What Explains Changes in Availability of Specialty Mental Health Services in Organized Settings?

HELEN NEWTON; JENNIFER HUMENSKY; HOWARD GOLDMAN; SUSAN H BUSCH

doi:10.1111/1468-0009.12592

. 2022 Dec 28;100(4):1166–1191. doi: 10.1111/1468-0009.12592

What Explains Changes in Availability of Specialty Mental Health Services in Organized Settings?

HELEN NEWTON ^1,^✉, JENNIFER HUMENSKY ², HOWARD GOLDMAN ³, SUSAN H BUSCH ¹

PMCID: PMC9836237 PMID: 36575952

Abstract

Policy Points.

Community mental health facilities often do not offer the full range of evidence‐based clinical and support services for individuals with serious mental illness. Facilities were no more likely to offer six of seven services studied in 2019 compared with 2010 in both Medicaid expansion and nonexpansion states.
For‐profit facilities generally experienced the largest declines in service availability, while public facilities experienced the smallest declines with small increases in availability of select services.
New payment models that incentivize the offer of specialty support services may be needed to encourage adoption of clinical and support services by specialty mental health organizations.

Context

Community mental health facilities often do not offer the full range of evidence‐based clinical and support services for individuals with serious mental illness. This creates equity issues, particularly when low‐income and minority communities have access to fewer facilities. Medicaid expansion might encourage facilities to offer these services. However, this decision may also be affected by facility ownership type or mediated by service cost structure, particularly in the absence of innovative payment mechanisms. In this study, we determine whether and how Medicaid expansion and facility ownership are associated with changes in specialty mental health service availability in organized settings over time.

Methods

We estimated two‐way fixed effects models using six cross‐sections of the National Mental Health Services Survey and compared changes in facility‐reported offering of seven services from 2010 to 2019 (54,885 facility years): psychotropic medication, case management, family psychoeducation, psychiatric emergency walk‐in services, supported employment, assertive community treatment, illness management, and recovery services. We tested whether Medicaid expansion and facility ownership (private for‐profit, private not‐for‐profit, public) were associated with differential changes in service availability from 2010 to 2019.

Findings

Overall, facilities were no more likely to offer nearly all services in 2019 than 2010. We found smaller declines for psychotropic medication and psychiatric emergency walk‐in services among facilities in Medicaid expansion states compared to declines in non‐Medicaid expansion states (6.3 (95% CI 95% CI = 1.8‐10.7) and 5.5 (95% CI = 0.2‐10.8) percentage points respectively). For‐profit facilities experienced the largest declines in availability from 2010 to 2019, while public facilities experienced the smallest declines and some increases in availability of select services.

Conclusions

Specialty mental health services are still not widely offered in community outpatient settings despite significant investments in Medicaid, although Medicaid expansion was associated with slower declines in availability. New payment models that incentivize outpatient facilities to offer clinical and support services may be needed.

Keywords: access to care, community health, mental health, organization of care, alternative payment models

Serious mental illness (smi) can cause significant functional disability, morbidity, and premature mortality. ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ Specialty mental health services—including clinical services such as psychotropic medication, support services, and crisis stabilization services—reduce symptom severity and improve functional ability and quality of life. ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ The Levels of Care Utilization System (LOCUS) guidelines developed by the American Association of Community Psychiatrists identify the types of services that should be availability in community‐based settings. ¹⁴ Some states have used these guidelines to define the services that facilities should offer, though oversight may be limited. ¹⁵ For individuals requiring medically monitored community‐based services (LOCUS level 4)—of whom most are individuals with severe and persistent illness—guidelines note that clinical, supportive, and crisis stabilization and prevention services should all be offered. ¹⁶ Access to these services is of concern to policymakers including those at the Substance Abuse and Mental Health Services Administration (SAMHSA) and state Medicaid programs. Although access can be defined in several ways we focus here on system‐level access or service availability at the facility level, which has precipitously declined. ¹⁷ , ¹⁸ , ¹⁹ From 2010 to 2017, the share of community mental health facilities offering support and crisis stabilization services has decreased by half nationally, though demand for outpatient mental health services has increased over the same time period.^20‐23 It is particularly important that facilities offer evidence‐based services in settings where there are few facilities, such as many rural communities, low‐income communities, and communities of color. ²⁴ Other measures of access such as patient‐level utilization are also important but offering by providers is a critical precursor to patient use.

Public insurance expansions—through Medicaid expansion—could explain some of the changes in specialty mental health service availability. Medicaid expansion was expected to increase access to services by insuring previously uninsured adults, thus reducing financial barriers to care. ²⁵ , ²⁶ Yet, if providers fail to offer a service, lower cost is unlikely to affect use, particularly when patients have access to a single or few providers. Whether Medicaid expansion led to more community mental health facilities offering these services has not been studied. By increasing insurance coverage, expansions may have increased facilities’ financial solvency and thus made them more apt to offer these services to patients. ²⁷ However, increases in Medicaid insured patients may have incentivized facilities to focus on services that are covered by Medicaid, are more favorably reimbursed, or that more easily adhere to Medicaid regulatory requirements. ²⁸ In addition, traditional per‐service reimbursement is unlikely to cover the start‐up costs needed to build referral networks, train staff, or develop new infrastructure to offer support services. Thus, whether and how Medicaid expansion affected facility decisions to offer specialty mental health services is unknown.

