Benefits and Drawbacks of Police Integration Into Assertive Community Treatment Teams

Individuals with severe mental illness are at elevated risk for criminal justice involvement but are more often victims than perpetrators of criminal behavior (1). The criminalization hypothesis posits that, because of a lack of mental health resources, those unable to access such supports are more likely to engage in behavior that results in justice system involvement (2). Indeed, mental illness is two to three times more prevalent within Canadian prisons than in the general population (3). Increased arrest rates for people with mental illness may be due to factors that also increase risk for criminal involvement (e.g., unemployment, low income, or substance use) (4). Police encounters for minor offenses involving people with mental illness are also more likely to lead to arrest than those involving people without mental illness (5).

Since deinstitutionalization, a lack of effective behavioral health services has meant that police are often responsible for addressing mental health issues, which is time consuming and outside of their competency and which can cause significant negative impacts (6). Cross-system collaborations between mental health and police services are hence on the rise (7). In response, concerns have been raised that the integration of mental health and criminal justice may cause further entrenchment in the criminal justice system and further stigmatize individuals as dangerous.

Assertive community treatment (ACT) consists of integrated teams of professionals providing intensive, tailored outreach to individuals with severe mental illness to support community living, psychosocial rehabilitation, and recovery. The first effort to integrate police into ACT was the forensic ACT (FACT) model, which arose to address the needs of individuals with severe mental illness and criminal justice involvement in an integrated manner (Box 1) (8–12). Some FACT staff feel that including a criminal justice partner on the team provides useful legal leverage to motivate clients who have previous arrests and incarcerations to engage in treatment, avoid arrest, and abstain from substance use, whereas other staff feel that this integration problematically blurs the distinction between health care and criminal justice activity (13). This tension may be due to different goals of health care (i.e., client health and well-being) and public safety (i.e., community protection) (14). Similarly, whereas some clients feel that legal leverage leads to problematic coercion to enter treatment, others feel that FACT is a better alternative to a continuing cycle of arrests, substance use, homelessness, and trauma (15–17).

An alternative to the FACT model is ACT-PI: police integrated into ACT teams on a full-time basis (18). Instead of segregating clients with a history of criminal justice involvement, ACT-PI integrates police officers into existing ACT teams, simultaneously serving clients both with and without forensic involvement. Unlike FACT, ACT-PI teams do not solely serve those who have perpetrated crime, but also those who have been victimized, as well as those who have little criminal justice involvement. ACT-PI teams were created to reduce the risk for stigma associated with FACT involvement, to provide better access to police for clients without a history of criminal justice involvement but who may need police services (e.g., to report experiences of victimization), and to distribute the caseload of clients coping with behavioral dysregulation across teams to reduce staff burnout. The ACT-PI model allows staff to systematically target underlying social and economic factors that affect all individuals with mental illness regardless of criminal justice involvement. See online supplemental material for a description of the police officers' role on the ACT team.

To date, very few ACT programs utilize an ACT-PI model, and no empirical evaluations of this approach exist. The purpose of this study was to conduct the first qualitative evaluation of the ACT-PI model. Using qualitative interviews with both clients and staff from the four ACT-PI teams in Victoria, British Columbia, we examined ACT-PI client and staff perspectives regarding the involvement of police officers in ACT-PI teams.

Methods

ACT-PI Officers

Police officers on ACT-PI are integrated, full-time team members who are physically present and accessible to staff and clients. They attend team meetings and share information with the team (e.g., feedback on how clients are doing in the community based on police contacts overnight) as part of prevention-focused care plans. Officers provide education and support to staff and clients regarding choices and consequences from the judicial system’s point of view. The officers are also involved in the care of clients without judicial system involvement, supporting staff (e.g., attending outreach appointments to housing units considered unsafe by the Health Authority) and clients (e.g., managing unsafe situations such as when a client is uncomfortable with someone in their home). At the time of this research, three officers (two women and one man) were embedded full-time with the four local ACT teams.

Participants

ACT-PI clients.

We interviewed 21 current ACT-PI clients in September and October 2017. To preserve anonymity of ACT clients, build trust, and encourage honest responding, we did not systemically collect extensive demographic information. However, clients frequently mentioned relevant demographic information during their interviews, and Box 2 shows selected characteristics gleaned from these statements. Nearly all clients (approximately 90%) reported at least some casual contact with ACT-PI officers (e.g., going for coffee or riding in the front seat of patrol car), and several reported extensive, repeated contacts. The few clients who had not had direct contact with the officers had knowledge about them through their acquaintances or had expectations based on their own beliefs and life histories.

