Investing in Certified Community Behavioral Health Centers to Fulfill Their Promise

Certified community behavioral health centers (CCBHCs) were established under section 223 of the Protecting Access to Medicare Act (PAMA) of 2014. CCBHCs had the goal of expanding access to care for people with behavioral health needs “regardless of ability to pay and place of residence” (1, 2). PAMA, as well as subsequent legislation authorizing additional grants to CCBHCs, emphasized expanding services for serious mental illness, substance use disorders, co-occurring illnesses, and serious emotional disorders in children. The federal government has made CCBHCs a central component of the strategy to expand access to and quality of behavioral health services. It has done so by spending

CCBHCs had the goal of expanding access to care for people with behavioral health needs “regardless of ability to pay and place of residence”

$1.7 billion in grants to establish and expand CCBHCs since 2016, in addition to funding many CCBHCs through Medicaid (3). The recent Bipartisan Safer Communities Act (BPSCA) of 2022 directed an estimated $8.5 billion over 10 years to expand the section 223 demonstration program into a national initiative (4).

This evaluation is focused on the initial demonstration sites and offers a useful profile of the activities and indicators of quality of care offered by the original cohort of CCBHCs.

To date, the Office of the Assistant Secretary for Planning and Evaluation at the U.S. Department of Health and Human Services has systematically evaluated the program (5). This evaluation is focused on the initial demonstration sites and offers a useful profile of the activities and indicators of quality of care offered by the original cohort of CCBHCs. However, it does not gauge performance relative to any benchmarks such as community mental health centers or other community-based mental health programs (6). Furthermore, it does not indicate whether CCBHCs are meeting other performance requirements, particularly whether they serve counties that have high need for care and low economic resources.

In this study, we considered the geographic distribution of CCBHCs relative to indicators of rates of mental illness and the economic resources of communities across the United States. We examined this distribution in light of the rules governing the various approaches to funding CCBHCs and the programmaticrequirements for certification of CCBHCs. We conclude by proposing some modifications to policies for establishing CCBHCs to meet the goal of expanding access to care for people with behavioral health needs “regardless of ability to pay and place of residence.”

CCBHCs were established in 2014 under section 223 of PAMA. The language of the legislation and subsequent funding announcements suggest four central aims of CCBHCs: to improve the quality of specialty behavioral health care; expand access to timely care regardless of ability to pay or location; provide a one-stop shop for specialty behavioral health services, with an emphasis on crisis services; and improve care coordination for people with serious mental illnesses and substance use disorders.

The language of the legislation and subsequent funding announcements suggest four central aims of CCBHCs: to improve the quality of specialty behavioral health care; expand access to timely care regardless of ability to pay or location; provide a one-stop shop for specialty behavioral health services, with an emphasis on crisis services; and improve care coordination for people with serious mental illnesses and substance use disorders.

Central to the development of the CCBHC program was the creation of criteria that would define organizations most likely to realize the goals of the program. (The government is revising these criteria, but for this article, we focused on the criteria in use at the end of 2022.) The certification requirements involve six categories of criteria: staffing, availability and access to services, care coordination, scope of services, quality and other reporting, and organizational structure and authority (7). Within each of these categories are detailed criteria that each organization must meet to qualify as a CCBHC. Of economic significance are criteria related to staffing and the scope of services that must be available.

Notable among the staffing criteria are requirements that each CCBHC have a chief executive officer and full management team; a medical director (who can also be the chief executive officer) who must be a psychiatrist, except in shortage areas; and a medical professional who can prescribe buprenorphine (the specific practitioners with this ability vary according to states’ scope-of-practice regulations). The team should include peer support staff; clinicians credentialed as substance use disorder specialists; and people with expertise in trauma, serious emotional disturbance of children and adolescents, serious mental illnesses among adults, and co-occurring mental and substance use disorders.

These staffing criteria are aligned with CCBHCs’ scope of service requirements. CCBHCs are responsible for directly providing crisis services that include 24-hour mobile crisis teams, emergency crisis intervention services, and crisis stabilization. Crisis services offered by CCBHCs are meant to encompass suicide crisis response and substance use disorder–related services. In addition, a CCBHC must directly provide screening, assessment, diagnosis and risk evaluation services, person-centered treatment planning, and outpatient mental health and substance use disorder treatment services.

Crisis services offered by CCBHCs are meant to encompass suicide crisis response and substance use disorder–related services.

The remaining five required services can be directly provided by CCBHCs or supplied through arrangements with so-called designated collaborating organizations. These services comprise outpatient primary care screening; community-based mental health care for veterans; targeted case management; peer, family, support, and counselor services; and psychiatric rehabilitation services.

