Abstract
This study presents an adaptation of the Crisis Intervention Team Model (CIT) to a jail setting. Pre-post surveys and interviews assessed changes in corrections officers’ (CO) knowledge of and attitudes toward mental health. Cell Removal Team (CRT) services assessed the impact of CIT on the use of this specialized unit. Results indicate positive changes in CO attitudes, increased de-escalation skills, and an abrupt decrease in the level of CRT usage, with results sustained in the 8-month follow-up period.
Introduction
In 2016, over 700,000 individuals were incarcerated at any given time, and 10.6 million admissions were recorded in US city or county jails (Kang-Brown et al., 2018; Subramanian et al., 2015; Zeng, 2018). Statistics show that individuals with severe mental illness (SMI) are incarcerated at a disproportionate rate when compared to the general population, and this proportion has been steadily rising since the 1980s (Steadman et al., 2009; Vogel et al., 2014; Zeng, 2018). This phenomenon can be attributed to a combination of factors, including lack of available emergency treatment due to the deinstitutionalization of psychiatric facilities and a dearth of adequate community mental health support. As a result, there is increased interaction between individuals with SMI and law enforcement officers (Lamb & Weinberger, 2005), the majority of whom do not have sufficient training in recognizing and reacting to signs and symptoms of psychological distress. Thus, individuals with SMI are often incarcerated rather than directed to treatment (Compton et al., 2014b; Lamb & Weinberger, 2005).
Correctional facilities have since become the primary care facility for individuals with SMI, despite not being designed as such (Kerle, 2016; Lamb & Weinberger, 2005; Torrey et al., 2010). Research demonstrates that individuals with no history of SMI are at increased risk for psychological distress while incarcerated. Notably, individuals with a history of SMI experience heightened symptoms and are at a greater risk for self-injurious behaviors in carceral settings (DeHart et al., 2009; Dvoskin & Spiers, 2004; Steadman et al., 2009; Vogel et al., 2014). These behaviors are often viewed as dangerous and non-compliant, increasing an individual’s likelihood of interacting with a Cell Removal Team (CRT) that is deployed in crisis situations to restrain or isolate an inmate (DeHart et al., 2009; Dvoskin & Spiers, 2004). Although mental health professionals caution that these actions might worsen symptoms of psychological distress, they are often perceived by correctional staff to be necessary, as they lack other resources for assisting these individuals (Deiter et al., 2000; Subramanian, et al., 2015).
In response to the mishandling and unnecessary deaths of inmates with SMI by law enforcement officers, a model called the Crisis Intervention Team (CIT) was developed to teach officers how to identify signs of psychological distress and strategies for interacting with individuals in this situation. CIT also connects officers with community mental health supports so that individuals with SMI can be diverted to treatment rather than incarcerated (Steadman & Morrissette, 2016; Watson, et al., 2010). CIT programs have been conducted with law enforcement officers in jurisdictions across the United States, and research has demonstrated numerous positive outcomes (Bonfine et al., 2014; Compton et al., 2008, Kubiak, et al., 2017; Comartin et al., 2019; Watson et al., 2010). Advocates for CIT recognize the need to expand and specialize the model for Correctional Officers (COs) (Kerle, 2016). However, no studies to date have examined the effect of CIT in correctional settings. Thus, the current study aims to address this gap. Various data sources are used to assess change in CO knowledge, attitude, and behavior before and after CIT implementation in one metropolitan jail. Pre-post survey instruments and interviews are used to analyze change in knowledge and attitude; change in CO behavior is measured with administrative reports written by CRT officers.
Background
Since the 1980s, the number of incarcerated individuals with SMI has been steadily increasing (Vogel et al., 2014). The increased use of incarceration disproportionately affects vulnerable populations, particularly individuals with SMI (Steadman et al., 2009; Subramanian, et al., 2015; Vogel et al., 2014; Zeng, 2018). SMI is defined as mental disorders that consistently and extensively interfere with a person’s life and ability to perform life tasks (Compton et al., 2014a). A 2002 study conducted by the National Commission on Correctional Health Care suggests that the percentage of incarcerated individuals with SMI is approximately 15% of men and 30% of women (National Institute of Mental Health [NIMH], 2017; Steadman et al., 2009). Similarly, a study conducted by the Bureau of Justice Statistics found that 26% of jail inmates reported symptoms of serious psychological distress, and 44% had received a diagnosis by a mental health professional in the past (Bronson & Berzofsky, 2017). This percentage is higher than in the general population, where only 5% have serious psychological distress (SPD) (Bronson & Berzofsky, 2017).