ACT-PI staff.

We interviewed 22 ACT-PI staff members in January and February 2018, with representation from all four teams. Most (N=16, 73%) staff participants were female, and half had worked at ACT-PI for >5 years. The sample represented many professions, including nursing, social work, psychiatry, addictions, mental health, and psychiatric rehabilitation.

Recruitment

ACT-PI clients.

All ACT-PI clients were eligible to participate. We held drop-in times during which ACT-PI clients could participate at their convenience to encourage as broad a cross-section of participation as possible. These drop-ins were advertised on flyers posted in ACT-PI offices, as well as in the community, and took place in social service and health agencies familiar to the clients. Clients participated in an individual confidential interview and were given a $10 gift card.

ACT-PI staff.

ACT-PI team leaders sent e-mail invitations to participate to staff members. Interested staff were directed to contact the research team. ACT-PI employers were not aware of who participated. Staff who agreed to participate were interviewed individually in a confidential space at the ACT-PI office. No compensation was provided.

Procedure

The university and health authority research ethics boards of Island Health Authority and the University of Victoria approved the research design. Participants were provided with information regarding the purpose and goals of the study and gave informed consent regarding their rights as research participants. Interviewers discussed how confidentiality would be protected and emphasized the researchers’ independent role from the health authority and police department. Interviews were audiotaped and transcribed verbatim. The first two authors (C.L.C., E.M.W.) conducted all interviews. ACT-PI staff were recruited from a larger pool of approximately 60 staff to ensure representation from each ACT-PI team as well as from each professional designation. ACT-PI clients were drawn from a larger pool of approximately 320 current clients. ACT-PI clients and staff continued to be interviewed until data saturation was reached (19).

Measures

Interview questions (10 for staff, nine for clients) were semistructured and open ended. For example, both clients and staff were asked, “From your perspective, what are some of the key benefits and drawbacks of having a police officer involved with your ACT team?” Client-specific questions included, “Is your relationship with your ACT officer different than any interactions you’ve had with other police officers?” Staff-specific questions included, “Does police involvement on the ACT team affect the services that you are able to provide to your clients?” Interview questions were created by the first two authors in consultation with the local health authority, police department, and representatives of the local ACT-PI teams. Interviews with clients generally lasted 15 minutes, and interviews with staff lasted 30–60 minutes.

Data Analysis

Transcripts were analyzed by the first two authors to capture the primary themes from both participant groups (20). Descriptive codes were applied to small portions of data—a few sentences of text or less (21). Individual codes were then thematically grouped, and subcategories were developed on the basis of shared content. The other two authors (K.N.D., R.F.) independently reviewed the themes for accuracy and consistency. Feedback from these two researchers formed the basis of a fuller group discussion, in which all researchers worked collaboratively to refine the conceptual frame guiding the coding scheme. The original coders then reviewed the previously coded transcripts in light of the revised coding scheme and made adjustments.

Coders liberally used reflexive coding memos to track emerging patterns, ideas, and questions and to maintain discussion about the data as the analyses developed. To further ensure trustworthiness of the analyses, the first two authors reviewed the themes and their descriptions with a steering committee (comprising health care and police representatives) and an advisory committee of individuals who had lived experience with the ACT-PI program (currently or in the past). Family members and social service providers (e.g., supportive housing) who had extensive contact with the program were also included in this review process. The advisory board agreed with our themes and often elaborated on them.

Results

Results are organized into benefits and drawbacks of police integration into the ACT-PI program. An online supplement presents further information about each theme.

Benefits

Most clients (N=17, 81%) and staff (N=20, 91%) identified benefits of officer involvement. Five themes emerged across client and staff responses (Table 1).

Relationship building.

Staff and clients consistently highlighted as an essential benefit the supportive relationships that ACT-PI officers form with clients. Officers are seen as compassionate, respectful, and helpful and as establishing good rapport. The interactions that clients had with ACT-PI officers were described as more empathic, more humane, and less stigmatizing than relationships with regular members of the police service. ACT-PI officers took time to get to know the clients personally (e.g., asking how they were or taking them out for a cup of coffee). These relationships also provided practical support for both victims and perpetrators of crime (e.g., assistance in obtaining a restraining order against a violent partner) because ACT-PI officers knew the history of clients well enough to provide concrete crisis responses. Building trust was acknowledged to take considerable time for clients with a history of negative interactions with police but was viewed as quite helpful. Staff highlighted that these relationships could provide a healing role for clients because they experienced officer support and advocacy, rather than exclusively enforcement.

