Counties with a higher percentage of uninsured residents were also less likely to have youth-serving hospital-based MH treatment facility in the unadjusted and adjusted comparisons. The model-adjusted probability of having any youth-serving hospital-based facility was more than ten percentage points lower (p<0.001) in counties classified in the second, third, and fourth quartile of the percentage of uninsured residents, compared to counties classified in the lowest quartile of the percentage of uninsured residents.
Table 5.
County-level characteristics associated with the geographic availability of youth-serving specialty mental health treatment facilities
| # of counties (N=3,141) | Hospital-based Facilities | Community-based Facilities | |||
|---|---|---|---|---|---|
| % Counties with any facility§ (Unadjusted) | Percentage Point Difference± (Adjusted) | % Counties with any facility§ (Unadjusted) | Percentage Point Difference± (Adjusted) | ||
| Percentage of county residents living in poverty | |||||
| 1st quartile (<10.8%) | 780 | 16.4% (Ref) | (Ref) | 66.9% (Ref) | (Ref) |
| 2nd quartile (10.8% - 14.1%) | 783 | 13.4% | 2.7%*** | 61.7%* | −1.3% |
| 3rd quartile (14.1% - 18.2%) | 786 | 14.4% | 6.0%*** | 63.4% | 1.0% |
| 4th quartile (≥18.2%) | 792 | 5.4%*** | 4.6%*** | 57.3%*** | 1.5% |
| Percentage of county residents who are Nonwhite | |||||
| 0-10% | 1,333 | 6.8% (Ref) | (Ref) | 59.5% (Ref) | (Ref) |
| 10%-25% | 814 | 15.7%*** | 2.9%*** | 67.9%*** | 0.9% |
| 25%-50% | 693 | 17.9%*** | 4.8%*** | 63.1% | −2.5% |
| 50% and higher | 301 | 15.3%*** | 3.2% | 57.8% | −12.1%** |
| Percentage of county residents who are <=19 | |||||
| 1st quartile (<26.5%) | 783 | 10.7% (Ref) | (Ref) | 57.0% (Ref) | (Ref) |
| 2nd quartile (26.5% - 28.2%) | 789 | 13.8% | −3.1% | 65.2%*** | 4.2% |
| 3rd quartile (28.2% - 30.1%) | 784 | 14.8%* | −1.2% | 66.2%*** | 5.0%* |
| 4th quartile (≥30.1%) | 785 | 10.2% | −4.1% | 60.9% | 7.2%* |
| Percentage uninsured | |||||
| 1st quartile (<13.4%) | 793 | 22.8% (Ref) | (Ref) | 76.7% (Ref) | (Ref) |
| 2nd quartile (13.4% - 16.9%) | 782 | 9.5%*** | −10.2%*** | 66.9%*** | −4.4% |
| 3rd quartile (16.9% - 20.9%) | 781 | 9.4%*** | −11.8%*** | 55.7%*** | −15.1%*** |
| 4th quartile (≥20.9%) | 785 | 7.8%*** | −13.8%** | 49.8%*** | −16.6%*** |
| Percentage insured with Medicaid | |||||
| 1st quartile (<14.3%) | 739 | 12.5% (Ref) | (Ref) | 52.2% (Ref) | (Ref) |
| 2nd quartile (14.3% - 19.4%) | 774 | 16.3%* | 3.2%** | 66.5%*** | 14.5%*** |
| 3rd quartile (19.4% - 25.2%) | 810 | 14.0% | 1.5% | 66.3%*** | 16.3%*** |
| 4th quartile (≥25.2%) | 818 | 7.1%*** | −1.8% | 63.5%*** | 17.4%*** |
| Urban/suburban and rural location | |||||
| Location in urban/suburban county | 1,786 | 21.0% (Ref) | (Ref) | 72.2% (Ref) | (Ref) |
| Location in rural county | 1,355 | 1.0%*** | −24.0%*** | 49.2%*** | −4.6%* |
Notes:
p<0.05
p<0.01
p<0.001
Statistical tests were conducted using adjusted Wald tests.