Multiple compounding factors contribute to the overrepresentation of individuals with SMI in the criminal justice system, including the lack of available psychiatric hospital beds (Lamb & Weinberger, 2005; Vogel et al., 2014), lack of access to mental health treatment in the community (Compton, et al., 2014b; Dupont & Cochran, 2000; Lamb & Weinberger, 2005; Subramanian, et al., 2015), and increased interaction between individuals with SMI and law enforcement officers (Lamb & Weinberger, 1998, 2005; Vogel et al., 2014). As a result, correctional settings have become the primary care facility for this population, as there are more people with SMI behind bars than in psychiatric settings (Lamb & Weinberger, 2005; Torrey, et al., 2010). Correctional facilities are expected to provide both punitive and rehabilitative functions, tasks that seem inherently contradictory and difficult to coordinate (Dvoskin & Spiers, 2004). Most jails are not equipped to provide adequate rehabilitative functions, especially regarding SMI. Thus, symptoms are often exacerbated in correctional settings (Steadman, et al., 2009; Subramanian, et al., 2015; Vogel et al., 2014).
The environmental conditions of correctional facilities have also been cited as potential contributors to higher rates of mental illness for inmates, including the threat of violence and victimization, lack of social supports, loss of autonomy, and physical limitations that lead inmates to experience increased feelings of anger, fear, and humiliation (Dvoskin & Spiers, 2004; Subramanian, et al., 2015; Vogel et al., 2014). Incarcerated individuals also experience higher rates of suicide and self-harming behaviors when compared to the general population (Fazel et al., 2016; Vogel et al., 2014). Therefore, incarcerated individuals with SMI often pose greater behavioral challenges for COs, which can increase the likelihood that they will experience use of force and inmate compliance tools such as physical restraints or tasers (DeHart et al., 2009; Dvoskin & Spiers, 2004). Although professionals caution against the use of restraints or isolation for individuals exhibiting self-injurious behavior, it is often deemed necessary in correctional facilities to maintain order and security (Deiter et al., 2000; Dvoskin & Spiers, 2004). While it has been well documented that inmates with SMI are likely to experience heightened symptoms during incarceration, individuals with no previous mental health diagnoses may experience a degree of psychological distress from the jail’s environment, which could potentially turn into a mental health crisis (DeHart et al., 2009; Dvoskin & Spiers, 2004).
Crisis Intervention Team Model
CIT has a track record of success with law enforcement (Dupont & Cochran, 2000). The goal of CIT is to train police officers to identify symptoms of SMI and learn about community resources where they can divert individuals to mental health treatment (Steadman & Morrissette, 2016; Watson, et al., 2010). The CIT model includes several core elements (Dupont et al., 2007)—three essential for model implementation: (1) a 40-h training for law enforcement, (2) crisis response services where officers can divert individuals in crisis, and (3) community partnerships to problem-solves issues that arise between community-based treatment providers and law enforcement agencies (Dupont & Cochran, 2000). The founders of CIT stress that successful implementation is dependent upon partnerships between law enforcement, community mental health agencies, and individuals with SMI and their families (Bonfine et al., 2014; Dupont & Cochran, 2000; Steadman & Morrissette, 2016).
To date, there have been numerous studies that have demonstrated positive outcomes for both officers and individuals with SMI as a result of CIT. For officers, these outcomes include an increase in knowledge about mental health and ability to identify symptoms of mental illness, an increased confidence in dealing with individuals with SMI, and a change in attitude toward this population (Bonfine et al., 2014; Compton et al., 2008; Kubiak et al., 2017; Watson et al., 2010). While these knowledge and attitude changes are relatively consistent across studies, Taheri’s (2016) meta-analysis found null effects in rates of diversion from jail. However, there is evidence suggesting that CIT-trained officers are more likely to transport individuals with SMI to crisis drop-off centers and community mental health services than their non-CIT trained counterparts (Kubiak et al., 2017; Taheri, 2016). Despite advocates for CIT expressing the need for specialized CIT trainings for correctional settings (Kerle, 2016), no studies were found that investigated the efficacy of training COs in jails (For a recent student on CIT training for COs in prisons: Canada et al., 2020). Therefore, this study asked: Does the CIT model have an impact in a county jail setting?