Facility ownership could also influence the availability of specialty mental health services. Though public and not‐for‐profit agencies have historically operated most community mental health facilities, insurance expansions—through Medicaid expansion, dependent coverage expansion under the Affordable Care Act (ACA), and increased generosity of coverage due to the Mental Health Parity and Addiction Equity Act (MHPAEA)—could lead to an increase in for‐profit ownership and changes in the types of services offered. ²⁹ , ³⁰ , ³¹ For example, private equity firms have invested in behavioral health treatment facilities and cited insurance expansions as their motivation. ³² , ³³ , ³⁴ Per the annual National Mental Health Services Survey (N‐MHSS), the share of facilities operated by for‐profit organizations doubled between 2010 and 2020 from 10% to 21%. ³⁵ , ³⁶ Evidence from the dialysis, nursing home, and substance use disorder treatment systems suggests that for‐profit ownership could lead to reduced access to support services that are labor intensive, as private equity firms commonly achieve efficiencies through downsizing staffing ratios. ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² For‐profit facilities, unlike public or not‐for‐profit facilities, often do not have access to public grant funds and may find that specialty services that are team based or require specifically trained staff or new infrastructure (e.g., Assertive Community Treatment, supported employment) are unprofitable. If these facilities prioritize private pay patients, support services provision may be less of a focus since these services may not be covered and those patients may have less disabling conditions.

As noted above, specialty mental health services differ in their cost structure and these differences likely moderate the impact of Medicaid expansion and facility ownership on changes in specialty mental health service availability. All services include some start‐up costs (here considered sunk costs that cannot be recouped), fixed costs, and marginal costs. One‐time start‐up costs, which are unlikely to be reimbursed under traditional Medicaid reimbursement, may include specialized staff training, developing referral relationships with community partners, and investing in technology or other infrastructure. ⁴³ , ⁴⁴ Fixed costs are high in team‐based models like Assertive Community Treatment (ACT) or supported employment, where more staff are needed per client served and staff spend time reviewing and coordinating care in addition to seeing clients face to face (see conceptual model, Table 1). ⁸ , ⁴⁵ These services may also require strict adherence to program standards related to program admission, staffing, service components and program evaluations such as those required by Illness Management and Recovery programs. In addition, these team‐based services are all optional Medicaid benefits and some states may not reimburse them in some eligibility categories. Finally, some services are frequently covered by insurers at reimbursement that covers marginal costs. For example, medication management, which typically requires a psychiatrist or other prescriber to see clients individually, is often covered by insurers at reimbursement that may approach the costs of providing the service.

Table 1.

Specialty Mental Health Services With and Without High Start‐Up Costs

	Target Patient Population
	Patients With Serious Mental Illness	Patients Without Serious Mental Illness
	(Medicaid eligible prior to expansion)	(not Medicaid eligible prior to expansion)
*Services Without* High Start‐Up Costs**
Delivered by one clinician (vs. team‐based); mandatory federal Medicaid benefit, more flexible reimbursement requirements and/or more easily reimbursable	Family Psychoeducation	Psychotropic Medication Case Management
*Services With* High Start‐Up Costs**
Team‐based; optional federal Medicaid benefit, more stringent reimbursement requirements and/or more difficult to reimburse	Illness, Management, and Recovery (IMR) Assertive Community Treatment (ACT)	Walk‐in Psychiatric Emergency Services
	Supported Employment

Open in a new tab

In this study, we compare changes in availability of a range of specialty mental health services from 2010 to 2019 to test whether availability of specialty mental health services in organized settings is associated with Medicaid expansion from 2010 to 2019, distinguishing between services with high start‐up costs and those without (see conceptual model, Table 1). Given the dramatic changes in facility ownership over this time period, we then examine whether certain ownership types had differential trends in provision of these services.

These efforts are important because they have implications for how alternative payment models might be structured to encourage provision of specialty mental health services. Alternative payment models, like other value‐based payment strategies, align financial incentives with clinical performance to reward high‐quality, cost‐effective care. ⁴⁶ Linking financial incentives to clinical performance could improve access to specialty mental health services, but the structure of these models, such as whether the model targets primary care or specialty settings, and the payment strategies used to reimburse for specialty mental health services, could dictate changes in service availability. By studying the effects of these two major changes in the specialty mental health treatment landscape—Medicaid expansion and changes in facility ownership—and by distinguishing changes for services closer to a traditional clinical model and those that are team‐based, we hope to glean information that will be useful to policymakers as they consider how and whether alternative payment models can increase the availability of these specialty services.

Methods

Study Population

We used responses to six waves (2010, 2014‐2019) of the National Mental Health Services Survey (N‐MHSS) to identify a sample of 54,885 outpatient mental health facility‐years. The N‐MHSS, an annual survey fielded by SAMHSA, collects administrative data from all known public and private mental health treatment facilities nationally, excluding those operated by the Department of Defense, jails and prisons, and independent practitioners ³⁶ with a response rate that ranged from 87% to 92% for the years included in this analysis. This sample frame includes all private and public psychiatric hospitals, nonfederal general hospitals with separate psychiatric units, US Department of Veterans Affairs medical centers, residential treatment centers for children and adults, community mental health centers, outpatient day treatment or partial hospitalization mental health facilities, and multisetting (non‐hospital) mental health facilities. Responding facilities can choose to not be included in the publicly available responses. This percentage is typically small, around 3%‐4%. For this analysis, we excluded any facilities who reported only caring for clients in inpatient or residential settings, or exclusively caring for children aged 0‐17. Our study sample therefore included all outpatient mental health treatment centers, including community mental health centers, federally qualified health centers that are licensed mental health treatment providers, and outpatient mental health facilities affiliated with medical groups or hospital systems.

Main Outcomes and Measures

In the N‐MHSS, mental health treatment facilities reported offering certain treatments and ancillary services for serious mental illness. We used responses to seven of these questions as our primary outcome measures and measured whether mental health facilities offered: psychiatric emergency walk‐in services, psychotropic medication, family psychoeducation, case management, supported employment, assertive community treatment, and illness management and recovery services (see question text in Online Appendix eTable 1). These services were all designated as essential community‐based mental health services in Mental Health: A Report of the Surgeon General, published in 1999, and comprise the continuum of services—including clinical, support, and crisis stabilization—required to serve clients with complex needs in community‐based settings. ¹⁷ Organized mental health settings are still the only setting in which patients can access many mental health support and crisis stabilization services in the community. While patients may receive some services such as psychotropic medication in the primary care setting or from office‐based solo practitioners, for individuals with severe and persistent illness the availability of services in one location is likely to increase use. As shown in Table 1, we distinguished between services without high start‐up costs (psychotropic medication, case management, family psychoeducation), and services with high start‐up costs (psychiatric emergency walk‐in services, supported employment, illness management and recovery, and assertive community treatment). ⁴⁷ , ⁴⁸ , ⁴⁹ , ⁵⁰