Adjusted percentage point difference based on results from logistic regression models that control for all covariates in the table, total population in county, and state indicators
N=8 counties (in Connecticut) and N=39 counties (8 in Connecticut, 3 in Delaware, 5 in Rhode Island, 23 in Wyoming) were excluded from each regression model, respectively, due to perfect prediction (e.g., all counties in a specific state had availability of the respective type of facilities that serve youth).
When examining the distribution of youth-serving community-based MH treatment facilities across counties, demographic characteristics and insurance coverage were associated with the geographic availability of these resources (Table 5). Counties with a higher percentage of children and adolescents (p<0.05) and counties with a higher percentage of residents enrolled in Medicaid (p<0.001) were more likely to have any youth-serving community-based facility. Supplemental analyses (not shown) revealed that the percentage of residents enrolled in Medicaid was also positively correlated with the number of youth-serving community-based facilities per capita in a given county (p<0.001). On the other hand, counties with a higher percentage of racial/ethnic minorities (p<0.01), rural counties (p<0.05), and/or counties with a higher percentage of uninsured residents (p<0.05) were less likely to have any youth-serving community-based facility. For example, the model-adjusted probability of having any youth-serving community-based facility was 12.1 percentage points lower (p<0.01) in counties with more than 50% of Nonwhite residents compared to counties in which less than 10% of the residents were Nonwhite.
DISCUSSION
We found that youth-serving hospital- and community-based MH facilities commonly provide an array of psychotherapeutic and psychosocial services in addition to psychotropic medication management, and that the vast majority of these facilities offer services to low-income populations. We observed systematic differences in the types of facilities that offer youth services – with larger, privately-owned facilities being the most likely to serve this population. Lastly, we identified substantial geographic gaps in the youth-serving specialty MH treatment infrastructure and key county-level correlates associated with the availability of these facilities across communities.
When examining the types of facilities that were most likely to serve youth, results indicated that privately-owned for-profit and non-profit MH treatment facilities were more likely to serve this population than publicly owned facilities. Since the time of this survey, there have been reports of increased investment in privately owned, for-profit corporations expanding their role in the organization and delivery of specialty MH services, as investors have identified this as a sector with the potential to offer high margins of return and growth opportunities.25,37 At the same time, the role of the public sector in the organization and delivery of services has declined. State mental health agencies experienced more than $4.4 billion in budget reductions between 2008 and 2013,38 resulting in an estimated 9% reduction in the number of state psychiatric hospital beds, the closure of state psychiatric hospitals, and a reduction of community-based MH programs.22,23,38 Future studies should examine the implications of these structural changes for the accessibility of specialty MH services for youth in general, as well as for low-income youth in particular.
When examining the geographic availability of specialty MH treatment facilities, only 12% of counties have a hospital-based MH treatment facility that serves youth compared to one-fourth of counties have any hospital-based MH treatment facility. In addition, nearly all of the youth-serving hospital-based facilities are located in urban and suburban areas. These findings provide important baseline information to assess how ongoing trends in the behavioral health care system–particularly consolidation–may affect the geographic accessibility of hospital-based MH services for youth in the coming years. The consolidation of multiple hospital-based MH facilities into fewer facilities could result in a reduction of the geographic accessibility of youth services.39 Therefore, future studies should assess the extent to which geographic access to youth services in hospital-based facilities is diminished as a result of the ongoing consolidation of facilities and organizations.
Unlike youth service availability for the hospital-based infrastructure, youth service availability in community-based MH treatment facilities minimizes county-level geographic gaps. While only 48% of counties with any hospital-based MH facility have at least one hospital-based facility that serves youth, 95% of counties with any community-based MH treatment facility have at least one of facility serving youth. These findings highlight the need to leverage other health care resources to address the geographic gaps in counties that are least likely to have these specialty facilities – that is rural counties, counties in which more than half of the population is Nonwhite, counties with a higher percentage of uninsured residents, and/or counties with a lower percentage of Medicaid enrollees. The need to leverage additional resources to address these gaps is especially critical as health insurance coverage for MH services continues to expand under the MHPAEA and the ACA.40
One strategy to leverage existing resources would entail the expansion of specialty MH services for youth in existing primary care safety-net facilities such as federally qualified health centers and rural health clinics. This strategy may be particularly beneficial for communities that lack sufficient population density to support a specialty MH treatment facility. However, it is important to bear in mind that the availability of psychotherapeutic services is especially important for youth because these services are often recommended as first-line treatments for many common youth MH disorders.10-12 Thus, if the expansion of MH services in other settings (e.g., primary care safety-net facilities) is considered as a mechanism to improve geographic access to MH care, these findings highlight the need to be cognizant of the range and types of services most specialty MH facilities provide and the resources required to provide similar services in these other settings.