Method
Case Description
This study investigated the use of CIT in a county jail in a large metropolitan area in the Midwest. The county is home to 1.2 million individuals (U.S. Census Bureau, 2016). The majority of residents (73%) are White, with a higher than average median income ($69,850; 8.7% living in poverty) and education (93.5% high school diploma, 45% completed a bachelor’s degree or higher) than the national average. The county sheriff’s office is the administrator of the county jail which has a capacity of 1,664 beds. In 2017, there was a total of 18,124 jail admissions. The jail has a 24-h health clinic with seven beds and an additional 18-cell unit used for officer observation (i.e., suicide watch, violence risk, etc). This county was part of a larger study evaluating successful requests for funds for jail diversion services offered through the state’s department of health and human services. The study was reviewed by one author’s institutional review board, which deemed the study as “non-research” due to its evaluative nature and lack of research manipulation.
The sheriff’s office previously provided CIT training to 20% of patrol officers and was now providing it to the county jail COs. Eight, 8-h sessions were held between May and July of 2017 by a certified CIT trainer who adapted the training to the correctional setting. The training program covered various mental health diagnoses (schizophrenia, bi-polar, etc.) and symptomology, as well as information about psychotropic medications and their side effects. Suicide in correctional settings was also a training topic. COs were trained in de-escalation techniques through the use of role-play scenarios. This instruction covered the same material as CIT for police officers, but was condensed and geared toward the specific interactions and situations faced by COs. There were a total of 307 jail COs at the time of the first training.
Data Sources
Three data sources were used to assess CO knowledge, attitudes, and behavioral changes. Knowledge and attitude changes were assessed through two sources: (1) pre/post surveys for all COs who took the training and (2) pre/post-interviews with a stratified sample of ten COs. Administrative data in the form of officer reports were used to assess whether CIT training impacted COs’ behavior. All three data sources are described below.
Pre/Post-Instrument
The same survey instrument was given to all COs immediately before and after the 8-h training session. The survey asked for their gender (male/female) and the number of years of service. They were asked about their level of agreement with the following statements, “I have a strong desire to take this training” and “Issues around mental health/illness are a serious problem for officers”.
Behavioral Outcomes Scale (BOS)
Next, the survey presented a scenario about an inmate named David who is hearing voices and threatening suicide, while making no attempts to put officers or other inmates in danger. This scenario, along with eight corresponding questions, were adapted from the Behavioral Outcome Scale (BOS) de-escalation skills subscale, which has shown good reliability and validity in prior CIT research (Broussard et al., 2011). For the current study, we adapted the scenario and questions for the jail/inmate context. The adapted BOS questions asked officers if certain statements or behaviors would be “negative or positive” when interacting with David. Scores ranged from 1 (very negative) to 4 (very positive), with a total possible score ranging from 8 to 32. An example is, “Keeping some space between you and David while you talk to him” (see Table 1 for all questions).
Pre-Post Instrument Results for BOS, Attributions Scale, and Officer Efficacy.
p < .05, **p < .01, ***p < .001; n.s., not significant.
Attributions scale
Six questions of attribution previously validated by Corrigan and colleagues (2003) were asked to evaluate COs’ feelings toward someone like David. The larger set of questions, created by Corrigan and colleagues (2003), was shown to be valid and reliable in general settings, but has not been assessed with the justice-involved population. COs were asked “what best indicates your thoughts about yourself and David’s situation”, with response items ranging from 1 (not at all) to 5 (extremely). Total possible score was 5 to 25. An example question is, “How controllable do you think David’s present condition is?”
Officer efficacy
Last, we asked two questions to assess how confident COs were that they could intervene with someone like David. Similar to the Attributions Scale, COs were asked “what best indicates your thoughts about yourself and David’s situation”. Responses ranged from 1 (not at all) to 5 (extremely), with a total possible score of two to ten. As an example, COs were asked how confident they felt in their ability to effectively communicate with someone like David.
Data analysis
Almost all COs participated in the 8-h training (99.7%, n = 306), with 301 (98.4%) taking the pre-test and 290 (94.8%) the post-test. Officers were asked a series of questions to create a unique identifier which was used to match their pre and post surveys. In total, 48 individuals were not included in the analyses because they did not complete one of the surveys, or their unique identifier was not completed. Three remaining individuals were removed from the analysis due to missing half or more of the Behavioral Outcomes or Attributions questions. Aside from these cases, missing responses were imputed with the series mean. Overall, 255 (83.3%) were successfully matched. Some questions were reverse coded, with a positive reported mean change score meaning that COs used more appropriate de-escalation techniques or their feelings toward individuals with mental illness became less stigmatizing and more understanding of their needs. Each question on the instrument was assessed individually using paired samples t-tests. Total scores for each scale are also presented; however, these should be interpreted with caution due to low reliability scores on the BOS (pre = .26, post = .40) and attributions scale (pre = .34, post = .25; Officer Efficacy (pre = .85, post = .85).
