Using GIS to Assess MH Hot Spots and Police Utilization of Crisis Centers: The Impact of Crisis Intervention Teams Training

Background: Preventing incarceration of persons with serious mental illness (SMI) requires a community response. Crisis Intervention Team (CIT) facilitates such a response with a 40-hour police officer training, community advisory boards, and development of a 24-hour mental health (MH) crisis center.  CIT research has shown increases in officer self-efficacy and enhanced recognition of SMI, as well as additional transports to hospital ER. However, ER can be expensive and lack linkage to long-term MH services.  Moreover, response might be better targeted if MH and law enforcement professionals knew where MH ‘hotspots’ were located or analyzed whether proximity to services impacted officer decision-making. The current study asks the following questions: 1) Where do MH crises cluster and does call location influence drop-off location? 2) After CIT implementation, do officers increase the drop-offs to the crisis center and decrease drop-offs to ER?

Methods: A Midwestern county of 1.23 million people with a 24-hour MH crisis center implemented CIT officer training in 2015. All MH and suicide call reports from the county sheriff’s officers from 2014 through 2016 (N=1,617) were coded. Call reports included call time/date, drop-off location, demographics, necessity of physical health care, and if the responding officers were trained in CIT. Interrupted time series (ITS) was used to examine officer drop-offs (crisis center or hospital) pre and post CIT training. QGIS (Quantum Geographic Information System) was used to identify regions with high call volumes (hot spots) and calculated differences in proximity to the crisis center from the call location. Logistic regression (LR) was used to assess predictors associated with crisis center drop-offs. Predictor variables included: age, gender, race, type of call, and differences in distance.

Results: QGIS illustrates that 50% of all calls occurred in 9% of the department’s jurisdiction (39% in just 3%) demonstrating distinct ‘hot spots’. ITS revealed an increase in drop-offs to the crisis center in the month after training (β=10.0, p<.003), which was sustained 5 months afterward (β=9.7, p<.016). Decreases in ER drop-offs immediately following training (β= -11.2, p<.037), were also sustained 5 months later (β= -13.5, p<.036). The LR predicted crisis center drop-offs post CIT training (χ²(8,N=723)=98.264), successfully classifying 73% of the calls and accounting for 18% of the  variance. After controlling for other variables, CIT officers were 2.5 more likely (Wald=20.736, p<.001) to transport to the crisis center than non-CIT officers. Finally, for every one-mile increase in the distance between the call location and the crisis center, officers were .96 times less likely (Wald=6.290, p<.05) to take the individual to the crisis center.

Conclusions/Implications: Evidence suggests CIT training, in conjunction with a 24-hour MH drop-off location, leads to changes in a community’s response to MH crises.  Instead of the short-term, high-cost, revolving door of a hospital ER, CIT’s referrals to potentially long-term, low-cost, MH treatment may lead toward a more financially sustainable and socially just solution for people living with SMI.  The spatial clustering of MH calls could encourage data-driven deployment of MH professionals and CIT officers to targeted hot spots to support future problem-solving.