Countywide Implementation of CIT: Multiple Methods and Alternative Outcomes

Background: Prevention of incarceration for persons with serious mental illness (SMI) has been a long-standing goal of both jail and mental health (MH) professionals. Although intervention models span the continuum from pre-arrest to prison reentry, pre-arrest diversion via the Critical Intervention Team (CIT) model is in high demand. CIT uses a combination of collaborative problem solving advisory boards, 40 hour officer training, and enhanced MH resources that allow 24 hour access.  While there is great demand for this approach, the desired outcomes of decreased arrests have eluded most evaluation studies.  In an effort to expand measureable outcomes and elucidate the effect of CIT, researchers triangulate their methods and data sources to ask the following questions: Does CIT training 1) enhance skills and knowledge 2) increase utilization of community MH resources; 3) decrease utilization of hospital ER; and 4) change officer perceptions?

Methods:  This County of 1.23 million conducted three CIT training sessions mid-year 2015, involving 79 officers (24% of the countywide sheriff force). Multiple methods included standardized instruments for pre/post-test training measurement; semi-structured interviews with a purposive sample of 15% of those trained; and use of administrative data in the form of sheriff records (776 reports flagged for MH in 2015) and two years of crisis center data detailing 719 drop offs.  Analyses used paired t-tests for pre/post training differences and interrupted time series (ITS) to assess change over time related to sheriff actions and crisis center drop offs. Inductive analysis was for interviews.

Results: Training significantly improved officer de-escalation skills (t(66)=4.199, p<.001) and knowledge of community resources (t(66)=8.722, p<.001). Outcomes did not differ by gender, education level, or years in law enforcement. There was a two fold increase in crisis center utilization when comparing 2014 and 2015 (263 v 456);  ITS demonstrated that use was 38.46 times higher than what was projected in the month after training (p<.001) and this effect remained 6 months later when use was 27.35 times higher than what would be projected (p<.001). Although ER drop offs appeared to gradually slope downward over the 1 year period, there was no significant difference between projected and actual rates. Officers’ confirm that CIT challenged and changed perceptions of MH and expanded their knowledge of community resources; they commend the training and advisory board with the streamlined process of drop off at the crisis center as compared with the ER.

Conclusions/Implications:  Outcomes indicate that CIT is successful in enhancing law enforcement knowledge and changing practice. Utilization of a less costly and more appropriate community resource (crisis center) demonstrates that the training took effect right away and was sustained, with no decay in the training effect over the 7-month post-training period. Although there was no change in ER drop offs, a single year of data may not be long enough to detect change, or there may be differing needs. Social workers employed in MH settings will need to creatively work with local law enforcement to enhance alternatives to incarceration for those with SMI, using alternative measures of success.