Opting Out of Medicaid Expansion: Impact on Encounters With Behavioral Health Specialty Staff in Community Health Centers

This study used the 2012 Uniform Data System data set, which is maintained by the Health Resources and Services Administration. Health center grantee organizations report patient demographic data annually at the organization level, including insurance mix, the number of full-time–equivalent mental health and substance use disorder treatment specialty staff members, the costs of the services provided by these specialists, and the number of encounters they provided. These data do not include behavioral health services provided by primary care staff, case managers, and peers. Descriptive statistics were calculated for 2012, stratified by Medicaid expansion status in 2014.

Previous research indicates that patients remain loyal to health centers when they gain insurance (24). State Medicaid expansion decisions are likely to affect health center revenues because a higher percentage of patients will become insured in expansion states. Using the projected case mix by insurance status and total patient caseload and following the methodology reported by Ku and colleagues (3), we estimated the total potential health center revenue in 2020 that will not accrue to health centers if states continue to opt out of Medicaid expansion; we derived these estimates by multiplying the projected case mix by the payment gap for patients with each type of insurance. The payment gap was defined as the difference between the average cost and the average reimbursement per patient, divided by the average cost per patient. For marketplace plans, the payment gap was calculated as the average of Medicaid and private insurance estimates.

These figures were used to estimate the revenue that could be used to support the provision of behavioral health services delivered by behavioral health specialists on staff in health centers. Because health centers are not-for-profit organizations, they use revenues to cover the costs of providing services to patients. We assumed that in 2020, health centers will spend the same percentage of total revenues to cover services provided by mental health and substance abuse treatment specialty staff as they did in 2012. The percentage of total costs, including administrative overhead (described as “allocation of facility and nonclinical support services” in the Uniform Data System), that was allocated to encounters with behavioral health specialty staff stayed constant between 2009 and 2012. So, the 2012 percentages of total costs used to provide behavioral health specialist services were used to estimate the percentage of total costs allocated to behavioral health services in 2020: 4.9% for mental health specialist services and .7% for substance abuse treatment services. As noted in the discussion of limitations below, this assumption does not factor in the potential increase in demand for behavioral health services.

The number of encounters that might occur in opt-out states if these states expand Medicaid by 2020 was calculated by dividing the estimated revenue that would have been used to cover behavioral health services by the estimated cost per encounter with behavioral health specialty staff in 2020. The 2020 cost per encounter was estimated by dividing the total number of encounters reported by mental health and substance abuse treatment specialty staff in 2012 by the costs of providing these services, assuming a 4% annual growth rate. Based on the proportion of patients in each nonexpansion state in 2012, state-level estimates were produced by dividing the additional behavioral health encounters projected for 2020. Stata version 12.0 and Microsoft Excel 2010 were used for the analyses.

At baseline in 2012, health centers in nonexpansion states were more likely to be in the South and were less likely to be in the Northeast, West, or Midwest, reflecting the characteristics of the states that did not expand Medicaid for the 2014 plan year (Table 1). Health centers in nonexpansion states were more likely to be located in rural areas; 57.6% of health centers in nonexpansion states are in rural areas, compared with 39.2% of health centers in states that expanded Medicaid. Health centers in states that did not expand Medicaid were smaller and less likely to have quality accreditation or recognition as a patient-centered medical home in 2012, compared with health centers in states that expanded Medicaid by 2014.

A higher percentage of health center patients in nonexpansion states were uninsured at baseline (42.9%, versus 34.8% in expansion states). In addition, a lower proportion of health center patients were covered by Medicaid in nonexpansion states in 2012, prior to ACA implementation. Slightly over one-third of patients in expansion states were covered by Medicaid (38.6%), compared with 29.7% in states that opted out of expanding by 2012. Differences in the race-ethnicity mix of health center patient caseloads were also observed, with a higher percentage of Hispanic patients and a lower percentage of black patients in health centers located in states that expanded Medicaid by 2014 compared with states that did not.

At baseline in 2012, health centers in states that opted out of expanding Medicaid in 2014 were disproportionately rural, located in the South, smaller, and less likely to have attained accreditation for quality of care or recognition as a patient-centered medical home. Health centers in opt-out states also had more uninsured patients and lower Medicaid caseloads compared with health centers in states that expanded Medicaid by 2014. These current disparities in uninsured and Medicaid case mix will only be exacerbated. Our findings highlight the impact of insurance case mix on the provision of nonrequired services in health centers, such as mental health and substance use disorder services. If all states expand Medicaid by 2020, we estimate that nearly $230 million in additional revenue could accrue to health centers in states that are currently opting out. If health centers are able to hire additional staff, this revenue could cover the costs of over 70,000 encounters with mental health and substance use disorder specialists, in addition to general medical, dental, and other services.