Several study limitations are noted. Because the data are cross-sectional, causality cannot be determined in the associations estimated in the regression analysis. Second, the age of the data poses another limitation. However, these findings provide an important baseline to examine how current trends in the behavioral health care system may affect the accessibility of youth MH services. Another data limitation is the inclusion of general hospitals in the sample only if they had a separate inpatient psychiatric unit. Thus, the data do not include general hospitals that admit psychiatric patients to general medical units41 or general hospitals with outpatient MH services if these facilities do not have a separate psychiatric inpatient unit. In addition, it is not possible to ascertain from the survey data whether a given facility provides services for youth in a specific setting. For example, although a psychiatric hospital may offer treatment in an inpatient and outpatient setting, services for youth may only be provided in one of these settings. Lastly, the available data provide information about service availability at the facility level, and do not include information about the number of individuals that received any particular service.
Notwithstanding limitations, this study provides the first available information about the differences in the types and locations of specialty MH treatment facilities that serve youth, and the extent to which gaps exist in the geographic availability of this infrastructure. These data provide an important foundation to understand the state of the specialty MH system for our nation's youth, and for future research to assess how trends in the behavioral health care system will continue to affect the accessibility of youth MH services.
Acknowledgments
Research reported in this publication was supported by the National Institute of Mental Health (K01MH095823). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
REFERENCES
- 1.Merikangas KR, He JP, Burstein M, et al. Lifetime prevalence of mental disorders in U.S. adolescents: results from the National Comorbidity Survey Replication--Adolescent Supplement (NCS-A). Journal of the American Academy of Child and Adolescent Psychiatry. 2010 Oct;49(10):980–989. doi: 10.1016/j.jaac.2010.05.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Merikangas KR, He JP, Burstein M, et al. Service utilization for lifetime mental disorders in U.S. adolescents: results of the National Comorbidity Survey-Adolescent Supplement (NCS-A). Journal of the American Academy of Child and Adolescent Psychiatry. 2011 Jan;50(1):32–45. doi: 10.1016/j.jaac.2010.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.The White House . Now Is The Time: The President's plan to protect our children and our communities by reducing gun violence. Washington D.C.: Jan 16, 2013. [Google Scholar]
- 4.Stroul B. Issue Brief -- System of care: A framework for system reform in children's mental health. Georgetown University Child Development Center, National Technical Assistance Center for Children's Mental Health; Washington D.C.: 2002. [Google Scholar]
- 5.Burns BJ, Hoagwood K, Mrazek PJ. Effective treatment for mental disorders in children and adolescents. Clinical child and family psychology review. 1999 Dec;2(4):199–254. doi: 10.1023/a:1021826216025. [DOI] [PubMed] [Google Scholar]
- 6.American Academy of Child and Adolescent Psychiatry [July 15, 2015];Inpatient Hospital Treatment of Children and Adolescents. 1989 https://www.aacap.org/aacap/policy_statements/1989/Inpatient_Hospital_Treatment_of_ Children_and_Adolescents.aspx.
- 7.Mental Health America [2015, July 6];Hospitalization. 2015 http://www.mentalhealthamerica.net/hospitalization.