Improving safety.

Both clients and staff frequently mentioned physical and emotional safety as benefits of ACT-PI officer involvement. Often this involved safety for both clients and staff in common office areas. Several clients remarked that this improved sense of safety increased the likelihood that they would access services at ACT-PI offices. ACT-PI officers also increased clients’ sense of safety by intervening to protect them from victimization or exploitation by dangerous individuals in their community. Finally, nearly all staff mentioned that ACT-PI officers enhanced staff safety, enabling the staff to deliver more effective services in high-risk situations such as clients experiencing agitation or aggression, as well as in housing situations deemed unsafe for staff to enter alone because of dangerous activity.

Embeddedness.

Both clients and staff remarked that ACT-PI officers were viewed as integral, trusted team partners whose involvement resulted in better services to clients. Communication was seen as a two-way street—policing information to the teams and health care information to the officers—resulting in more holistic care. With health care knowledge, ACT-PI officers could communicate with patrol officers regarding clients’ mental status and violence risk, leading to improved law enforcement responses. Similarly, teams could provide better care with ready access to specialized police knowledge (e.g., nuanced knowledge of arrests, warrants, and protection orders). ACT-PI officers helped interpret police reports, helped navigate the courts, and shared unique knowledge with staff (e.g., drug education). ACT-PI officers were also points of connection with other service providers, including supportive housing, and used their authority in the community to facilitate resources that other team members might not have the power to procure on their own (e.g., persuading social service agencies to prioritize ACT-PI clients for faster housing referrals).

Prevention oriented.

Clients and staff discussed the ability of ACT-PI officers to focus on prevention, although these two groups conceptualized this differently. Clients described officers as identifying and preventing problems before they escalated, instead of reacting to late-stage decompensation. Clients also described ACT-PI officers as being more likely than other police to help clients stay in their homes during periods of psychiatric distress. Maintaining housing and a general sense of dignity were seen by many clients as key to long-term well-being.

Staff discussed prevention in the context of police responses to emergencies. ACT-PI officers could respond to questions and requests faster than other officers, and teams could get clients treatment more quickly in emergency situations. Furthermore, the accessibility created predictability, because ACT-PI teams could manage their time well and plan proactively, rather than waiting for a crisis. For example, apprehending clients and bringing them to the hospital were described by staff to be quicker, more predictable, and less traumatic with ACT-PI officers.

Authority.

One benefit highlighted solely by staff was the authority that comes with being a police officer. As members of the police service, ACT-PI officers have the power to give legal consequences to clients. Many staff reported that clients often listen to ACT-PI officers more than they listen to other team members because of the officers’ authority. In some instances, staff perceived the ACT-PI officers to draw firmer boundaries for clients than did other staff, who were seen as more willing to negotiate. Some staff felt that the “mere presence” of a police officer had a calming effect on clients because the officers are familiar to the clients. In addition, the ACT-PI officers’ skill at deescalating potentially volatile situations was seen as more effective because of their authority. As a result, fewer disruptions occurred that interfered with ACT-PI team care (e.g., if the client is frustrated about how personal money is managed). Because ACT-PI officer authority enhanced deescalation success, both client and staff safety were enhanced.

Drawbacks

Many clients (N=8, 38%) and more than half of staff (N=12, 54%) explicitly stated that they believed that there were no drawbacks to ACT-PI officer integration. Ten clients and five staff raised at least one drawback, and nine additional staff identified challenges with having an officer on the team (not directly framed as a drawback). Among staff and clients, five drawback-related themes emerged (Table 2).

Risk for consequences.

One drawback mentioned only by clients was the fear of arrest or forced hospitalization. ACT-PI clients were sometimes concerned that the presence of police officers could increase the risk for negative consequences. Concerns about privacy also arose for clients who felt uncomfortable with health care providers sharing information with police.

Police stigma.

Several staff and clients also mentioned that experiencing public stigma due to interacting with police was a drawback. Clients expressed concern that people in their community would think they were police informants if they were seen talking to ACT-PI officers. Among staff, this stigma was more closely tied to fear of criminalizing mental health. In line with this, some clients mentioned particularly disliking police involvement with their care when police came to their home or escorted them in the officer’s vehicle, which was seen as stigmatizing or punitive.

