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

Abstract

Importance: The mental health crisis lifeline 988 will begin operating July 16, 2022. In the absence of appropriately trained first responders, including crisis intervention teams (CITs), persons experiencing behavioral health crises face the risk of incarceration and even death.

Objective: To assess county-level access to CIT in 2015 and 2020 and its association with area characteristics and state policies in 2020.

Design, setting, and participants: This cross-sectional study included 10 430 facilities from the 2015 National Directory of Mental Health Treatment Facilities and 10 591 facilities from the 2020 National Directory of Mental Health Treatment Facilities, attributed to 3142 US counties.

Exposures: Area measures included need (suicide, drug-related overdose mortality), rurality, and demographic characteristics. State-level policies included 5 Medicaid policies enacted prior to 2020 and 2 recent policies intended to assist implementation of the 988 lifeline.

Main outcomes and measures: Whether there was at least 1 facility that reported offering a CIT that handled acute mental health issues at the facility or off-site for each county in 2015 and, separately, in 2020.

Results: Most US residents (88%) resided in a county with at least 1 facility offering CIT, although half of US counties had no facility offering CIT (2015: 1537 of 3142 [49%]; 2020: 1512 [48%]). Almost 1 in 5 counties, representing 9% of the population, experienced a change in county-level access from 2015 to 2020. Unadjusted analyses indicated residents of counties without vs with CIT access were more likely to be older and uninsured (top quartile of percentage of residents aged >55 years: 502 of 1512 [33%] vs 283 of 1630 [17%]; P < .001; top quartile of percentage of residents uninsured: 500 [33%] vs 285 [17%]; P < .001) and were more likely be rural (frontier: 500 [33%] vs 144 [9%]; P < .001). Similar results, excluding counties in the top quartile of residents aged older than 55 years, were found in adjusted analyses. Counties without vs with CIT access were less likely to be in states that expanded Medicaid (788 [52%] vs 1102 [68%]; P = .01) and in states that allow Medicaid to pay for short-term stays in psychiatric hospitals (34 [2%] vs 73 [4%]; P = .02). Other Medicaid-related associations were not statistically significant in adjusted analyses.

Conclusions and relevance: In this study, most US residents lived in counties with access to at least 1 CIT, but fewer than half of US counties had such access. Policies to encourage facilities in rural counties to offer CIT may help realize the potential of the new lifeline.

Figure 1.. Changes in County-Level Access to Crisis Intervention Teams (CITs)

The 2015 and 2020 National Directory of Mental Health Treatment Facilities included information from 10 430 and 10 591 facilities, respectively. This figure uses responses from the 5050 and 5238 facilities listed in 2015 and 2020 National Mental Health Treatment Directories who responded yes to the question “Does this facility offer a crisis intervention team that handles acute mental health issues at this facility and/or off-site?” For each US county, we determined whether there was at least 1 facility that answered yes to this question. Half of counties had no facility offering CIT: 1537 of 3141 (49%) in 2015 and 1512 (48%) in 2020. Data from 2015 and 2020 was used to categorize counties into 1 of 4 categories: had no access to CIT in either year, lost access, gained access, and had access in both years. The share of the 2020 population that resided in each category of counties was then calculated.

Figure 2.. Adjusted Associations Between Area Characteristics and County-Level Access to Crisis Intervention Teams (CITs) in 2020

This figure shows the change in estimated probability and 95% CIs from a logistic regression model estimating county-level CIT in 2020 using all covariates listed in the Table, excluding measures of need. Logistic regression models were used to compute the adjusted associations between area and state Medicaid policy characteristics and adjusted SEs to account for correlation within state. All measures except rurality are binary and represent counties in the top quartile of that measure.

Figure 3.. Adjusted Associations Between State Medicaid Policies and County-Level Access to Crisis Intervention Teams (CITs) in 2020

This figure shows the change in estimated probability and CIs from a logistic regression model estimating county-level CIT access in 2020 using all covariates listed in the Table, excluding measures of need. Logistic regression models were used to compute the adjusted associations between area and state Medicaid policy characteristics and adjusted SEs to account for correlation within state. All measures are binary; federal grants and payments categories represent the counties in the top quartile of that category. CCBHC indicates Certified Community Behavioral Health Clinic; DSH, disproportionate share hospital; IMD, institutes of mental disease; and SAMHSA, Substance Abuse and Mental Health Services Administration.