Together, these requirements imply a scale of operations that is likely to be larger than that of the average community mental health center. Although this scale likely was intended, it may also pose a barrier to creation of CCBHCs in rural areas and smaller cities and towns.

CCBHCs have been developed and supported through several different funding mechanisms. Some were established by and receive revenues from Medicaid as part of the PAMA section 223 demonstration program. Beginning in 2018, other CCBHCs were established and received initial and sometimes supplemental funding under “expansion grant” programs administered by SAMHSA. As of July 2022, SAMHSA has completed three rounds of grantmaking, issuing notices of funding opportunities (NOFOs) in 2018, 2020, and 2021. These NOFOs resulted in 115, 402, and 170 expansion grants in each of these three years, respectively; note that the same organizations often received multiple grants, sometimes within the same NOFO (7).

Each state selected for the Medicaid section 223 demonstration program had a choice of two payment structures (1). The first, prospective payment system 1 (PPS-1), pays CCBHCs in the demonstration program a fixed per diem for each day that a Medicaid beneficiary is served by the CCBHC. The payment arrangement mimics that of federally qualified health centers. A state may layer additional payments onto PPS-1, using a quality bonus payment based on performance on six specified quality indicators. The second payment option is PPS-2, which pays a fixed base amount to each demonstration CCBHC for each month during which a Medicaid beneficiary has received care from the CCBHC. That amount may vary according to the segment of the population to which the person served belongs and on the basis of a mandatory quality bonus payment rating. Moreover, outlier payments can be made to the CCBHC for costs above a specified level. As previously noted, under the BPSCA, the section 223 programs will be substantially expanded.

CCBHCs solely funded by SAMHSA grants are also required to meet the same standards as those qualifying for the section 223 program, even though they do not participate in the PPS-1 or PPS-2 payment arrangements. These grants have increased the number of CCBHCs across the nation and more recently allowed for expansions of existing CCBHCs.

As noted, a fundamental goal of the CCBHC program is to expand access to care for people with behavioral health needs regardless of their ability to pay or place of residence. A central question in measuring progress toward that goal is to assess how CCBHC funding is being allocated. We therefore conducted a series of analyses to characterize the communities where CCBHCs have been established, with a focus on their estimated prevalence of serious mental illness, poverty rates, and population density.

As noted, a fundamental goal of the CCBHC program is to expand access to care for people with behavioral health needs regardless of their ability to pay or place of residence.

.

To determine where CCBHCs are located, we constructed a data set of all active CCBHCs as of July 2022. This data set was based on the National Council for Mental Wellbeing’s (NCMW’s) CCBHC Locator (8) and cross-referenced against SAMHSA’s online grants dashboard. We used the NCMW’s list of CCBHCs by state and counties served (9) to link each CCBHC to the counties in its service area (one CCBHC may serve multiple counties) and to categorize each clinic as a demonstration site, expansion grantee, or state-certified clinic. We also merged information about all expansion grants issued between 2018 and 2021, including date issued, funding source (denoting when a grant was financed by COVID-19 relief funds), and award size, available in SAMHSA’s grants dashboard. Because this analysis focused on federal CCBHC policies, we did not supplement the NCMW list with all sites certified through independent state programs. However, we did account for these sites in a robustness check that yielded the same results as in the primary analysis. (For a detailed description of the CCBHC data set construction and the robustness check, see the supplement.)

To determine where CCBHCs are located, we constructed a data set of all active CCBHCs as of July 2022.

We focused our examination on whether CCBHCs serve counties that have high need for care and low economic resources (high-need, low-resource [HNLR] counties). We measured need for CCBHC services and local resource availability by merging county-level rates of serious mental illness (defined as a diagnosable mental, behavior, or emotional disorder that causes serious functional impairment that substantially interferes with or limits one or more major life activities) and poverty (defined as 100% of the federal poverty level), respectively, into our data set. Our measure of the prevalence of serious mental illness was drawn from the National Survey on Drug Use and Health 2016–2018 substate estimates. We linked the substate-level serious mental illness prevalence estimates to the counties contained within the region. County-level poverty rates were sourced from American Community Survey 5-year estimates for the 2016–2020 period. We designated counties in the fourth quartile of both serious mental illness and poverty rates as HNLR counties. We also included county- and state-level contextual variables (such as population density, state mental health expenditures per capita, and whether a state expanded Medicaid) in our data set. (See the supplement for a detailed description of each variable and its source.)

The assessment of CCBHC placement focused on two sets of analyses. In the first, we compared the location of CCBHCs with county-level rates of serious mental illness and poverty by using descriptive tables and a heat map. These descriptions were followed by a statistical analysis with regression models to determine which local characteristics are most strongly associated with the establishment of a CCBHC in each county.