Pre/Post-Interviews
Paired pre and post-interviews were conducted with nine COs. A stratified sample was selected, based on gender (male/female), rank (deputy/administrator), shift (mornings, afternoons, and midnights) and primary position in the jail (at booking/release, the men’s or women’s unit, or the mental health observation unit). The sample was also stratified by training level, by selecting COs based on their membership on the Cell Removal Team (CRT). CRT is a group of COs who are highly trained in de-escalation techniques and appropriate use of force for challenging inmates. These COs were selected for interviews to determine if CIT training added additional benefits above and beyond the extensive training provided to this special unit.
Interviews lasted between 30 and 45 min. Pre-interviews were conducted 1 month prior to the first CIT training, and post-interviews were conducted 6 months after the last training. Waiting to perform post-interviews was purposeful, for time to pass before assessing the impacts after CIT was implemented. After each interview, the lead interviewer and note taker discussed responses and typed their notes together.
Instrument
Questions on the interview instruments were open ended. The pre-training interview instrument asked officers: what they hoped to gain from the training, how they recognize mental illness in inmates, what techniques they use to handle inmates with mental illness, and how frequently they felt they relied on CRT response for de-escalation. CRT interviewees were specifically asked about requests for their services within the jail: what happens before they are called, what they do when they are requested, and their initial response to a CRT situation. Finally, interviewees were asked about the resources available to inmates with mental illness within the jail and what value they thought CIT could bring to the jail. The post-training instrument revisited the individual’s expectations for CIT training, asked what knowledge, skill, or techniques they gained, and if they believed that situations involving inmates with mental illness had changed since the training. CRT officers were asked about any changes that had occurred on their team as a result of the training. Finally, all officers were asked if they experienced any barriers to implementing what they learned in CIT in their day-to-day roles.
Data analysis
A third team member coded and analyzed pre/post interview notes, which were reviewed to extract recurring themes and then organized in Microsoft Excel using a case-ordered descriptive matrix (Miles et al., 2014). Data were arrayed case-by-case according to COs’ characteristics (e.g., gender, position in jail, etc.) and their participation on the CRT. Further, pre/post analyses were conducted to assess changes within each officer.
Calls for CRT
Administrative reports written by CRT officers attending to incidents in the jail were obtained to assess the requests for this team to manage challenging inmates. To examine whether the CIT training had an impact on CO behavior, reports were analyzed in the 12 months prior to and 8 months after training began. Interrupted time series (ITS) analyses (McDowall et al., 1980) was used to determine the impact that CIT training had on the use of CRT within the jail. Time series models were fitted separately to the pre-training data points—the number of times the CRT was called into a contentious situation with an inmate in the 12 months prior to CIT training. The pre- and post-intervention slopes were compared statistically to test whether there was significant change in trajectory following training. In addition, the time trend resulting from the pre-intervention data points was projected past the point of CIT training in order to compare the number of monthly calls for service in the post-intervention model with the number that would be expected if the training had no significant effect and the pre-intervention trend continued. We made these comparisons at 1, 3, and 8 months after the training was implemented. For all models, we used autoregressive integrated moving average (ARIMA) methods, which account for autocorrelation and allows adjustments for nonlinearity and nonstationarity. In this case, we found that simple ARIMA (1,0,0) models, which account for autocorrelation between immediately adjacent points (lag 1), were sufficient to describe the data.
Results
Pre/Post-Training Surveys
Of the 255 COs matched on the pre/post-survey instruments, the majority were male (73.3%, n = 187). The average years spent working in the field was 12.2 years (SD = 9.7), and ranged from zero to 45 years. At pre-survey, 70.1% (n = 178) agreed that they had a strong desire to take the training, and 96.9% (n = 246) believed that mental health issues were a serious problem for law enforcement. At post-survey, both of these proportions had increased (desire = 89.7%, n = 226; and serious problem = 98.0%, n = 249).