Facilities also reported other facility characteristics and state location in the N‐MHSS. We used facility state location to identify facilities located in states that expanded Medicaid from 2014 to 2016 and self‐reported facility ownership to identify private for‐profit, private not‐for‐profit, and publicly run facilities in each year. Facilities also reported their primary treatment focus (a five‐category variable corresponding to mental health only, a mix of mental health and substance use disorder treatment, general health, or other) and the settings in which treatment services were offered (inpatient, residential, and outpatient settings). We used state‐level estimates from the Census to determine state racial and ethnic distribution and age distribution over the study period and estimates from the Bureau of Labor Statistics to determine state unemployment distribution over time.

Statistical Analysis

For each of our seven primary outcomes, we compared changes in service offering over time using two‐way fixed effects models. Our first set of models tested whether location in a Medicaid expansion state was associated with changes in service offering before and after 2014. The model included a binary variable identifying facility location in a 2014‐2016 Medicaid expansion state, the first predictor of interest, year after 2014 (a binary variable identifying years 2014‐2019, the post period), and the interaction of Medicaid expansion and post—the coefficient capturing the estimated impact of Medicaid expansion on service offering. We did not have survey responses prior to 2010 so we excluded facilities located in states that expanded Medicaid early (CA, CO, CT, DC, MA, MN, WA) from this model and compared 2014‐2016 expansion states to states who have not yet expanded or expanded within a year of the study end (VA and ME, which expanded in 2019).

Our second set of models tested whether facility‐reported ownership type, the second predictor of interest, was associated with service offering. For models estimating the relationship between facility ownership and service offering, the coefficients of interest were the interaction of ownership and year, which estimated the differential impact of ownership type on service offering over time.

In all models, for ease of interpretation, we used linear probability models. We adjusted for several facility‐level characteristics: whether the facility also included an inpatient site, the focus of treatment (mental health only, mental health and substance use disorder, general health, or other), and the annual state unemployment rate. We included year and state fixed effects and clustered standard errors at the state level. We tested several different specifications of the model as a robustness check, including restricting to facilities who reported accepting Medicaid as payment, aggregating facility‐years into a state‐year panel, and adjusting for facility size, and tested a logit (versus linear probability) model. Because the effects of Medicaid expansion on service availability could differ depending on the year that the state expanded Medicaid (2014, 2015, or 2016), we estimated this model using the Callaway Sant'Anna estimator—which accounts for dynamic treatment effects across multiple time periods—as a sensitivity analysis (eTable 4). ⁵¹ , ⁵²

Results

Characteristics of Outpatient Mental Health Treatment Facilities, 2010‐2019

Our final sample included 54,885 facility‐years. In 2010, a minority (13%) of outpatient mental health facilities reported having an inpatient site in addition to their outpatient facility and most (69%) reported providing only mental health treatment services versus mental health and substance use disorder treatment or general health services (Table 2). Facilities were more likely to report dual treatment (substance use disorder treatment and mental health services) or a general health focus in 2019 versus 2010. From 2010 to 2019, the share of mental health outpatient facilities owned by private for‐profit organizations more than doubled, from 8% to 20% of facilities.

Table 2.

Characteristics of Outpatient Mental Health Treatment Facilities Responding to the National Mental Health Services Survey (N‐MHSS) 2010, 2014‐2019

	Full Sample	2010	2019
Facility Characteristics
Facility‐years (N)	54,885	6184	8544
Multiple Settings
Also has an Inpatient Setting	9%	13%	9%
Treatment Focus
Mental Health	60%	69%	56%
Mental Health/Substance Use Disorder	37%	31%	39%
General Health	3%	0%	4%
State Characteristics
Age
Percent 0‐18	23%	24%	22%
Percent 18‐24	10%	10%	9%
Percent 25‐34	14%	13%	14%
Percent 35‐54	26%	28%	25%
Percent 55‐64	13%	12%	13%
Percent 65+	16%	13%	17%
Race
Percent White	79%	79%	78%
Percent Black	13%	13%	13%
Percent Asian	5%	4%	5%
Percent American Indian/Alaskan Native	1%	1%	1%
Percent Native Hawaiian/Pacific Islander	0%	0%	0%
Ethnicity
Percent Hispanic	14%	13%	14%
Sex
Percent Male	49%	49%	49%
Employment Status
Percent Unemployed	16%	19%	11%
Exposures/Predictors of Interest
Facility located in Medicaid Expansion state ^a	57%	20%	67%
Facility Ownership
Private for‐profit	16%	8%	20%
Private not‐for‐profit	62%	66%	59%
Public	22%	26%	21%

Open in a new tab

Notes: Facility Characteristics were derived from questions in the N‐MHSS. State Characteristics were derived from Census (age, sex, race), and Bureau of Labor Statistics (unemployment rate).

^{^a}

States that expanded Medicaid in 2014‐2016 are AK, AZ, AR, DE, HI, IA, IL, IN, KY, LA, MD, MI, MT, ND, NH, NM, NY, OH, OR, PA, RI, VT, and WV.

Availability of Specialty Mental Health Services

The majority of facilities offered psychotropic medication, case management, and family psychoeducation services—services without high start‐up costs.

Less than half offered emergency walk‐in services or illness management and recovery services, and only a quarter offered supported employment services or assertive community treatment (Figure 1). Mental health outpatient facilities were no more likely to offer six of the seven specialty mental health services studied in 2019 compared to 2010, with slightly more facilities offering family psychoeducation services.

Specialty Mental Health Service Availability Among Outpatient Mental Health Treatment Facilities, 2010‐2019.

*Note*: Figure 1 shows the unadjusted mean proportion of outpatient mental health facilities that reported using the seven services analyzed in this study. The N‐MHSS was either not fielded or did not include questions about these specific services 2011, 2012, or 2013, but we include these years here to appropriately scale the x‐axis.