This study indicates that expanding Medicaid would likely result in higher utilization of behavioral health services in health centers, highlighting an important indirect benefit of Medicaid expansion. The estimates in this study are consistent with previous research demonstrating that health centers’ capacity for the provision of nonrequired services, such as mental health and substance abuse treatment services, is particularly vulnerable to fluctuations in health center revenues (22). This increase in health center capacity is in addition to the positive impacts on access to care among patients who would become newly eligible for Medicaid coverage (25).

One implication of these findings is the critical importance of reducing uninsurance among health center patients by enrolling as many eligible individuals as possible. Outreach and enrollment activities are affected by state policies, so reducing the hurdles to effective outreach activities is critical (26). Also, streamlining enrollment requirements to enroll those who are eligible for Medicaid would be helpful, because at baseline, only 61.7% of eligible individuals were enrolled in Medicaid (27).

These findings also highlight the importance of team-based care, given the limited capacity for patient encounters with behavioral health specialists in health centers. Training for teamwork and for the use of collaborative and integrated models is needed to ensure that the behavioral health specialists practicing in health centers are supporting primary care providers as they care for patients with behavioral health needs (28). The most recent funding opportunity to expand mental health and substance use disorder treatment capacity specifically requires the provision of integrated care (29). It is important to note that, even as on-site behavioral health capacity in health centers increases, strong linkages to community-based behavioral health services are needed, particularly for services such as detoxification, residential treatment, and intensive outpatient treatment. Previous research shows that unmet need for mental health services is more likely to be found among patients with more severe mental illness, highlighting the need to ensure that health center patients have access to the full continuum of services (11).

Several limitations of the study should be noted, given that these results are sensitive to the assumptions used to generate the estimates. In particular, these analyses did not account for the potentially increased demand for behavioral treatment services because of expanded insurance coverage for these services. Health centers are required to be responsive to the service mix demands of health center users, by conducting periodic needs assessments and having patient-majority governing boards. Therefore, increased demand for behavioral health services among health center patients, as a result of increasing insurance coverage for behavioral health services and less cost sharing, might drive increased spending on behavioral health services as a percentage of total health center spending (30,31). Therefore, our estimates of the number of behavioral health encounters that could occur in 2020 if all states expand Medicaid by then might be conservative. Also, this study examined only encounters with behavioral health specialists, but some behavioral health services are likely to be provided by primary care staff. Information on behavioral health services provided by primary care staff, case managers, and peers is not available in the Uniform Data System; increasingly, behavioral health services are delivered by nonspecialists, so this is a critical shortcoming of the data set. Finally, there is a lag between revenue receipt and the ability to hire a specialist, so these projections might overestimate the additional number of encounters for patients with behavioral disorders that might occur in 2020 if all states expand Medicaid by then.

This study focused on the provision of behavioral health care by mental health and substance abuse treatment specialists in health centers. Future research should monitor the indirect impact of state Medicaid expansion decisions on other types of capacity and utilization within health centers, as well as throughout the health care delivery system. The impact on community mental health centers, rural health clinics, hospitals, and other safety net providers is especially critical to understand. Future research should also aim to bolster the evidence base for team-based care, so that behavioral health specialty staff can be most effectively utilized in health centers and other settings.

Addressing patients’ behavioral health needs is essential to care delivery transformation, and health centers are critical access points for medical and behavioral health services in underserved communities. As not-for-profit organizations with a mission and statutory mandate to increase access for underserved communities and populations, health centers use revenues to cover the costs of service provision. Given the limited number of behavioral health providers willing to accept patients who are uninsured or covered by Medicaid, health centers will remain a primary source of behavioral health treatment services for underserved populations. There is a clear need for on-site behavioral health services at most health centers, but financial constraints might limit the ability of health centers to expand their scope of services, particularly in states that opt out of expanding Medicaid.

The authors acknowledge Richard G. Frank, Ph.D., for reviewing the manuscript and providing insightful comments that shaped subsequent drafts. The authors are also grateful for the feedback from Kirsten Beronio, J.D., Ruth Katz, M.A., and Linda Elam, Ph.D., M.P.H.