The mean change in five of the eight BOS questions showed significant changes, with four in the positive direction and one in the negative direction. In responding to the scenario of David, COs positively increased their de-escalation scores around threatening language: segregation (Mean change = +.10, t(1) = 2.070, p < .05) and commanding David to calm down (Mean change = +.27, t(1) = .27, p < .001). In addition, COs positively increased their scores regarding physical actions that can be perceived as threatening, by having their hand on their baton/gun (Mean change = +.32, t(1) = 6.663, p < .001). COs positively increased their use of questioning David about his feelings (Mean change = +.30, t(1) = 6.216, p < .001). Conversely, COs negatively changed their scores about keeping space between themselves and David (Mean change = –.17, t (1) = –4.463, p < .001). None of the remaining scores significantly differed from pre- to post-test. The overall change on the BOS instrument was +.90 (t 1) = 6.079, p < .001), but this should be interpreted with caution due to the low reliability score of this scale (see Table 1).
All questions on the attributions scale significantly changed from pre- to post-test. Overall, COs felt less aggravated (Mean change = .14, t (1) = 2.489, p < .05) and frightened (Mean change = .30, t(1) = 3.900, p < .001) of David. They positively changed their perceptions of dangerousness (Mean change = .73, t(1) = 10.464, p < .001) of David. Also, significant increases were found in the COs helpfulness toward David (Mean change = .42, t(1) = 6.774, p < .001). Conversely, COs felt that David’s illness was more controllable at post-test (Mean change = –.25, t (1) = –3.294, p < .01). The overall change in the attributions scale was 1.34 (t (1) = 8.016, p < .001), but should be interpreted with caution due to low reliability scores of this scale (see Table 1).
Both questions related to CO efficacy resulted in positive changes at post-test. COs felt more confident in their communications with David (Mean change = .50, t (1) = 9.546, p < .001) and felt confident to calm him down (Mean change = .52, t(1) = 10.655, p < .001). The overall change in officer efficacy was 1.027 (t (1) = 11.549, p < .001) (see Table 1).
Pre/Post-Interviews
Ten COs were interviewed in April of 2017 before receiving CIT training. Due to one retirement, nine were interviewed in March of 2018 after the training. Of the initial ten COs, eight identified as White; two as Black. Eight COs identified as male; two as female. COs represented all shifts; four on days, three on afternoons, and three on midnights. COs ranged in experience from one to 30 years; five COs had five or less years of experience. COs’ positions varied throughout the jail, with the possibility of COs’ taking on more than one position. Six COs belonged to the CRT, with two holding the rank of CRT Team Leader. COs’ daily positions included six in Receiving/Release, four in the observation unit, one in the Clinic, and one in the women’s unit. One CO held a supervisory role in the observation unit.
Pre-CIT training
COs had limited knowledge of CIT during pre-interviews, although all were aware that it was related to their professional involvement with inmate mental health. While COs anticipated their role being limited due to the nature of their job to secure and incarcerate, all welcomed additional training. Prevailing themes around what COs hoped to learn from CIT included: assessing potential mental health situations, identifying mental health issues, and learning ways to improve communication with inmates. All COs mentioned the need to know more about mental health issues and how to handle relating situations.
Perceptions of mental health
Pre-interviews revealed three common perceptions about mental health and its relation to corrections. Based on regular interaction, COs perceived there to be an excessive number of individuals with mental health issues incarcerated due to lack of access to proper treatment outside of corrections and inability to control symptoms that result in criminal acts. Second, COs noticed a lack of mental health resources in and outside of the correctional system, with emphasis on diversion from incarceration and community resources. COs saw their greatest resource to be caseworkers and mental health professionals in the jail, although they were either unaware or dissatisfied with the scope of their work. Third, COs perceived the escalation of a mental health-related situation to depend on the inmate’s ability and/or willingness to communicate and the CO’s ability to recognize an inmate’s communication patterns.
COs mentioned being aware that individuals with mental health issues may not be aware of or be able to control inappropriate behaviors. COs’ perceptions of mental health could relate to their ability to recognize mental health situations; COs who shared the perception that individuals with mental illness cannot always control their behavior also shared similar strategies for recognizing mental health situations.
Recognizing and handling mental health situations
The most commonly reported technique for recognizing mental health situations was observing inmate behaviors. 89% (8/9) of COs claimed being able to distinguish poor behavior from mental illness by watching for sustained or repetitive behaviors, such as inappropriate handling of food or feces, self-harm, abrupt changes in emotion, incoherent communication, and/or peculiar body movements. Rarely (22%, 2/9), COs used the knowledge of other corrections personnel and available medical histories to recognize mental health related situations, which were often limited. Regardless of strategy to recognize a mental health situation, all COs admitted having difficulty and used a variety of strategies to handle such situations.