Effect of Medicaid Expansion

There were important changes in service offerings among facilities located in Medicaid expansion and nonexpansion states over time. The analytic sample for the Medicaid analysis included 23,955 facility‐years located in states that expanded Medicaid between 2014 and 2016 and 19,955 facility‐years located in states that had not expanded Medicaid during the study period. While the unadjusted availability of nearly every service did not increase in both Medicaid expansion and nonexpansion states, declines were smaller for all but two services in Medicaid expansion states (eTable2). The declines in case management and illness management and recovery services were the same in expansion and nonexpansion states (eTable2).

After adjusting for state and facility characteristics, the share of facilities offering psychotropic medication declined 6.3 percentage points less (95% CI = 1.8‐10.7) in Medicaid expansion states relative to nonexpansion states (Figure 2). This was also the case for psychiatric emergency walk‐in services. While the availability of psychiatric emergency walk‐in services declined in both expansion and nonexpansion states, the decline in availability was 5.5 percentage points less (95% CI = 0.2‐10.8) among facilities in expansion states relative to those in nonexpansion states. There were no statistically significant differences in changes in the availability of the other five services studied. Full regression results can be found in eTable 3. Results from sensitivity models and those that used the Callaway Sant'Anna estimator were similar in direction and magnitude to results from the main specification (eTables 4‐6).

Changes in Specialty Mental Health Service Availability in Outpatient Mental Health Treatment Facilities After Medicaid Expansion. [Colour figure can be viewed at wileyonlinelibrary.com]

*Note*: Figure 2 presents results from seven linear probability models that predicted each service (presented in the left most column). Primary predictors included whether the state in which the facility was located was part of the treatment group (expanded Medicaid in 2014‐2016 (1)) or not (state that has not, or has not yet (by 2019) expanded Medicaid (0)), an indicator for years after 2014 (post_2014), the interaction of treatment and post (the coefficient of interest, which is presented in this exhibit), facility characteristics including whether or not the facility included an inpatient site and the focus of treatment (mental health, mental health and substance use disorder, or general health), time‐varying state unemployment rate, and year and state fixed effects with standard errors clustered at the state.

Effect of Facility Ownership

In unadjusted results, family psychoeducation was the only service studied to increase in all ownership types. The unadjusted availability of the six remaining studied services declined among for‐profit facilities from 2010 to 2019 with declines ranging from ‐5 to ‐15 percentage points (eTable 7). Not‐for‐profit facilities also experienced declines although these were smaller in magnitude (ranging from ‐1 to ‐7 percentage points), while public facilities were more likely to also offer psychiatric emergency walk‐in services and psychotropic medication in 2019 compared to 2010 (eTable 7), with small declines in the other studied services, ranging from ‐2 to ‐5 percentage points.

In adjusted analysis, among services without high start‐up costs, for‐profit facilities were significantly less likely to offer psychotropic medication services in 2019 than 2010 (‐10.7 percentage points [95% CI = ‐16.7, ‐4.7], versus ‐1.5 percentage points [95% CI = ‐5.1, 2.2] among not‐for‐profits and a 0.4 percentage point increase among public facilities [95% CI = ‐2.4, 3.2]); and ‐14.4 percentage points less likely (95% CI = ‐23.7, ‐5.0) to offer case management services (relative to ‐2.1 percentage points among not‐for profit [95% CI = ‐5.7, 1.5] and ‐3.7 among public facilities [95% CI = ‐9.0, 1.6]) (Figure 3). Among services with high start‐up costs, for‐profit facilities were significantly less likely to offer psychiatric walk‐in emergency services relative to not‐for‐profit or public facilities in 2019 than in 2010 (‐12.4 percentage points [95% CI = ‐19.3, ‐5.5] versus ‐8.2 percentage points [95% CI = ‐13.6, ‐2.9] among not‐for‐profit and 1.6 [95% CI = ‐4.7, 7.8] among public facilities. Full regression results can be found eTable 8. Results from sensitivity analyses were similar in direction to the main specification (eTables 9 and 10).

Changes in Specialty Mental Health Service Availability Among For‐Profit, Nonprofit, and Publicly Operated Outpatient Mental Health Treatment Facilities, 2010‐2019. [Colour figure can be viewed at wileyonlinelibrary.com]

** indicates services for which the difference in difference estimate were statistically significant (p<0.05). E.g., change in availability among either public or not‐for‐profit operated facilities was significantly different from the change availability among for‐profit operated facilities.

*Note*: Figure 3 presents results from seven linear probability models that predicted availability of each specialty mental services studied (presented in the left most column). Primary predictors included 1) a categorical variable that identified whether the facility was operated by a non‐profit, public, or for‐profit organization, 2) an indicator for each year of data (2010, 2014‐2019), and the interaction of ownership and year (our coefficients of interest). We adjusted each model for facility characteristics—whether the facility included an inpatient site and the focus of treatment (mental health, mental health and substance abuse, or general health)—as well as time‐varying state unemployment rate, and state fixed effects with standard errors clustered at the state. The results plotted in this graph show the percentage point change in availability from 2010‐2019 by ownership type.

Limitations

Study limitations should be noted. We relied on survey data to measure facility‐level offerings of specialty mental health services so our outcome measures may be inflated due to response bias. Risk of response bias is minimal, however, given the high N‐MHSS response rate. Our sample is limited to facilities identified as mental health treatment facilities, though there are potentially some facilities or individual practitioners who provide mental health services but do not identify as mental health treatment facilities. As we are focused specifically on outpatient mental health treatment facilities, this exclusion is not as important, but does present a limitation in the generalizability of our findings. Importantly, we do not know how frequently clients received specialty mental health services, nor the number of clients served by facilities in our sample. Additionally, we cannot observe referral relationships that facilities may share. We only had one pre‐period (2010), as our outcomes of interest were not ascertained in the 2012 survey. Finally, as an observational study, we can only identify associations.