Escalating distress.

Some clients mentioned that it could be difficult to trust ACT-PI officers because of past traumatic police experiences. Staff similarly felt that clients might feel intimidated by officer involvement or that it could escalate violence potential. Staff also noted that officer involvement could result in resistance to engaging with the ACT-PI team.

Low availability.

A few clients and staff noted that ACT-PI officers were not always immediately available. For instance, when clients contacted an ACT-PI officer during a crisis, the officer was sometimes occupied and unable to provide immediate assistance. Interestingly, the two clients who discussed lack of availability also identified the theme of escalating distress, suggesting that clients with a history of police-related trauma might be most likely to perceive ACT-PI officers as unavailable or to feel the need for additional relationship building.

Changed teams.

A drawback highlighted by only one staff member was that officer involvement might change the makeup of ACT-PI teams. The staff member expressed concern that the ACT-PI program could move too far in the direction of clients with forensic involvement and exclude those with only mental illness. The staff member also noted that staff could become overly reliant on officers for deescalation and that the presence of police could change the nature of therapeutic encounters. We included this theme because the participant who endorsed it was one of the most long-term staff members we interviewed, and the theme added pertinent information.

Discussion

Across interviews, perceived advantages of police integration into ACT teams were prominent. The benefits that were identified were discrete and observed or experienced directly. The identified drawbacks were often hypothetical or suspected. Even when respondents mentioned drawbacks, they typically expressed the belief that the drawbacks were outweighed by benefits. Nonetheless, the perceived drawbacks may have a significant impact, and therefore we addressed them fully; they highlight considerations for deciding whether police involvement in a specific person’s care is appropriate. Further, our conclusions were drawn from a community with some (but not complete) mistrust of police and mixed views on police defunding; we note that they may not generalize to other communities where the police have not earned a basic level of trust or community support.

Both clients and staff emphasized the importance of an empathic, nonjudgmental relationship between ACT-PI officers and clients, consistent with past research that has highlighted the need for supportiveness and frequent contact for ACT and FACT clients (15, 22, 23). In addition, both clients and staff felt that the program was safer and more predictable with officer integration, and officer deescalation skills emerged as key to maintaining client and staff safety.

Safe and supportive relationships facilitated other perceived benefits, such as the greater likelihood of preventive, rather than reactive, interventions. When relationship building and prevention are interwoven, clients may experience emotional and practical support that maintains stability. The efforts of ACT-PI officers to support and get to know clients also facilitated the effective use of police officer authority when needed (10) (although we note that some clients viewed the risk for legal leverage as a drawback rather than a benefit). This authority was an asset to clients who were navigating the criminal justice system, because ACT-PI officers helped explain difficult processes and provided support.

The theme of embeddedness reflected the importance of coordinated care and ongoing problem solving as alternatives to punishment (21). Effective interagency collaborations between police and health care are guided by information sharing, joint decision making, and coordinated intervention (24). The embeddedness observed in the ACT-PI model is particularly important for clients with a history of criminal behavior, because it provides a coordinated approach to intervene with modifiable risk factors (e.g., housing instability) that increase clients’ vulnerability to both psychiatric instability and criminal behavior (25).

The integration of officers was not without perceived drawbacks. New clients in particular feared that officer involvement could lead to more criminal sanctions, although staff perceived ACT-PI officers as less likely than other officers to invoke criminal sanctions for minor offenses such as illicit substance use. Ongoing education on the role of police officers may be needed to establish trust and rapport (26).

Several clients expressed concern about the stigma of being seen with police. Some staff also feared clients would be misperceived by the public as violent, which is a common misperception of individuals with severe mental illness (27) that may be amplified by police presence. A less frequent, but important, concern was raised that clients with a history of negative police interactions may be distressed by the presence of ACT-PI officers. These concerns may be especially amplified for Black, Indigenous, and People of Color (BIPOC), because BIPOC clients may be disproportionately affected when police are involved.