As of July 2022, in total 432 CCBHCs served 621 counties across 42 U.S. states and Guam. The vast majority of CCBHCs (86%, N=372) served only one county; among the CCBHCs that served multiple counties, the average number served was seven and the maximum was 33 counties. Ten states and state equivalents (Delaware, Hawaii, Idaho, New Mexico, North Dakota, Puerto Rico, South Carolina, South Dakota, Washington, D.C., and Wyoming) were not served by any CCBHCs. (We
note that SAMHSA has recently provided funding to establish CCBHCs in Hawaii,Idaho, New Mexico, Puerto Rico, Washington, D.C., and Wyoming through CCBHC planning, development, and implementation grants announced on September 23, 2022.) Most CCBHCs were SAMHSA-funded expansion grantees; only 30 CCBHCs had never received an expansion grant (Table 1). The typical expansion grant award provides slightly less than $2 million per year.

We found that CCBHCs tended to serve populous urban areas: counties served by CCBHCs had an average population of approximately 280,000 residents, compared with roughly 60,000 residents in counties without CCBHCs. Furthermore, our analysis indicated that counties served by CCBHCs were >1.5 times more likely to be in metropolitan areas than were counties without a CCBHC (see Table OS2 in the supplement).

Table 1 shows that currently the bulk of CCBHCs, that is, 93% (N=402), are at least partially reliant on SAMHSA grant funding. Figure OS1 in the supplement shows a map that indicates the location of CCBHCs according to whether they were funded by the section 223 demonstration program. Given the important expansion of the demonstration program, the map can be expected to change over the coming years.

We found that CCBHCs tended to serve populous urban areas: counties served by CCBHCs had an average population of approximately 280,000 residents, compared with roughly 60,000 residents in counties without CCBHCs.

We further probed the characteristics of counties with and without a CCBHC by examining the share of counties served by a CCBHC according to their rates of poverty and serious mental illness. Table 2 summarizes this distribution, stratified by the type of funding mechanism supporting the CCBHCs. The table shows the share of counties served by CCBHCs in each quartile of serious mental illness and poverty, suggesting that the association between CCBHC presence and local need and resources may be slightly negative.

That is, the counties with the lowest rates of serious mental illness and poverty were more likely to be served by CCBHCs. An estimated 26% of counties in the first quartiles of serious mental illness and poverty (corresponding to the lowest rates for both) were served by any CCBHC, compared with 22% of counties in the fourth quartiles of serious mental illness and poverty (i.e., HNLR counties).

Table 3 further elucidates the relationship between poverty rates and the share of counties that are served by CCBHCs. About 20% of counties in the first quartile of poverty were served by CCBHCs, compared with 16% of counties in the fourth quartile, indicating that the highest-poverty counties were underrepresented among the counties served by CCBHCs. Similarly, 30% of the counties with the lowest rates of serious mental illness were served by CCBHCs, compared with 20% of those in the highest serious mental illness quartile.

Table 3 also shows that the counties most likely to have a CCBHC were those in the highest quartile of population density (i.e., number of residents per square kilometer), with 34% of counties in the fourth quartile of population density having a CCBHC, compared with 13% of counties in the first population-density quartile. One implication of this distribution is that CCBHCs are available to serve many people in poverty or with serious mental illness because they tend to be located in communities with large populations.

Figure 1 shows a heat map highlighting the counties where CCBHCs are located relative to whether a county is an HNLR county, illustrating the geographic dimensions of the summary statistics reported in Table 3. We noted a clustering of HNLR counties not served by CCBHCs in the south-central region of the country. Specifically, >60% of the HNLR counties not served by CCBHCs were in four states: Arkansas, Kentucky, Missouri, and Tennessee. We also noted that much of the west-central region was not served by CCBHCs, although the number of HNLR counties in that region is relatively low.

We examined these relationships further by using regression modeling to estimate the association between local factors and the likelihood of a county being served by a CCBHC. We included covariates for population density, poverty rate, serious mental illness prevalence, the non-White share of a county’s population, state Medicaid expansion status, state mental health expenditures per capita, the share of county residents with Medicaid and no other form of health insurance, and the share of county residents who are uninsured. The results of a regression analysis controlling for several local factors are reported in Table 4 and indicated that CCBHC presence was not significantly associated with poverty or serious mental illness rates. We briefly summarize the findings of the analyses below. (Complete results are shown in the supplement.)