- 8.Substance Abuse and Mental Health Services Administration . Data on Mental Health Treatment Facilities. BHIS Series S-69, HHS Publication No. (SMA) 14-4837.; Rockville, MD: 2014. National Mental Health Services Survey (N-MHSS): 2010. [Google Scholar]
- 9.Hoagwood K, Burns BJ, Kiser L, Ringeisen H, Schoenwald SK. Evidence-based practice in child and adolescent mental health services. Psychiatric Services. 2001 Sep;52(9):1179–1189. doi: 10.1176/appi.ps.52.9.1179. [DOI] [PubMed] [Google Scholar]
- 10.Birmaher B, Brent D, Bernet W, et al. Practice parameter for the assessment and treatment of children and adolescents with depressive disorders. Journal of the American Academy of Child and Adolescent Psychiatry. 2007 Nov;46(11):1503–1526. doi: 10.1097/chi.0b013e318145ae1c. [DOI] [PubMed] [Google Scholar]
- 11.Subcommitee on Attention-Deficit/Hyperactivity Disorder, Steering Committe on Quality Improvement and Management ADHD: Clinical practice guideline for the diagnosis, evaluation, and treatment of attention-deficit/hyperactivity disorder in children and adolescents. Pediatrics. 2011:2011–2654. doi: 10.1542/peds.2011-2654. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Connolly SD, Bernstein GA. Practice parameter for the assessment and treatment of children and adolescents with anxiety disorders. Journal of the American Academy of Child and Adolescent Psychiatry. 2007 Feb;46(2):267–283. doi: 10.1097/01.chi.0000246070.23695.06. [DOI] [PubMed] [Google Scholar]
- 13.Burns BJ, Costello EJ, Angold A, et al. Children's mental health service use across service sectors. Health Affairs. 1995;14(3):147–159. doi: 10.1377/hlthaff.14.3.147. Fall. [DOI] [PubMed] [Google Scholar]
- 14.Olfson M, Blanco C, Wang S, Laje G, Correll CU. National trends in the mental health care of children, adolescents, and adults by office-based physicians. JAMA Psychiatry. 2014 Jan;71(1):81–90. doi: 10.1001/jamapsychiatry.2013.3074. [DOI] [PubMed] [Google Scholar]
- 15.Heneghan A, Garner AS, Storfer-Isser A, Kortepeter K, Stein RE, Horwitz SM. Pediatricians' role in providing mental health care for children and adolescents: do pediatricians and child and adolescent psychiatrists agree? Journal of Developmental and Behavioral Pediatrics. 2008 Aug;29(4):262–269. doi: 10.1097/DBP.0b013e31817dbd97. [DOI] [PubMed] [Google Scholar]
- 16.Stein RE, Horwitz SM, Storfer-Isser A, Heneghan A, Olson L, Hoagwood KE. Do pediatricians think they are responsible for identification and management of child mental health problems? Results of the AAP periodic survey. Ambulatory Pediatrics. 2008 Jan-Feb;8(1):11–17. doi: 10.1016/j.ambp.2007.10.006. [DOI] [PubMed] [Google Scholar]
- 17.Rones M, Hoagwood K. School-based mental health services: a research review. Clinical Child and Family Psychology Review. 2000 Dec;3(4):223–241. doi: 10.1023/a:1026425104386. [DOI] [PubMed] [Google Scholar]
- 18.Substance Abuse and Mental Health Services Administration . Data on Mental Health Treatment Facilities. BHIS Series S-69, HHS Publication No. (SMA) 14-4837.; Rockville, MD: 2014. National Mental Health Services Survey (N-MHSS): 2010. [Google Scholar]
- 19.Bishop TF, Press MJ, Keyhani S, Pincus HA. Acceptance of insurance by psychiatrists and the implications for access to mental health care. JAMA Psychiatry. 2014 Feb;71(2):176–181. doi: 10.1001/jamapsychiatry.2013.2862. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Jacobs S, Wilk J, Chen D, Rae D, Steiner J. Datapoints: Medicaid as a payer for services provided by psychiatrists. Psychiatric Services. 2005 Nov;56(11):1356. doi: 10.1176/appi.ps.56.11.1356. [DOI] [PubMed] [Google Scholar]
- 21.The Henry J, Kaiser Family Foundation [July 14, 2015];State Health Facts: Health Insurance Coverage of Children 0-18. 2013 http://kff.org/other/state-indicator/children-0-18/.