Discussion

Using a national cohort of outpatient mental health facilities, we studied how Medicaid expansion and facility ownership explained changes in the availability of a range of specialty mental health services in organized settings from 2010 to 2019. Facilities that provide medically monitored non‐residential services—the highest level of community‐based treatment for individuals with severe and persistent mental illness—are expected to provide a continuum of clinical, support, and crisis stabilization services, yet availability of these essential services has decreased nationally. We found that facilities were no more likely to offer six of the seven studied services in 2019 compared with 2010. Facilities in both Medicaid expansion and nonexpansion states experienced declines, though declines were less for select services in Medicaid expansion states. The share of outpatient mental health facilities operated by for‐profit organizations more than doubled over the study period, and for‐profit facilities experienced the largest declines in service availability.

We found that changes in availability of services in Medicaid expansion and nonexpansion states was different depending on the service cost structure. Facilities in Medicaid expansion states saw no change in psychotropic medication availability—a service without high start‐up costs and favorable insurance reimbursement—after expansion. Even though the share of facilities offering psychotropic medication in Medicaid expansion states remained constant before and after expansion, a significantly smaller share of facilities in nonexpansion states reported offering psychotropic medication after 2014. Importantly, we also found that facilities in Medicaid expansion states were no less likely than those in nonexpansion states to offer services with high start‐up costs after expansion. Under conventional reimbursement mechanisms, start‐up costs may not be reimbursed. We expected that these services may decline in availability after expansion as facilities decided to instead align focus with reimbursement. Rather, we found that declines in the share of facilities offering psychiatric emergency walk‐in services after 2014 were smaller in expansion versus nonexpansion states. While it is reassuring that the declines in psychiatric emergency walk‐in services were smaller among Medicaid expansion states relative to nonexpansion states, the fact that the offer of these services declined in both expansion and non‐expansion states may have implications for the new mental health crisis line, 988. ⁵³ , ⁵⁴

Surprisingly, we found that for‐profit facilities experienced declines in services without high start‐up costs. For‐profit facilities had significantly larger declines in the availability of psychotropic medication and case management services relative to the declines in not‐for‐profit and public facilities, though like not‐for‐profit and public facilities, the availability of family psychoeducation services grew in for‐profit facilities in 2019 compared with 2010. The decrease in availability of psychotropic medication among for‐profit facilities is surprising as medication services are often favorably reimbursed, are often offered to patients with mild or moderate mental illness, and are without high start‐up costs (unlike emergency walk‐in services). However, to offer psychotropic medication services does require having a psychiatrist or another prescriber on‐site, which may be costly, especially because most communities report shortages of psychiatrists. ⁵⁵ For‐profit facilities were instead more likely to offer family psychoeducation services, which do not require clinical staff with a credential allowing prescribing and have more a flexible reimbursement structure. Decisions to offer family psychoeducation services rather than psychotropic medication could also suggest that for‐profit facilities may specialize in providing wrap‐around therapy services meant to complement, rather than substitute, the medications that can be offered by providers in other settings, including primary care.

We found that public facilities were the only facilities to increase offerings of emergency walk‐in services—a service with high start‐up costs—in addition to services without high start‐up costs, such as psychotropic medication and family psychoeducation, over the study period. Public facilities also saw the smallest declines or no change in availability among the other studied services. These increases suggest that Medicaid insurance expansions may build on other public financing mechanisms such as Community Mental Health Block Grants. Block grants are an important lever for SAMSHA and other federal stakeholders to invest in specialty mental health services, but alone are not sufficient to sustain these services. For example, states have set aside block grant funds to implement evidence‐based services for first episode psychosis (coordinated specialty care) but states often leverage favorable Medicaid reimbursement rates and other state dollars in addition to block grant funds to achieve financial sustainablity. ⁵⁶ Our finding that after Medicaid expansion public facilities reported increases in services like psychotropic medication and psychiatric emergency walk‐in services, and only small declines in supported employment—services that are frequently offered as part of coordinated specialty care services—could indicate that insurance expansions have improved access to services financed initially by block grants.

Our findings suggest that expanding coverage to more individuals will likely not be enough to ensure access to evidence‐based clinical and support services to individuals with severe and persistent mental illness. However, our findings may offer insight on how alternative payment models can be structured to expand access to specialty mental health services in organized settings. To date, the alternative payment model with the most enthusiasm and provider participation are ACO contracts—models where participating providers are held accountable for the total cost of care and quality for an attributed patient population. ⁵⁷ In theory, these models could encourage participating providers to develop referral relationships with outpatient mental health treatment facilities to provide specialty mental health services. However, these models may fail to encourage providers to offer services with high start‐up costs that are used by the relatively narrow set of the population with serious mental illness. Most ACOs report that their providers do not have the ability to refer patients to specialty support services, including many who even have a behavioral health facility participating in their ACO contract. ⁵⁸ This suggests that to be successful, new payment models specifically focused on specialty mental health organizations may be needed, and that these may need to include a focus on team‐based support services. Our work also suggests for‐profit facilities may be less apt to offer these support services. It has been argued that it may be possible to create value‐based payment models that align the incentives of for‐profit investors and health care purchasers and deliver high quality care, but this may be easier to achieve in primary rather than specialty care. ⁵⁹

In contrast to ACOs, prospective payment models in outpatient mental health settings could encourage facilities to offer certain services and thus expand availability of support services, which tend to have high start‐up costs. The Protecting Access to Medicare Act (PAMA) authorized the establishment of the Certified Community Behavioral Health Clinics (CCBHC) Demonstration, which reimburses participating facilities prospectively for the full cost of providing services (including start‐up and fixed costs) using a per visit rate in return for providing nine evidence‐based services, including psychiatric support services. ⁶⁰ Under this model, one third of facilities participating in the 2017‐2019 demonstration added emergency crisis services, and a quarter added supported employment and illness, management, and recovery services. ⁶¹ Of note, for‐profit facilities were not eligible for CCBHC status although they are allowed to contract with CCBHCs to provide some services and become designated collaborating organization (DCOs) with certain requirements. Similarly, some Medicaid managed care organizations have dedicated payments to providers to finance services with high start‐up costs. For example, Tennesee's Medicaid program, TennCare, currently has an arrangement in which Medicaid managed care organizations designate practice transformation payments to clinics to support the development of specific case management services for enrollees with serious mental illness. ⁶² Multi‐component models could also expand offer of specialty mental health services. One creative payment model initially proposed for first episode psychosis treatment included a prospective per‐case payment to reward organizations for identifying individuals for whom these specialty mental health services would be clinically appropriate and to cover case identification, retention and outreach that could not easily be attributed to an individual patient. ⁶³