Pre-CIT interview responses regarding strategies for handling mental health situations were primarily procedural rather than specific to the individual CO. COs recounted procedure to first attempt verbal de-escalation, refer to supervisors if needed, who would then refer to CRT if de-escalation could not be achieved. COs claimed CRT procedure was to attempt verbal de-escalation again and escalate to hands-on tactics if compliance was not given.
While procedure may be similar in every situation, six COs mentioned that they themselves or CRT adjust their tactics when mental health is a known concern. CRT members said that they were more likely to make multiple attempts at verbal de-escalation when a mental health issue was known. Others mentioned using communication techniques that include playing along with inmates’ delusions, being agreeable, and demonstrating active listening by paraphrasing and acknowledging concerns. These techniques were said to be used to build trust in order to avoid future escalation. “People just want to be heard”, one CO recounted, “A lot of times, someone knowing you’re listening is enough to get them to get off the ledge”.
Post-CIT training
COs seem to have already been familiar with some verbal de-escalation techniques pre-CIT, but they were not recognized as CIT or specific to mental health situations. Overall, COs claimed that CIT gave them broader understandings of mental health and communication, while also reinforcing previously held perceptions.
Perceptions of mental health
Two out of three COs (67%, 6/9) exhibited a positive change in overall perception of mental health. COs showed increased understanding for medication effects on behavior along with greater understanding of why individuals might not be able to control their behaviors. One CO, who had been in corrections for nearly 20 years and was at the time responsible for classification at booking, mentioned having gained a new understanding that not everyone with a mental health issue, “is a mess”. One CO positioned in the women’s unit noticed that judgments made from COs’ lack of understanding could lead to increased aggression toward inmates. COs also reported decreased judgments and increased empathy toward inmates post-CIT. One CO noted that increased exposure to mental health through CIT increased their ability to recognize mental health situations, which positively impacted their perception of such situations. The 33% (3/9) of COs who did not report positive change in perceptions of mental health reported that CIT supported their existing perceptions, the common one being that many individuals with mental health issues do not belong in jails. Overall, perceptions of mental health were either positively impacted or reinforced primarily through the ability to see reasons behind these behaviors.
Recognizing and handling mental health situations
COs’ abilities to recognize mental health situations improved since CIT but techniques remained similar. Two out of three COs (67%, 6/9) claimed that CIT improved or reinforced their ability to observe behaviors and differentiate between poor behavior and mental health. COs were non-specific about how their ability to recognize mental health behaviors had improved, but credited CIT and prior experience. Even though COs noticed personal improvements, they admitted a lack of complete knowledge around mental health and requested additional training.
While techniques for recognizing mental health situations remained relatively stable, techniques for handling situations showed positive change, with 77% (7/9) of COs claiming that CIT positively impacted their techniques for handling mental health situations. COs highlighted learning the skills of listening to and acknowledging, inmates’ concerns, showing mutual respect, building rapport, and remaining patient in behavior and voice.
Pre-CIT, COs reported using verbal de-escalation to play along with inmates’ delusions, intending to avoid further conflict. Post-CIT, COs reported spending more time communicating one-on-one with inmates and making greater efforts to understand their experiences and concerns. COs have begun asking inmates about their backgrounds, intending to build rapport and trust to limit future escalated situations. One CO remarked that they have to, “Give respect to get respect” and that it, “Feels good to de-escalate a situation even if it takes an hour”. The same CO claimed that they were less irritated with noncompliant inmates as they improved their communication.
While no change in official jail procedure or CRT response was seen, COs reported having observed other COs increase time spent communicating with inmates. All of the COs interviewed agreed that increased one-on-one communication with inmates led to fewer escalated situations, which in turn would lead to fewer CRT requests. Additionally, COs have noted that using CIT communication techniques have saved time because non-escalated situations do not require additional supervision and report filing.
Pre-to-Post Change
While no clear demographic patterns were observed around pre-post changes, some inferences can be made. It could be suggested that COs positioned in Receiving/Release are exposed to a larger number of inmates, with whom they have less opportunity to have sustained interaction. Additionally, these COs are often exposed to inmates coming in under the influence of a substance or are in crisis. Thus, mental health may be more difficult to determine.
CRT members did not stand out among other COs when it came to pre-post-CIT change. However, each member reported an improved or reinforced perception of mental health and improved communication with inmates. While these improvements were not unique to CRT members, CRT has been noticed by other COs to be less likely to take an aggressive approach when they encounter an escalating situation. Pre-CIT, CRT members anticipated their ability to use CIT techniques to be very limited. However, post-CIT, CRT members claimed to spend more time attempting verbal de-escalation when possible.