- 22.Honberg R, Kimball A, Diehl S, Usher L, Fitzpatrick M. State Mental Health Cuts: The Continuing Crisis. National Alliance on Mental Illness; Arlington, VA: Nov, 2011. [Google Scholar]
- 23.Bogira S. Starvation diet: coping with shrinking budgets in publicly funded mental health services. Health Affairs. 2009 May-Jun;28(3):667–675. doi: 10.1377/hlthaff.28.3.667. [DOI] [PubMed] [Google Scholar]
- 24.Santilli J. [August 5, 2014];Consolidation in the Behavioral Healthcare Market. 2012 http://knowsource.com/2012/10/consolidation-in-the-behavioral-healthcare-market/.
- 25.Hill J. [August 1, 2014];Behavioral Healthcare: Five Key Factors Are Driving Sector Consolidation. 2012 http://www.triple-tree.com/blog/2012/03/20/behavioral-healthcare-five-key-factors-are-driving-sector-consolidation/.
- 26.Cummings JR, Wen H, Druss BG. Improving access to mental health services for youth in the United States. Journal of the American Medical Association. 2013 Feb 13;309(6):553–554. doi: 10.1001/jama.2013.437. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Geller JL, Biebel K. The premature demise of public child and adolescent inpatient psychiatric beds : part I: overview and current conditions. The Psychiatric Quarterly. 2006;77(3):251–271. doi: 10.1007/s11126-006-9012-0. Fall. [DOI] [PubMed] [Google Scholar]
- 28.Substance Abuse and Mental Health Services Administration . Behavioral Health, United States, 2012. HHS Publication No. (SMA) 13-4797; Rockville, MD: 2013. [PubMed] [Google Scholar]
- 29.Centers for Disease Control and Prevention [February 8, 2013];CDC Wonder: Bridged-Race Population Estimates 1990-2012 Request. 2013 http://wonder.cdc.gov/Bridged-Race-v2012.HTML.
- 30.Cummings JR, Wen H, Ko M, Druss BG. Geography and the Medicaid mental health care infrastructure: implications for health care reform. JAMA Psychiatry. 2013 Oct;70(10):1084–1090. doi: 10.1001/jamapsychiatry.2013.377. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.National Survey of Mental Health Treatment Facilities. Center for Mental Health Services, Substance Abuse and Mental Health Services Administration; Rockville, MD: 2008. [Google Scholar]
- 32.Area Health Resource Files (AHRF) US Department of Health and Human Services, Health Resources and Services Administration, Bureau of Health Professions; Rockville, MD: 2011-2012. [Google Scholar]
- 33.ZIP code to HSA to HRR crosswalk file. The Dartmouth Institute for Health Policy and Clinical Practice; Lebanon, NH: 2008. The Dartmouth Atlas of Health Care. [Google Scholar]
- 34.The Dartmouth Atlas of Health Care. [October 30, 2013];Data by Region. 2013 http://www.dartmouthatlas.org/data/region/.
- 35.United States Census Bureau [July 1, 2015];Geographic Terms and Concepts - Core Based Statistical Areas and Related Statistical Areas. 2015 http://www.census.gov/geo/reference/gtc/gtc_cbsa.html.
- 36.Stata Statistical Software [computer program]. Version 13.1. College Station, TX: 2013. [Google Scholar]
- 37.Kutscher B. Modern Healthcare. Chicago: 2013. Investors eye behavioral health: Market sees growth as more payers cover treatment. [PubMed] [Google Scholar]
- 38.National Association of State Mental Health Program Directors . The Vital Role of State Psychiatric Hospitals. National Association of State Mental Health Program Directors Medical Directors Council; Alexandria, VA: 2014. [Google Scholar]
- 39.O'Regan D. Concerns raised over consolidating state's mental health hospitals. The Legislative Gazette. Sep;30:2013. [Google Scholar]
- 40.Garfield RL, Zuvekas SH, Lave JR, Donohue JM. The impact of national health care reform on adults with severe mental disorders. American Journal of Psychiatry. 2011 May;168(5):486–494. doi: 10.1176/appi.ajp.2010.10060792. [DOI] [PubMed] [Google Scholar]
- 41.Mansbach JM, Wharff E, Austin SB, Ginnis K, Woods ER. Which psychiatric patients board on the medical service? Pediatrics. 2003 Jun;111(6 Pt 1):e693–698. doi: 10.1542/peds.111.6.e693. [DOI] [PubMed] [Google Scholar]