These concerns point to several critical recommendations. The officers currently self-select into the role on ACT-PI teams by applying when an opening in a team is available. Officers are selected on the basis of their attitudes and skills related to mental health promotion. This practice should be maintained, rather than assigning officers to rotate through this position, because not every officer would be a good fit for ACT-PI teams. Ideally, officers should reflect the demographic characteristics of the population being served (e.g., include members of racial-ethnic minority groups). Further, officers should be trained in trauma-informed approaches generally and deescalation skills specifically to prevent arrests of individuals experiencing acute distress (28). ACT-PI officers could help train non–ACT-PI officers to minimize the possibility of traumatic interactions more broadly across the system. For some individuals, officer involvement may be contraindicated, such as when there is a history of police violence and lack of trust. Including ACT-PI officers in these cases would compromise the development of trust with the rest of the team. Efforts to promote trust in police within the wider community are also required to reduce the perception of public stigma of interacting with a police officer. Public trust is a substantial challenge in communities with a history of overpolicing, including profiling and use of excessive force.

Only a few ACT-PI officers are available to address the needs of all ACT clients across four ACT teams. As a result, ACT-PI officers may be unable to respond outside of crises. ACT-PI programs require a critical mass of availability to ensure that officers are able to conduct relationship building and preventive work and not just crisis response. Otherwise, officers cannot become well known or trusted, and the benefits of embeddedness are diminished. Ideally, officers should be members of ACT-PI for several years and should be available in the evening and on weekends to further reduce client interactions with unknown officers.

The ACT-PI model seeks to balance services for individuals with and without a history of forensic involvement. The availability of officers on teams may result in an overreliance on officers for deescalation, a skill that all team members should have. Further, if trust does not develop, the presence of an officer could alter the nature of therapeutic relationships. In addition, officer presence may tip the balance toward admitting those with forensic backgrounds, which might prevent some individuals with severe mental illness from receiving much-needed services.

In general, clients who voiced the most drawbacks to police integration were relatively new to ACT-PI and often discussed hypothetical drawbacks, whereas established clients tended to have more positive views of officers based on their experiences. Just as developing trust in the ACT program takes time (29), clients may initially experience apprehension but become more comfortable with ACT-PI officers over time.

Limitations

The data used in this study were based on opinions of ACT-PI clients and staff and thus are inherently subjective and possibly influenced by various biases (e.g., antipolice stigma or motivation to support colleagues). Further, although we interviewed a diverse range of clients, we do not know how representative the sample was of the overall ACT-PI clientele. We also do not know whether our findings would generalize to staff who chose not to participate. Finally, we interviewed only clients who were available in the community; we therefore may have missed the most severely distressed or criminally involved clients.

The clients we interviewed, overall, had fairly limited experience with police. Further, given our sample size, we were unable to make comparisons within each group of participants (e.g., among staff members of different professional backgrounds). Notably, we could not systematically analyze how views of the benefits of police integration differed between clients with and without a history of criminal justice involvement (or victimization).

We did not systematically collect data on racial-ethnic background and therefore were unable to compare perspectives of BIPOC and non-BIPOC clients. Current and past abuses of power by police create substantial obstacles to supportive relationships with police in BIPOC and other communities. Consequently, the level of trust likely necessary for ACT-PI to be effective may not exist in these communities. That said, even in the context of police mistrust, having the stable presence of a police officer on the ACT team could be protective because the ACT-PI officer’s knowledge of individual clients could minimize implicit biases and promote personalized responding. This would be an improvement over dealing only with patrol officers, who do not have individualized knowledge of ACT clients.

Future Directions

Future research should include perspectives of professionals who have contact with the ACT-PI program, such as ACT-PI officers themselves and professionals within criminal justice, mental health, and social services. Further, randomized controlled trials comparing ACT-PI with ACT or FACT models are needed to assess the effectiveness of each model for improving both client and staff well-being. Finally, it would be important to compare experiences of clients with and without criminal involvement, as well as clients who do or do not experience discrimination based on one or more identities.

Conclusions

Clients and staff perceived numerous benefits of the ACT-PI model but also noted several potential drawbacks. Police integration may prevent clients from falling through the cracks between mental health and law enforcement, reduce the risk for traumatic police experiences, and lead to greater housing and psychiatric stabilities. With certain precautions, ACT-PI may effectively meet the needs of clients with and without a history of criminal involvement. This prevention-oriented long-term care model is complementary to efforts to address acute mental health crises, where collaboration between mental health specialist and law enforcement can also be valuable. As communities grapple with ways to support individuals with severe mental illness, the incorporation of police officers into ACT teams represents an innovative effort to integrate the health care and criminal justice systems, as police officers join the mental health team and prioritize client care and support.