Conclusion

Community mental health facilities were no more likely to offer most specialty mental health clinical and support services in 2019 than 2010. Even with significant investments in Medicaid, access to specialty mental health services in organized settings had the largest declines in for‐ profit organizations, while public facilities had smaller declines or even small increases. New payment models that incentivize the offer of specialty support services may be needed to encourage adoption of these models by specialty mental health organizations.

Funding/Support: US Department of Health and Human Services, National Institutes of Health, National Institute of Mental Health (R01MH106635).

Conflict of Interest Disclosures: None.

Supporting information

Technical Appendix

eTable 1: N‐MHSS question text and response sets

eTable 2: Availability in specialty mental health services by state Medicaid expansion status, 2010‐2019

eTable 3: Regression results for Medicaid expansion models (Figure 2 in text)

eTable 4: Estimated effects of Medicaid expansion on availability of specialty mental health services using the Callaway Sant'Anna estimator

eTable 5: Comparisons of coefficient of interest for Medicaid expansion models across different adjustment specifications

eTable 6: Comparison in coefficients for Medicaid expansion models across different sensitivity analyses

eTable 7: Availability of specialty mental health services by facility ownership type, 2010‐2019

eTable 8: Regression results for ownership models (Figure 3 in text)

eTable 9: Comparisons of coefficient of interest for ownership models across different adjustment specifications

eTable 10: Comparison in coefficients of interest for ownership models across different sensitivity analyses

