Earlier this month, the School of Government’s Criminal Justice Innovation Lab released our final report evaluating the Burlington Police Department’s mental health co-responder program. This report is the culmination of a two-year partnership between the School, the Burlington Police Department, RHA Health Services, and the North Carolina Association of Chiefs of Police. The evaluation speaks to the impact of the co-responder program on the police department, the clients served, and the broader community. We also included recommendations on data collection for other agencies that may be interested to conduct similar evaluations. This post summarizes the findings and the recommendations.
About the Program
The Burlington Police Department established their Law Enforcement Crisis Counselor (LECC) program in 2015. The crisis counselor, a trained mental health professional who is embedded in the police department, has three primary functions:
- Co-responding on scene to calls related to behavioral health crises and other social issues. After responding to a call, determining that an individual is in crisis, and securing the scene, an officer may request the crisis counselor co-respond to the call.
- Providing real-time consultation to officers remotely. The crisis counselor may be called upon to assist officers remotely if the situation does not require an in-person co-response. For example, an officer might call for consultation if they are interacting with someone they know is currently connected with the crisis counselor or has been in the past
- Connecting individuals to services in response to referrals. Officers often make after-the-fact referrals to the crisis counselor if they interact with an individual who they believe could use additional support. Once the crisis counselor receives a referral, they will follow up with that individual and work to connect them to a variety of services, like behavioral health treatment, housing, and food assistance.
Key Findings
The report divides findings into four sections: program scope, program outcomes, criminal justice outcomes, and program satisfaction. I share some of the key findings from each section below. Our analyses were based on 911 dispatch data, case management data from the crisis counselor, jail data, court records data, surveys, and interviews.
Program Scope
- Proportionally, a relatively small number of calls for service are eligible for crisis counselor intervention. From 2013 to 2024, 4.74% (25,774) of all calls for service received by the Burlington Police Department were potentially eligible for LECC intervention, meaning they included a behavioral health component that fell within the scope of the program. Importantly, we learned from the crisis counselor that not all calls that might fall under a “behavioral health” umbrella are appropriate for the LECC. For example, the counselor rarely responded to overdose calls. We took these program-specific considerations into account when designing our analyses.
- A minority of clients have repeat interactions. On average, the crisis counselor served over 300 unique individuals per year. 17.88% of clients across all years had a repeat interaction with the crisis counselor.
Program Outcomes
- Calls for service with a co-response almost never end in arrest. Since 2021, only one call involving an on-scene co-response ended in arrest. Most calls with a co-response are resolved at the scene.
- Calls for service with a co-response are less likely to result in transport to the emergency room. Our analyses showed that behavioral health calls without a co-response are more likely to result in emergency room transport (26.2%) versus behavioral health calls that have a co-response (18.75%).
- Engaging a crisis counselor does not save officer time on-scene, but time savings may be reflected in other ways. On average, calls with a co-response took 21 minutes longer than other behavioral health calls and 47 minutes longer than non-behavioral health calls. This finding shows that co-responses do not save officer time directly at the scene of a call, but time savings from co-responses may be reflected in other ways like reducing the amount of time officers spend with clients at the hospital. In surveys, some officers reported that they strongly believed the LECC program saved them time, while others disagreed.
Criminal Justice Outcomes
- The crisis counselor intervention is seemingly not enough to divert clients from future involvement in the criminal justice system. Criminal justice analyses show that a minority of LECC clients have any history of criminal court involvement (28.9%) or incarceration in the local jail (16.4%). However, those clients that do have a history of justice system involvement continue to have future interactions with the justice system. We saw that clients continue to acquire new criminal charges and be booked into jail at similar rates before versus after their LECC interaction.
Program Satisfaction
- Stakeholders across the board are highly satisfied with the program. Based on survey and interview data, the police department, community partners, and clients are all highly satisfied with the LECC program. Stakeholders attribute much of the success of the program to the specific embedded model used in Burlington, where the counselor is housed in the department, participates in department activities, and builds trust and rapport with the officers.
Recommendations
On page thirty-two of the report, we present four data-related recommendations for other jurisdictions that may be interested to conduct similar evaluations of alternative responder programs. I include a summary of those recommendations here, but you can find additional detail in the report.
- Put procedures in place for robust data collection from the start. One notable limitation of this evaluation is the fact that the LECC program began in 2015, but case management data was only available starting in 2021. Other communities that are interested in implementing alternative responder programs may benefit from having standardized data collection procedures in place from the start. This could include investing in case management software or simply creating an Excel spreadsheet with information on client interactions and outcomes. In Appendix A of the report, we provide a detailed list of data points that agencies might want to track.
- Create common definitions between alternative response data and other data sources. When developing data collection procedures for an alternative responder program, agencies may benefit from first reviewing the data collection systems that are already in place (i.e. Computer-Aided Dispatch data, Record Management Systems). When possible, ensure that the terminology and definitions used for the alternative responder program align with the existing definitions in current data collection.
- Use unique identifiers to tell the full story of a call for service. For each call involving an alternative response, it may be beneficial for stakeholders to think about data collection as capturing the who, what, when, where, and why of each event. To tell the full story, you typically have to combine multiple data sources. For example, 911 call data can tell you when, where, and why a call took place. But that data usually cannot tell you reliable information about the subject of the call or what services were provided at the scene. To connect different data sources, consider including a unique event identifier in every piece of documentation related to calls receiving an alternative response.
- Consider additional training and collaboration with telecommunications to identify calls eligible for alternative response. In real time and with limited information, it can be difficult for telecommunicators to differentiate between behavioral health concerns and a different type of emergency. Additional training and consultation may be beneficial to help telecommunicators identify and flag when a call may involve a behavioral health crisis.
If you have any questions about the evaluation, please feel free to reach out to me at hturner@sog.unc.edu. We have additional resources on alternative response available on our website.
And, if your jurisdiction is looking to implement a new alternative responder program, evaluate an existing program, or discuss data collection practices, we would love to connect! Our team is here to help.