Medicaid Data Show Wide Differences in Mental Health Care in the United States
• Research Highlight
The Medicaid program, administered by the Centers for Medicare and Medicaid Services, provides health coverage to millions of Americans, including eligible low-income adults, children, pregnant women, elderly adults, and people with disabilities. As the program is a state-federal partnership, its implementation can vary widely across states. A new study supported by the National Institute of Mental Health revealed wide differences in rates of mental health care among Medicaid enrollees based on where they live in the United States.
K. John McConnell, Ph.D., and colleagues at Oregon Health and Science University examined data from millions of Americans aged 19–64 years enrolled in Medicaid in 2018. The study represents the first extensive use of a new database of national Medicaid claims, known as the Transformed Medicaid Statistical Information System Analytic Files, designed to provide reliable, deidentified data for research purposes.
The database included data from 42 states and Washington, D.C., after excluding states with excessive missing or incomplete data. The researchers divided states to create 393 economic areas—groupings of counties within states of at least 100,000 people and with similar socioeconomic features. For each economic area, the researchers calculated the number of emergency department visits for mental health conditions and outpatient mental health visits. They used emergency department visits as a proxy for access to and quality of outpatient mental health care, assuming that people would turn to emergency rooms when they could not access or faced a shortage of outpatient mental health services.
Medicaid claims showed variations in rates of emergency department visits for mental health conditions between states and between economic areas within states. For instance, the state with the most emergency department visits had almost five times more visits than the state with the fewest visits. Similarly, within many states, there was a wide range between the economic areas with the highest and lowest number of emergency department visits.
There was also large geographic variation in outpatient mental health visits, which were positively, but only moderately, correlated with emergency department visits. In general, economic areas with high rates of emergency department visits also had high rates of outpatient care. However, there were also areas with high rates of emergency department visits but low outpatient mental health visits, and vice versa. Frequent use of emergency services for mental health treatment might reflect a high degree of need in those areas and, in the absence of available outpatient care, a high degree of unmet need.
As a final step, the researchers compared emergency department visits for certain types of mental health conditions:
- Anxiety disorders compared to schizophrenia and other psychotic disorders
- Depressive disorders compared to suicidal ideation and intentional self-harm
Emergency department visits differed for each condition. For instance, economic areas composed of large cities had higher rates of emergency department visits for schizophrenia but relatively few visits for anxiety. In contrast, smaller cities tended to have more visits for anxiety. Even for depression and suicidal ideation, which are clinically linked conditions, rates of emergency department visits were not the same across areas. These results show that combining different mental health conditions can mask critical differences in in the number or type of services Medicaid enrollees receive for certain conditions.
Together, the findings highlight the nationwide use of emergency departments for mental health care, while emphasizing wide variations in rates of use between states and between mental disorders. The frequent use of emergency services for mental health care shown in this study might, in some cases, indicate a high degree of unmet need or a lack of access to outpatient mental health services.
Regional differences in Medicaid recipients’ mental health care experiences also demonstrate the importance of tailoring strategies to specific populations. The authors emphasize the need for context-specific, local solutions, which might start with looking at how mental health benefits are administered by state Medicaid agencies, determining the availability of local mental health providers, or identifying the most common types of mental disorders in communities.
Because this is a new database, there are several considerations to note. First, eight states were not included in the analyses due to missing or incomplete data. Second, the database provides limited information on disparities in care received by people of different races and ethnicities. Third, a subset of the Medicaid population who may be especially likely to face barriers to care—people dually eligible for Medicaid and Medicare—were excluded from the analyses. Last, regional differences in coding mental health diagnoses or submitting mental health claims may have led to underreporting in some areas. As these Medicaid files become more widely used, improvements in data quality and addition of more common data elements may help overcome these limitations.
Initial analyses using this new database demonstrate the value of large national data to understand trends in diagnosis, service use, and mental health. This study also identified limitations that state administrators, data analysts, and others might want to address. Although these files are more comprehensive and reliable than prior Medicaid claims data, further research can provide additional insight on their quality and consistency. Future research and policy implementation can capitalize on nationally representative data to enhance the quality and value of Medicaid mental health programs and thereby help improve mental health outcomes for millions of Americans nationwide.
McConnell, K. J., Watson, K., Choo, E., & Zhu, J. M. (2023). Geographical variations in emergency department visits for mental health conditions for Medicaid beneficiaries. Health Affairs, 42(2), 172–181. https://doi.org/10.1377/hlthaff.2022.00796