When COs on the observation unit were asked how their position impacted their interaction with mental health situations, it was found that an excess of inmates with mental health concerns may have caused the ward to become a place to cycle inmates in crisis. COs on the observation unit remarked that this left inmates with more stable or more manageable mental health issues to be housed in general population. COs positioned in every other role echoed the perception that the number of inmates with mental health concerns has increased. This points to the importance of continued and expanded CIT, which COs requested.
In pre-interviews, COs mentioned that their knowledge of mental health resources was limited beyond the caseworkers in the jail. Further, COs admitted not knowing the responsibilities of the caseworkers or mental health professionals other than the ability to prescribe medications and advise behavior watches. COs noticed no change in resources or knowledge around resources post-CIT, except for the single CO who was repositioned from midnight to day shift and gained access to on-site caseworkers as opposed to on-call.
Calls for CRT
In the 12 months pre-training period, there were a total of 361 calls for CRT to assist with challenging inmates in the jail. The average calls for CRT team was 30.1 per month, and ranged from 24 to 40 per month. In the 8 month post-CIT period, there were a total of 131 calls for the CRT, with an average of 16.4 per month. CRT calls in the post-period ranged from 4 to 24. The time-series model in Figure 1 illustrates both the projected and observed monthly use of CRT to escalated situations in the jail. The graph shows a large decrease in the number of monthly calls for CRT following the start of CIT training. This change was significant and largely persisted through the observed period: in the month after CIT training, the post-intervention model showed 20.3 fewer requests for CRT than projected from the baseline time trend (p < .01); 3 months after the training 18.7 fewer requests were made for CRT than projected (p < .01). Eight months post-training found 14.6 fewer calls for CRT, which neared statistical significance (p = .063).
Interrupted time series analysis of calls for cell removal teams.
In the pre-period, the number of calls for CRT remained relatively stable month-to-month, with a slight, but not significant, positive slope (B = 0.4, p = .311). The slope for the post-training period was not significantly different from the pre-training slope (difference in B = 0.8, p = .425), indicating no significant decline over the post-training months. These findings suggest that CIT training resulted in an abrupt decrease in the level of monthly CRT usage, and that over the 8-month period following training, this change persisted.
Discussion
The primary goal of this study was to determine if the CIT model has an impact in a correctional setting, noting that previous research on the CIT model has been with law enforcement. Training COs in CIT is important because correctional settings have become multidimensional primary care facilities for individuals with SMI that are expected to serve both punitive and rehabilitative functions, often regardless of the functional capacity to do so (Dvoskin & Spiers, 2004; Lamb & Weinberger, 2005; Steadman et al., 2009; Subramanian et al., 2015; Torrey et al., 2010; Vogel et al., 2014).
Major findings from this exploratory study showed that CIT training had a positive impact on COs and their experiences with SMI-related situations. COs exhibited overall positive changes in their knowledge of mental illnesses, as well as in their attitudes toward individuals with SMI. In addition, significant changes were found in CO de-escalation skills and their level of efficacy in handling a SMI-related situation. The training goal of increased use of verbal de-escalation and decreased use of physical response was achieved, with rates of monthly CRT calls showing a significant and persistent decline, and officer interviews confirming the COs perceived changes among COs and their colleagues.
To date, this is the first study to assess the use and impacts of CIT in a jail setting. And similar to the recent evaluation study of CIT training in prison (Canada et al., 2020), there were similar positive outcomes. Additionally, prior research on use and impact of CIT with law enforcement officers showed consistent improvements in officers’ knowledge and attitudes about mental health (Bonfine et al., 2014; Compton et al., 2008; Kubiak et al., 2017; Watson et al., 2017). COs in this study also showed significant improvements in their skills in handling SMI-related situations and attitudes toward individuals with SMI post-CIT. Skills gained from CIT for both law enforcement and correctional officers were intended to help shift their approach with individuals with SMI, from using more rehabilitative measures, and away from punitive tactics.