Click here for additional data file.^{(75.2KB, docx)}

References

1. Germack HD, Noor‐E‐Alam M, Wang X, Hanrahan N. Association of comorbid serious mental illness diagnosis with 30‐Day medical and surgical readmissions. JAMA Psychiatry. 2019;76(1):96‐98. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Hendrie HC, Lindgren D, Hay DP, et al. Comorbidity profile and healthcare utilization in elderly patients with serious mental illnesses. Am J Geriatr Psychiatry. 2013;21(12):1267‐1276. [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Cloutier M, Aigbogun MS, Guerin A, et al. The economic burden of schizophrenia in the United States in 2013. J Clin Psychiatry. 2016;77(6):764‐771. [DOI] [PubMed] [Google Scholar]
4. Howell S, Yarovova E, Khwanda A, Rosen SD. Cardiovascular effects of psychotic illnesses and antipsychotic therapy. Heart. 2019;105(24):1852‐1859. [DOI] [PubMed] [Google Scholar]
5. McGinty E, Baller J, Azrin S, Juliano‐Bult D, Daumit G. Quality of medical care for persons with serious mental illness: A comprehensive review. Schizophr Res. 2015;165(0):227‐235. [DOI] [PMC free article] [PubMed] [Google Scholar]
6. Spivak S, Cullen B, Eaton WW, Rodriguez K, Mojtabai R. Financial hardship among individuals with serious mental illness. Psychiatry Res. 2019;282:112632. [DOI] [PubMed] [Google Scholar]
7. Walker ER, McGee RE, Druss BG. Mortality in mental disorders and global disease burden implications. JAMA Psychiatry. 2015;72(4):334‐334. d [DOI] [PMC free article] [PubMed] [Google Scholar]
8. Drake RE, Mueser KT, Torrey WC, et al. Evidence‐based treatment of schizophrenia. Curr Psychiatry Rep. 2000;2(5):393‐397. [DOI] [PubMed] [Google Scholar]
9. Kane JM, Robinson DG, Schooler NR, et al. Comprehensive versus usual community care for first episode psychosis: Two‐year outcomes from the NIMH RAISE Early Treatment Program. Am J Psychiatry. 2016;173(4):362‐372. [DOI] [PMC free article] [PubMed] [Google Scholar]
10. Christensen TN, Wallstrøm IG, Stenager E, et al. Effects of individual placement and support supplemented with cognitive remediation and work‐focused social skills training for people with severe mental illness: A randomized clinical trial. JAMA Psychiatry. 2019;76(12):1232‐1240. [DOI] [PMC free article] [PubMed] [Google Scholar]
11. Dixon L, Goldman H, Srihari V, Kane J. Transforming the treatment of schizophrenia in the United States: The RAISE Initiative. Annual Review of Clinical Psychology. 2018;14:237‐258. [DOI] [PMC free article] [PubMed] [Google Scholar]
12. Valenstein M, McCarthy JF, Ganoczy D, et al. Assertive community treatment in veterans affairs settings: impact on adherence to antipsychotic medication. Psychiatr Serv. 2013;64(5):445‐451, 451.e1. [DOI] [PubMed] [Google Scholar]
13. Biasi B, Dahl MS, Moser P. Career Effects of Mental Health. Social Science Research Network; 2020. [Google Scholar]
14. American Association of Community Psychiatrists . LOCUS: Levels of Care Descriptions. Level of Care Utilization System for Psychiatric and Addiction Services. https://sites.google.com/view/aacp123/resources/locus. Accessed May 6, 2022.
15. McBain RK, Eberhart NK, Breslau J, et al. How to Transform the U.S. Mental Health System: Evidence‐Based Recomendations. RAND Corporation; 2021. https://www.rand.org/pubs/research_reports/RRA889‐1.html#:~:text=Strengthen/20mental/20health/20parity/20regulation,interventions%20for%20serious%20mental%20illness. Accessed November 2, 2022. [Google Scholar]
16. American Association of Community Psychiatrists . LOCUS: Levels of Care Descriptions. Published 2016. www.comm.psych.pitt.edu/find.html. Accessed November 2, 2022.
17. US Department of Health and Human Services . Mental Health: A Report of the Surgeon General. U.S. Department of Health and Human Services, Substance Abuse and Mental Health Services Administration, Center for Mental Health Services, National Institutes of Health, National Institute of Mental Health; 1999. [Google Scholar]
18. Penchansky R, Thomas JW. The concept of access: definition and relationship to consumer satisfaction. Med Care. 1981;19(2):127‐140. [DOI] [PubMed] [Google Scholar]
19. Donabedian A. Evaluating the Quality of Medical Care. Milbank Q. 2005;83(4):691‐729. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Cummings JR, Smith JL, Cullen SW, Marcus SC. The changing landscape of community mental health care: Availability of treatment services in national data, 2010–2017. PS. 2020;72(2):204‐208. [DOI] [PubMed] [Google Scholar]
21. US Department of Health and Human Services . National Survey on Drug Use and Health 2018 . US Department of Health and Human Services, Substance Abuse and Mental Health Services Administration, Center for Behavioral Health Statistics and Quality; 2019. http://datafiles.samhsa.gov/. Accessed November 2, 2022.
22. Mojtabai R, Olfson M. National trends in mental health care for US adolescents. JAMA Psychiatry. 2020;77(7):703‐714. d [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Germack HD, Drake C, Donohue JM, Golberstein E, Busch SH. National trends in outpatient mental health service use among adults between 2008 and 2015. Psychiatr Serv. 2020;71(11):1127‐1135. [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Cummings JR, Allen L, Clennon J, Ji X, Druss BG. Geographic access to specialty mental health care across high‐ and low‐ncome US communities. JAMA Psychiatry. 2017;74(5):476‐484. [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Zuvekas SH, McClellan CB, Ali MM, Mutter R. Medicaid expansion and health insurance coverage and treatment utilization among individuals with a mental health condition. J Ment Health Policy Econ. 2020;23(3):151‐182. [PubMed] [Google Scholar]
26. Garfield RL, Zuvekas SH, Lave JR, Donohue JM. The impact of national health care reform on adults with severe mental disorders. AJP. 2011;168(5):486‐494. [DOI] [PubMed] [Google Scholar]
27. Hackbarth NS. Financing Integrated Care for Adults with Serious Mental Illness in Community Mental Health Centers: An Overview of Program Components, Funding Environments, and Financing Barriers. RAND Health; 2015. https://www.rand.org/content/dam/rand/pubs/working_papers/WR1000/WR1084/RAND_WR1084.pdf. Accessed November 2, 2022. [Google Scholar]
28. Frank RG, Goldman HH, Hogan M. Medicaid and mental health: be careful what you ask for. Health Aff (Millwood). 2003;22(1):101‐113. [DOI] [PubMed] [Google Scholar]
29. Beronio K, Glied S, Frank R. How the Affordable Care Act and Mental Health Parity and Addiction Equity Act greatly expand coverage of behavioral health care. J Behav Health Serv Res. 2014;41(4):410‐428. [DOI] [PubMed] [Google Scholar]
30. Breslau J, Stein BD, Yu H, Burns RM, Han B. Impacts of the dependent care expansion on the allocation of mental health care. Adm Policy Ment Health. 2019;46(1):82‐90. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Burns ME, Wolfe BL. The effects of the Affordable Care Act adult dependent coverage expansion on mental health. J Ment Health Policy Econ. 2016;19(1):3‐20. [PMC free article] [PubMed] [Google Scholar]
32. Brown B, O'Donnell E, Casalino LP. Private equity investment in behavioral health treatment centers. JAMA Psychiatry. 2020;77(3):229‐230. [DOI] [PubMed] [Google Scholar]