Across the different regression models, population density was significantly, consistently, and positively associated with whether a county is served by a CCBHC. The most rural counties (i.e., completely rural or having <2,500 residents and not adjacent to a metro area) had an average population density of 6.1 residents per square kilometer, whereas the most urban counties (i.e., those in metro areas of ≥1 million residents) had an average population density of 543.4 residents per square kilometer. Holding all other variables constant in a regression analysis, we found that going from the population density typical of the most rural counties to that of the most urban counties was associated with an approximately 28-percentage-point increase in the likelihood of being served by a CCBHC (from 7% to 35%). (See full results of this analysis in the supplement.)

Counties in states that expanded Medicaid were significantly more likely to be served by a CCBHC. However, this observation appeared to be mediated through the Medicaid demonstration clinics, because the relationship between Medicaid expansion and CCBHC presence was attenuated when counties with clinics participating in the section 223 demonstration program were excluded from this analysis. We found some weak evidence of an association between poverty and a county being served by a CCBHC, but this relationship was not consistently significantly different from zero across the regression analyses.

This analysis revealed several factors associated with the establishment of CCBHCs. The decisions of states and communities about applying for CCBHC funding, in conjunction with decisions by SAMHSA and the Centers for Medicare and Medicaid Services regarding which states and localities to fund, result in population-dense, urban communities being more likely to be served by CCBHCs. Because communities with large populations also have large numbers of people in poverty or with serious mental illness, CCBHCs are providing services to such individuals. However, one consequence is that CCBHCs are not serving communities with lower population densities even if they have high rates of poverty and serious mental illness. Particularly striking are our results about the differential likelihood of being served by a CCBHC between communities with high or low population densities, with the latter being less likely to have CCBHC services. This observation means that, at least in part, CCBHCs are falling short of the goal of increasing access irrespective of a person’s ability to pay and residence.

These disparities likely stem from several factors. Program requirements stipulate that clinics should have a relatively large scale, evidenced by the fact that CCBHCs are on average 50% larger than an average community mental health clinic (our calculation from a National Mental Health Services Survey; data not shown). Thus, rural and small metropolitan counties may not have patient volumes that would allow for a CCBHC to operate at an efficient scale, given the extensive requirements for certification. This trend is compounded in states that have not expanded Medicaid. A second factor concerns the fact that allocations of funds require states or localities (depending on the funding mechanism) to apply for funding, favoring states and communities with an existing mental health infrastructure and sufficient financial resources. Communities with high levels of poverty and low population densities may not have an adequate mental health and human services infrastructure to offer a convincing case for funding and thus may be discouraged from applying for grant funds.

A second observation stemming from our results is that most CCBHCs rely on grant funding from SAMHSA. This imbalance is slated to change as the section 223 demonstration continues to expand. Although evidence shows that grant-based funding approaches can effectively support establishment of new treatment capacity, reliance on grant funding to sustain established CCBHCs is risky because of the confluence of shifting priorities and tight budgets. The recent potential to shift from an emphasis on ad hoc grants to Medicaid funding through the expanded demonstration program offers greater possibilities for sustainable support of CCBHCs. Moreover, recent SAMHSA funding announcements propose to “certify CCBHCs that represent diverse geographic areas, including rural and underserved areas” (10).

These observations point to a reconsideration of the role of SAMHSA grants and entry criteria, especially in the context of major expansions of the section 223 demonstration program. Expanding services to areas with lower population densities likely requires an approach somewhat different from that of allocating SAMHSA grant funds. Two features might be modified. The first would build on flexibilities incorporated into the most recent round of SAMHSA grantmaking that explicitly aim to build infrastructure and capacity to develop a CCBHC, sometimes by allocating grant funds to CCBHCs before they have met all certification requirements. A second modification might profitably seek to identify which requirements are essential to supporting the types of access and quality envisioned when the PAMA legislation was enacted. Existing requirements were based on experience, perceptions of best practice, and informal evidence. Now that the program has been in place for several years, an assessment to identify the most necessary ingredients could lead to changes in requirements that might offer the ability to create CCBHCs that could efficiently operate on a smaller scale than currently existing CCBHCs and would still provide high-quality services. As noted above, SAMHSA is undertaking a revision of federal requirements for CCBHCs.

Finally, in the interest of longer-term sustainability of the CCBHC program, the central role of SAMHSA grants might also be reconsidered. Relative to the ad hoc SAMHSA grants on which 93% of CCBHCs are at least partially dependent, Medicaid provides a more reliable long-term funding path for CCBHCs. As the section 223 program expands, the role of SAMHSA grants might be shifted toward providing targeted, timely funding to help new CCBHCs join the section 223 program and supplement CCBHCs in communities that have high rates of uninsured residents.