Diversion, as it relates to CIT, was illustrated differently for law enforcement and correctional officers. For prior research with law enforcement, diversion involved officers utilizing arrestee drop-off sites at community mental health partners (Kubiak et al., 2017; Taheri, 2016; Watson et al., 2010). In contrast, diversion within the county jail setting was indicated by a decrease in requests of the CRT, along with COs’ awareness of the mental health unit and professionals in the jail. The relationship between CIT and diversion into treatment is unknown in this study, as COs reported limited awareness of the capabilities of these professionals in assisting individuals with SMI, outside of verbal de-escalation. Jail practices of mental health identification and referral by COs to jail-based mental health services has shown positive, yet inconsistent practices across multiple jails (Kubiak et al., 2020). Future work should assess the transition from jail-based mental health treatment into the community, and the relationship between treatment engagement and a return to jail.
Implications
This jail-based study revealed that COs benefit from CIT training, by increasing their knowledge and attitudes about individuals with SMI and changing their behaviors toward this population. Yet, the remaining essential components of the CIT model should be given additional attention. COs desired more training on mental illness after CIT training, and there was still a lack of knowledge about caseworkers in the jail. The model requires a more collaborative approach, with COs and mental health workers problem-solving cases together. Administrators who are considering CIT training in correctional settings should establish communication pathways between these roles, and may even consider cross-training COs and caseworkers in each others’ daily tasks. The building of these collaborative networks can likely lead to benefits for the institution, the staff, and jailed individuals with SMI.
Where CIT is implemented, jail administrators need to be aware of the amount of time COs spend in verbally de-escalating a suspected SMI situation. During interviews, COs positioned in Receiving/Release expressed that they were not always permitted enough time to interview or converse with inmates, likely due to the fast-paced and high-volume admissions/ discharges environment. COs observed that this was a barrier to becoming aware of inmates’ mental health needs in order to provide assistance or referrals for proper care. In some cases, this led COs to not attempt verbal de-escalation. For COs in other areas of the institution, more time is also be needed “up front” to de-escalate potential crisis incidents, compared to time after a crisis spent in debriefing and report writing.
The requests from COs for additional resources suggest needed procedural changes. Related to in-facility caseworkers, COs’ primary concern was onsite access to them or other mental health professionals after traditional work hours. One CO, who had been moved from the midnight to day shift during the period between pre- and post-interviews, observed that the ability to assist with SMI situations had improved markedly with access to onsite caseworkers. Other COs noted that when caseworkers were able to assist with SMI situations, they were not able to assist beyond the point of de-escalation; additional knowledge of resources for individuals to access outside of the correctional system was recommended. Jail administrators can respond to such proposals by providing COs with a compilation of community mental health resources to provide to inmates as a low-cost option. This could also include the inmate’s familial or other support systems with which COs could build rapport and discuss care options at reentry. Given these findings, the hiring of additional caseworkers to support all shifts is warranted.
Importantly, COs’ concerns about mental health resources within and outside the correctional system emphasize the need for policy change at local, state, and federal levels. First, the allocation of funds for CIT training is necessary for creating a stabilizing environment in the jail. Similarly, seeking funding for additional caseworkers would build on the jail’s mental health services capacity. Last, the critical issue of SMI-supportive community-based services at jail inmate re-entry is crucial to addressing SMI jail recidivism.
Limitations
While this exploratory study found positive outcomes from CIT in a jail setting, the investigation has its limitations. First, the location and context of this study is in an affluent county with considerably more mental health resources, both inside the jail and in the community. In addition, COs have engaged in other trainings that have included skills similar to those in CIT. As a result, these findings may not be generalizable to other jails with fewer resources.
Whereas item analysis for each question on the survey instrument showed mostly positive changes, the reliability of scales used in prior CIT studies was not found in this study’s applications. Accordingly, the cumulative scores on the BOS and Attributions scale must be interpreted with caution, and future research should build measures for CIT that are useful in this context.
This study was also challenged by the use of administrative data for the outcome variable. The use of CRT reports was used as a proxy for the number of critical incidents that occurred in the jail but was not a direct indicator of mental health severity in the jail as not all CRT calls for service are related to mental health. Future CIT jail studies should include outcome measures that directly assess situations related to mental health crises or individuals in distress.
Conclusion
This exploratory research assessed the use of the CIT model in a county jail. Significant positive changes were found in CO knowledge about mental illness, CO attitudes toward this population, and CO behaviors surrounding SMI situations. These preliminary findings support the growing use of CIT in correctional settings as the incarceration of the mentally ill shows no sign of declining.
Footnotes
Acknowledgements
The authors thank the Michigan Department of Health and Human Services and the Governor’s Mental Health Diversity Council for funding this project.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the Governor’s Mental Health Diversion Council and the Michigan Department of Health and Human Services.