33. Bachhuber MA, Southern WN, Cunningham CO. Profiting and providing less care: comprehensive services at for‐profit, nonprofit, and public opioid treatment programs in the United States. Med Care. 2014;52(5):428‐434. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Beetham T, Saloner B, Gaye M, Wakeman SE, Frank RG, Barnett ML. Admission practices and cost of care for opioid use disorder at residential addiction treatment programs in the US. Health Affairs. 2021;40(2):317‐325. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Substance Abuse and Mental Health Services Administration . National Mental Health Services Survey (N‐MHSS): 2019. Data on Mental Health Treatment Facilities . Substance Abuse and Mental Health Services Administration; 2020. [Google Scholar]
36. National Mental Health Services Survey . U.S. Department of Health and Human Services, Substance Abuse and Mental Health Services Administration, Center for Behavioral Health Statistics and Quality; 2010. https://datafiles.samhsa.gov/. Accessed November 2, 2022.
37. Shields MC. Effects of the CMS’ public reporting program for inpatient psychiatric facilities on targeted and nontargeted safety: Differences between for‐profits and nonprofits. Med Care Res Rev. Published online March 12, 2021:1077558721998924. [DOI] [PMC free article] [PubMed] [Google Scholar]
38. Shields MC, Rosenthal MB. Quality of inpatient psychiatric care at VA, other government, nonprofit, and for‐profit hospitals: A comparison. Psychiatr Serv. 2017;68(3):225‐230. [DOI] [PubMed] [Google Scholar]
39. Braun RT, Yun H, Casalino LP, et al. Comparative performance of private equity‐owned US nursing homes during the COVID‐19 pandemic. JAMA Netw Open. 2020;3(10):e2026702. [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Wheeler JR, Fadel H, D'Aunno TA. Ownership and performance of outpatient substance abuse treatment centers. Am J Public Health. 1992;82(5):711‐718. [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Nahra TA, Alexander J, Pollack H. Influence of ownership on access in outpatient substance abuse treatment. J Subst Abuse Treat. 2009;36(4):355‐365. [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Wang V, Maciejewski ML. Patient outcomes and dialysis consolidation‐two big to fail? JAMA Netw Open. 2019;2(5):e193962. [DOI] [PubMed] [Google Scholar]
43. McHugo GJ, Drake RE, Whitley R, et al. Fidelity outcomes in the National Implementing Evidence‐Based Practices Project. Psychiatr Serv. 2007;58(10):1279‐1284. [DOI] [PubMed] [Google Scholar]
44. Drake RE, Bond GR, Goldman HH, Hogan MF, Karakus M. Individual placement and support services boost employment for people with serious mental illnesses, but funding is lacking. Health Aff (Millwood). 2016;35(6):1098‐1105. [DOI] [PubMed] [Google Scholar]
45. Mancini AD, Moser LL, Whitley R, et al. Assertive community treatment: facilitators and barriers to implementation in routine mental health settings. Psychiatr Serv. 2009;60(2):189‐195. 10.1176/ps.2009.60.2.189 [DOI] [PubMed] [Google Scholar]
46. Alternative Payment Model Framework and Progress Tracking (APM FPT) Work Group. Alternative Payment Model (APM) Framework: Final White Paper . Health Care Payment Learning & Action Network; 2016. https://www.milbank.org/wp‐content/uploads/2017/11/HCP‐LAN‐White‐Paper‐APM‐Framework.pdf. Accessed November 2, 2022. [Google Scholar]
47. Dixon L, McFarlane WR, Lefley H, et al. Evidence‐based practices for services to families of people with psychiatric disabilities. Psychiatr Serv. 2001;52(7):903‐910. [DOI] [PubMed] [Google Scholar]
48. Phillips SD, Burns BJ, Edgar ER, et al. Moving assertive community treatment into standard practice. Psychiatr Serv. 2001;52(6):771‐779. [DOI] [PubMed] [Google Scholar]
49. Mueser KT, Corrigan PW, Hilton DW, et al. Illness management and recovery: a review of the research. Psychiatr Serv. 2002;53(10):1272‐1284. [DOI] [PubMed] [Google Scholar]
50. Bond GR, Becker DR, Drake RE, et al. Implementing supported employment as an evidence‐based practice. Psychiatr Serv. 2001;52(3):313‐322. [DOI] [PubMed] [Google Scholar]
51. Rios‐Avila F, Sant'Anna P, Callaway B. CSDID: Stata module for the estimation of Difference‐In‐Difference modesl with multiple time periods. Published 2022. https://econpapers.repec.org/software/bocbocode/s458976.htm. Accessed November 2, 2022.
52. Callaway B, Sant'Anna PHC. Difference‐in‐differences with multiple time periods. Journal of Econometrics. 2021;225(2):200‐230. [Google Scholar]
53. Federal Communications Commission . Implementation of the National Suicide Hotline Improvement Act of 2018 . Federal Register; 2020:57767‐57783. https://www.federalregister.gov/documents/2020/09/16/2020‐16908/implementation‐of‐the‐national‐suicide‐hotline‐improvement‐act‐of‐2018. Accessed May 9, 2022.
54. Newton H, Beetham T, Busch SH. Association of Access to Crisis Intervention Teams with County Sociodemographic Characteristics and State Medicaid Policies and Its Implications for a New Mental Health Crisis Lifeline. JAMA Network Open. 2022; 5(7): e2224803. 10.1001/jamanetworkopen.2022.24803 [DOI] [PMC free article] [PubMed] [Google Scholar]
55. Andrilla CHA, Patterson DG, Garberson LA, Coulthard C, Larson EH. Geographic variation in the supply of selected behavioral health providers. Am J Prev Med. 2018;54(6 Suppl 3):S199‐S207. [DOI] [PubMed] [Google Scholar]
56. Gruttadoro D, Blair E, Kimball A, Usher L, Diehl S, Carlock H. First Episode Psychosis: A Guide to State Expansion. National Alliance on Mental Illness; 2017. https://www.nami.org/getattachment/Extranet/Advocacy/FEP‐State‐Advocacy‐Toolkit/FEP‐State‐Advocacy‐Guide.pdf. Accessed November 2, 2022. [Google Scholar]
57. Kaufman BG, Spivack BS, Stearns SC, Song PH, O'Brien EC. Impact of accountable care organizations on utilization, care, and outcomes: A systematic review. Medical care research and review : MCRR. Published online December 2017:1077558717745916‐1077558717745916. [DOI] [PubMed] [Google Scholar]
58. Busch SH, Tomaino M, Newton H, Meara E. Access to mental health support services in accountable care organizations: A national survey. Healthcare. 2022;10(1):100613. [DOI] [PMC free article] [PubMed] [Google Scholar]
59. Powers BW, Shrank WH, Navathe AS. Private equity and health care delivery: Value‐based payment as a guardrail? JAMA. 2021;326(10):907‐908. [DOI] [PubMed] [Google Scholar]
60. US Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation . Certified Community Behavioral Health Clinics Demonstration Program: Report to Congress, 2018 . Washington, D.C.; 2019. https://aspe.hhs.gov/reports/certified‐community‐behavioral‐health‐clinics‐demonstration‐program‐report‐congress‐2018. Accessed October 28, 2021. [Google Scholar]
61. US Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation . Certified Community Behavioral Health Clinics Demonstration Program: Report to Congress, 2019 . Office of the Assistant Secretary for Planning and Evaluation, US Department of Health and Human Services; 2020. https://aspe.hhs.gov/reports/certified‐community‐behavioral‐health‐clinics‐demonstration‐program‐report‐congress‐2019. Accessed October 28, 2021. [Google Scholar]
62. Soper MH, Matulis R, Menschner C. Moving towards value‐based payment for Medicaid behavioral health services. Center for Health Care Strategies; 2017. https://www.chcs.org/media/VBP‐BH‐Brief‐061917.pdf. Accessed November 2, 2022. [Google Scholar]
63. Frank RG, Glied SA, McGuire TG. Paying for early interventions in psychoses: A three‐part model. PS. 2015;66(7):677‐679. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials