Research in the Real World: Studying
Chicago Police Department’s Crisis
Intervention Team Program

Research on Social Work Practice
20(5) 536-543
ª The Author(s) 2010
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1049731510374201
http://rswp.sagepub.com

Amy C. Watson1

Abstract
Police agencies across the country are struggling to respond to significant number of persons with serious mental illness, who are
landing on their doorsteps with sometimes tragic consequences. Arguably, the most widely adopted approach, the Crisis
Intervention Team (CIT) model, is a specialized police-based program designed to improve officers’ ability to safely and effectively
respond to mental health crises. Conducting research on CIT is challenging; thus, the evidence base is limited. In this article, the
author reviews the emerging literature, present a conceptual model of CIT effectiveness, and describe a study of CIT in Chicago.
Findings from Chicago suggest that CIT is increasing linkage to services and reducing use of force in encounters with persons with
mental illness. Lessons learned are discussed.
Keywords
criminal justice, intervention, mental health, mixed methods
At the same time we struggle to get systems—agencies and
clinicians to implement evidence-based interventions—other
interventions are getting ahead of the data and spreading like wildfire (Horvitz-Lennon, Donohue, Domino, & Normand, 2009).
Proceeding ahead of the data seems to be particularly common
in the criminal justice/mental health arena, where jurisdictions are
struggling to respond to significant numbers of persons with serious mental illness, who are landing on their doorsteps with sometimes tragic consequences. Across the country, collaborations of
stakeholders have come together to develop innovative responses.
Based on anecdotal successes, some of these models (e.g., mental
health courts, Crisis Intervention Teams [CITs]) have taken hold
and spread quickly.
Slowing the pace of the dissemination of promising models
to allow the evidence base to catch up is unrealistic. Likewise,
it may not be possible for political, ethical, or practical reasons
to conduct randomized controlled trials of interventions involving multiple systems. However, we can systematically develop
an evidence base that allows us determine whether these interventions are effective, for whom and under what conditions. In
this article, I discuss the emergence of the CIT model as an
intervention to improve police response to mental health crisis,
its rapid dissemination, and the limited but growing evidence of
its effectiveness. Then, I will describe an effort to systematically examine the CIT program in Chicago and summarize
lessons learned and findings to date.

Veysey, & Morrissey, 1999). Data on the number of people
police are directing to mental health services are limited. However, estimates of the prevalence of mental illness in U.S. jails
and prisons range from 6% (Teplin, 1990) to 16% (Ditton,
1999) suggesting that police are arresting sizable numbers.
Arguably, the most widely adopted law enforcement
approach to address this issue, the CIT model, is a specialized
police-based program designed to improve police officers’
ability to safely and effectively respond to mental health crises.
The model emerged out of grassroots collaboration between
law enforcement, advocates and mental health providers following the tragic shooting of a man with serious mental illness
by a Memphis, Tennessee police officer in 1988 (Dupont &
Cochran, 2000). The model involves specialized training for
officers, collaboration with community providers, a central
psychiatric crisis drop off center, and shifts in organizational
policies and procedures. Officers volunteer to receive 40 hr
of training provided by mental health clinicians, consumer and
family advocates, and police trainers. Dispatchers are trained
to identify mental disturbance calls and assign these calls to
CIT trained officers. If necessary, the CIT officer uses deescalation techniques and assesses whether referral to services
or transport for mental health evaluation is appropriate. Early
reports from Memphis suggest that the CIT program has

1

Jane Addams College of Social Work, University of Illinois at Chicago

Police, Persons With Mental Illness and CIT
Approximately 10% of all police contacts with the public
involve persons with mental illness (Deane, Steadman, Borum,

Corresponding Author:
Amy C. Watson, 1040 W Harrsion St, Chicago, IL 60607, USA
Email: acwatson@gmail.com

536
Downloaded from rsw.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016

 Watson

537

reduced arrests and increased safety and diversion to mental
health services (Dupont & Cochran, 2000).
Enthusiasm about the CIT model has spread quickly as police
agencies struggle to demonstrate greater responsiveness to persons with mental illness. Currently, there are hundreds of CIT
programs in over 39 states (Compton, Bahora, Watson, &
Oliva, 2008; University of Memphis CIT Center National Directory, n.d.). While suffering from a variety of methodological
limitations, the emerging body of literature suggests that CIT
programs are producing some of the intended effects. After
briefly reviewing this literature, I will discuss how the nature
of CIT and police settings presents significant barriers to conducting methodologically rigorous research.

Evidence to Date: Research on CIT
A goal stated by many CIT programs is to divert persons with
mental illness away from the criminal justice system to mental
health services. An early study (Steadman, Deane, Borum, &
Morrissey, 2000) used police reports to compare arrest rates
from three departments, each using a different specialized
response model. All three departments had relatively low arrest
rates for mental health calls. The rate for the department with
CIT (Memphis) was the lowest at 2%. These data are sometimes used in support of the claim that CIT reduces arrests of
persons with mental illness, despite the absence of a pre-CIT
comparison.
Using dispatch data on mental disturbance calls in Akron,
Ohio, Teller, Munetz, Gil, and Ritter (2006) found that in
post-CIT implementation there was an increase in the number
of mental health–related calls identified; increases among CIT
officers in transports to the hospital for psychiatric evaluation
and increases in the proportion of transports that were voluntary. They did not find a decrease in arrests.
Another important goal of CIT programs is to increase
safety for officers and persons with mental illnesses. Dupont
and Cochran (2000) have reported an association between CIT
implementation in Memphis and decreased use of highintensity police units such as Special Weapons and Tactics
(SWAT) teams and a lower rate of officer injuries. However,
a study using administrative data from another urban police
department did not find an association between CIT implementation and SWAT call outs (Compton, Demir, Oliva, & Boyce,
2009). There is scant quantitative data available on CIT’s effectiveness for reducing injuries. However, in a qualitative study
(Hanafi, Bahora, Demir, & Compton, 2008) officers reported
that application of their CIT skills reduces the risk of injury
to officers and persons with mental illness.
Closely related to safety and injuries is use of force. Skeem
and Bibeau (2008) examined violence risk and use of force
using police reports for CIT calls. Officers used force in only
15% of the events with serious to extreme violence potential.
When they did use force, they relied on low-lethality methods.
While the study did not involve a comparison group, the
authors indicate the findings suggest CIT may be improving
safety in potentially risky encounters.

A number of additional outcomes of CIT programs have been
explored in the literature. For example, one study found that
CIT officers are able to accurately identify individuals with
mental illness (Strauss et al., 2005). Studies have also shown
CIT training is associated with improvements in attitudes and
knowledge about mental illness (Compton, Esterberg,
McGee, Kotwicki, & Oliva, 2006) and officers’ confidence
in identifying and responding to persons with mental illness
(Wells & Schafer, 2006).
Thus, evidence to date is guardedly positive. Given the
methodological and resource constraints inherent in evaluating
applied interventions, none of these studies included control
groups or randomized designs, some did not have any type of
comparison group. Most relied on available documentation,
which may have been limited in its reliability.
Barriers to research on CIT. The lack of randomized controlled
or other rigorously designed studies of CIT is not for lack of
knowledge and skill on the part of researchers interested in the
area. Conducting research on CIT presents the usual applied
research issues along with a few unique challenges that limit
design options. These include the voluntary nature of officer
CIT participation (Council of State Governments [CSG],
2002), researchers’ lack of ability to dictate police department
personnel assignments, and administrative data systems that
do not consistently record mental health status information. In
the following sections, I discuss how we addressed these challenges in designing our study of Chicago Police Department’s
(CPD) CIT program.

Examining CIT in Chicago
The Conceptual Model
Prior to embarking on research of CIT in Chicago, my colleagues and I developed a conceptual model of CIT effectiveness (Watson, Morabito, Draine, & Ottati, 2008). As social
work researchers, we considered not only the person and the
effects of training but also contextual factors that likely determine the opportunities and options for officers to apply knowledge and skills on the job and ultimately the effectiveness of
CIT. Our model includes constructs related to three levels of
the officer’s work context, the organization, the local mental
health system, and the community.
CIT outcomes. The primary goals of CIT programs are to
reduce arrests of persons with mental illness, increase diversion to mental health services, and improve safety. We consider how calls are resolved in terms of arrest, linkage to
services, and contact only/no action. We also examine injuries to officers and call subjects. However, because injuries
are rare, we also examine likely precursors-subject resistance
and officer use of force. Finally, we include officer use of
CIT skills.
Officer-level factors. It has been noted that officers that volunteer for CIT may have characteristics that reduce the likelihood
injury in encounters with persons with mental illness and
537

Downloaded from rsw.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016

 538

Research on Social Work Practice 20(5)

Officer
Characteristics
Demographics
Familiarity
with MI

Completion of
CIT Training

Treatment
Linkages
Availability
Perception of
police officer

Officer and
Encounter
Outcomes
Interaction
of CIT
&
Treatment
linkages

Skills with PSMI
Use of Force

Organizational
factors

Violence by
PSMI

Saturation

Injuries

Champion

Arrests

Community
Characteristics
Social disorganization
Crime rates

Linkage to
treatment as
disposition

Figure 1. Effectiveness of CIT. CIT ¼ Crisis Intervention Team. Adapted from ‘‘Improving Police Response to Persons With Mental Illness:
A Multi-Level Conceptualization of CIT,’’ by A. C. Watson, M. S. Morabito, J. Draine, and V. Ottati, 2008, International Journal of Law &
Psychiatry, 31, pp. 359–368. Copyright 2008 with permission from Elsevier.

increase diversion (Watson & Angell, 2007). These characteristics may be associated with pre-CIT knowledge, attitudes,
and skills as well as the effect of CIT training on outcomes.
Organizational level. When fully implemented, CIT is an
organizational intervention involving shifts in operating practices in relation to persons with mental illness. Hence, organizational factors are important to conceptualize when
considering CIT implementation and effectiveness. We include
two organizational factors in our model, CIT saturation and the
existence of a champion. Recommendations for CIT staffing
ranges from 15% to 25% of all patrol officers to ensure
24/7 CIT coverage (Reuland & Cheney, 2005). However, the
optimal numerical saturation level has not been empirically
tested. Our model conceptualizes saturation as a factor influencing implementation and outcomes. We also consider attitudinal saturation, defined as the extent to which officers
accept CIT as a valuable approach to responding to persons
with mental illness that is supported by their supervisors and
administration. The acceptance of CIT as legitimate and willingness to volunteer may be influenced by the presence of a
‘‘champion,’’ a well-respected peer or superior that sends the
message that CIT participation and cooperation is valued
within the agency. The presence of a ‘‘champion’’ may positively influence both numerical and attitudinal saturation
(Rogers, 2003).
Mental health system level. Another core element of CIT is the
development of linkages between police and mental health providers (CSG, 2002) in order to facilitate diversion of persons
with mental illness from the criminal justice system to appropriate mental health treatment. This can only occur if appropriate mental health resources exist and if officers perceive them

as plausible, efficient, and consistent with resolving the
situation (Finn & Stalans, 2002). Hence, we include both the
availability of mental health services and the officers’ perception of them in our conceptual model.
Community level. Finally, we include community characteristics in our model as the broader social context in which police
interact with persons with mental illness is likely to influence
what police are expected to respond to and resources available
to for doing so. Specifically, we consider conditions that are
indicators of social disorganization and associated with crime,
violence, policing, and outcomes for persons with mental illness (Klinger, 1997; Sampson & Laub, 1990). Variations in
community characteristics related to social disorganization
may shape officers’ workload demands, resources for managing demands, and their ability to effectively implement CIT
responses.
Taking into account the complexities of police work in varied organizational service system and community environments, our conceptualization of CIT effectiveness is
illustrated in Figure 1. In this model, the impact of CIT training
is moderated by availability of mental health services. If appropriate mental health treatment linkages are not (or are perceived as not) available, application of CIT skills may seem
futile. The two organizational factors, saturation and the presence of a champion, also moderate CIT training. Community
characteristics define the context in which officers interact with
members of the public.

Applying the Model in Chicago
In 2005, the CPD began implementation of CIT in 2 of its 25
police districts. Approximately, 40 officers and supervisors

538
Downloaded from rsw.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016

 Watson

539

from each completed 40 hr of training developed and provided
by a team of police academy instructors, community mental
health providers, family members, and consumers. Early success of the program led to the Department’s decision to implement CIT citywide in late 2006.
Designing the study. Given the size and structure of the CPD,
we made the decision early to focus on districts as our organizational unit instead of the department as a whole. While all
districts are subject to the same departmental policies and procedures, each district has its own command structure, culture,
and community and mental health resource context. Ideally,
to study CIT’s effectiveness, we would have designed a randomized controlled trial that randomized districts to implement
CIT and officers in the treatment districts to CIT training.
There were significant barriers to implementing such a methodologically rigorous design, forcing us to be resourceful and
make some trade-offs in design, sampling, and measurement.
First, we were not able to randomize officers, districts, or
calls to CIT. A key element of the CIT model is that officers
must volunteer to undergo CIT training (CSG, 2002). That
way, CIT training is used to build the knowledge and skills
of officers that already have an interest in responding more
effectively to mental health crisis calls. Randomizing officers
to the training would violate a core element of the program.
One option for addressing this issue was to randomize volunteers to training or a waitlist. However, as outside researchers,
we did not have control over personnel assignments to training
or regular duties. For the same reason, randomizing calls to be
handled by CIT-certified officers or districts to implement CIT
were not options. Thus, instead of a control group, our design
included the two CIT pilot districts and two comparison districts each selected because it was similar to one of the pilot
districts in terms of availability of mental health services and
community characteristics but very early in the process of CIT
implementation. Within these districts, we also compared CIT
officers to those that had not completed CIT training.
Sampling and measurement. The options available for identifying and measuring outcomes of calls involving persons with
mental illness presented another barrier. Prior studies of CIT
have sampled mental disturbance calls using existing records
(Steadman et al., 2000; Teller, Munetz, Gil, & Ritter, 2006).
However, the existing documentation of calls in Chicago did
not reliably allow us to identify calls involving persons with
mental illness. At the time of the study, dispatch personnel had
not been formally trained to identify mental health calls up
front and there was no formal mechanism to recode calls later,
if a mental health component was identified. Report forms are
only completed on a subset of calls, for example, when force is
used or a subject is transported to the hospital, and the presence
of a mental health component may not be noted. The consistency and thoroughness with which these forms are completed
is questionable, particularly in busier districts where downtime
to complete paperwork is rare. Given the limitations of using
call-level documentation, we decided against using formal
records as our primary data source. Instead, we chose to sample

officers rather than calls and collect information that would
allow us to conduct analysis using call or officer as the unit
of analysis.
In order To elicit information on mental disturbance calls,
we designed an interview tool that included questions about the
most recent mental disturbance calls officers responded to as
well as all of the mental disturbance calls they responded to
in the past month. We were interested in all calls involving
adults with mental illness. Thus, we instructed officers: ‘‘This
call does not have to necessarily include a hospitalization of the
subject but could have other outcomes, such as arrest, referral
to services or informal action. Additionally, this does not have
to be a call that you documented or did paper work on.’’ This
definition of mental illness relies on the perceptions of officers,
as it is the officer’s perception of illness that matters in identifying a potential subject for CIT intervention.
Officers were asked a similar set of questions about their
most recent and past month calls. Interview items covered
how the subject came to their attention, characteristics of the
subject, the subject’s level of resistance, officer’s use of force,
injuries to the officer and subject, other situational characteristics, the outcome of the call, perception of mental health system resources, and officer use of CIT skills. Many of the items
and scales were adapted from existing measures (see Watson
et al., 2009). The next section of the interview included items
exploring officers’ perceptions of mental health resources in
their district, the CIT program, the organizational support for
CIT, and the existence of CIT champions. The final section of
the interview included questions related to officer demographics, work assignment, and personal familiarity with mental illness.
In order To provide texture and depth to the quantitative call
data obtained from officer interviews, we conducted qualitative
interviews in the study districts to examine perceptions of mental disturbance calls and barriers and facilitators of the CIT program. Additionally, we included available dispatch data and
CIT call forms for the study districts during the study period for
purposes of triangulation.
In summary, our study design involved examining CIT outcomes in two fully implemented pilot districts and two comparison districts selected because they were similar in mental
health resources and community characteristics but much earlier in the CIT implementation process. We sampled CIT and
non-CIT officers from all four study districts and conducted
in person interviews. We also conducted phone follow-up interviews at 1, 3, and 6 months using a slightly abbreviated interview tool. Qualitative interviews were conducted and dispatch
data and CIT forms were requested.
The study districts. Data on district characteristics were compiled from U.S. Census 2000 (U.S. Census Bureau, (2001),
Chicago Department of Public Health (2007) listings of mental
health providers, CPD Annual Report for 2007, and personnel
lists provided by the districts. Pilot district A is located in a
highly disadvantaged, predominantly Black, high-crime area
with relatively few mental health services other than the
539

Downloaded from rsw.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016

 540

Research on Social Work Practice 20(5)

emergency room at the local hospital. At the time of data
collection for this study, 17.4% of the supervisory (Sergeant
[Sgt] and above) staff and 10.14% of nonprobationary patrol
officers were CIT certified. District C, which is immediately
South of District A, served as the comparison. District C is similar to District A in crime rates, community characteristics, and
mental health service availability, although with a few pockets
of slightly less disadvantage. At the time of baseline data collection, only 2.0% of the district’s supervisory staff and 4.7%
of patrol officers were CIT certified.
Pilot District B is less homogenous than Districts A and C
with relatively lower crime rates and social disadvantage and
much higher density of mental health service providers. At the
time of baseline data collection, 19.5% of the supervisory staff
and 13.9% of nonprobationary patrol officers were CIT certified. Immediately west of Pilot District B, Comparison District
D is similar in community characteristics and density of mental
health resources. It shares a designated hospital for police
transports for psychiatric evaluation. At the time of baseline
data collection, 10.81% of the district’s supervisory staff and
3.95% of nonprobationary patrol officers were CIT certified.
Sampling. We used personnel lists to select a total of 80–86
officers from each study district to invite to participate. All CIT
officers (41, 40, 13, and 16, respectively, from Districts A, B,
C, and D) were invited. We used proportionate random sampling based on the number of officers assigned to each watch
to select enough non-CIT officers to make 80–86 in each district for a total of 333 officers invited.
Recruitment. One week prior to beginning data collection,
members of the research team attended all roll calls in each of
the four districts and announced the study. The next day, we
delivered letters to the districts for the selected officers inviting
them to participate and giving them a number to call, if they preferred not to be contacted. Officers were contacted in person
while on duty via their watch commander. Including six that
declined by phone prior to meeting with the researchers, a total
of 58 declined participation. We were unable to meet with
59 officers due to medical leaves, military leaves, and retirements. Our response rate was 65%, if the officers we were unable
to contact are included in the denominator or 79% if they are not.
The sample. Our sample of 216 officers included 170
(78.7%) male and 46 (21.30%) female officers. In terms of race
and ethnicity, our sample roughly reflected the composition of
the CPD (2008) with 109 (50.90%) non-Hispanic White, 73
(34.10%) Black, 37 (17.1%) Hispanic/Latino, 6 (2.8%) Asian
Pacific Islanders (Asian PI), 2 (0.90%) Native American, and
24 (2.9%) other. The mean age of participants was 40.96
(SD ¼ 9.13) and mean years on the force 11.46 (SD ¼ 7.56).
CIT officers on average were about 4 years older (M ¼
43.62, SD ¼ 9.46 vs. M ¼ 39.06, SD ¼ 8.41, t(211) ¼ 3.703,
p ¼ .000) and had about 3 more years of service (M ¼ 13.26,
SD ¼ 7.85 vs. M ¼ 10.15, SD ¼ 7.09, t(214) ¼ 3.053, p ¼
.003) than non-CIT officers. In terms of rank, 183 (84.7%) were
patrol officers, 7 (3.2) were field training officers (FTOs), and
24 (11.1%) were supervisory (Sgt or Lieutenant). In all, 61

(28.20%) worked in District A, 49 (22.70%) in District B,
58 (26.90%) in from District C, and 48 (22.20%) in District D.
A total of 91 participants (42.10%) were CIT trained.
Looking at patrol officers and FTOs (who are more likely to
be first responders than supervisors) only and excluding the
75 officers indicating no past month mental disturbance calls,
the mean number of past month calls involving an adult with
mental illness was 5.14 (SD ¼ 5.20). CIT officers reported
more calls per month (M ¼ 6.43, SD ¼ 6.53) than non-CIT
officers (M ¼ 3.86, SD ¼ 2.93, t(110) ¼ 2.689, p ¼ .008). This
may be because mental disturbance calls are appropriately
being dispatched to CIT officers or that CIT officers are better
able to recognize signs and symptoms of mental illness in
calls not preidentified.
Follow-up interviews and response rates. The study design
included 1, 3, and 6 month phone follow-up interviews, presumably during the officer’s off duty time. At the 1 month
follow-up, we were able to complete interviews with 139
(64.4%) participants. Follow-up rates dwindled at 3 and
6 months with 108 (50%) and 98 (45.4%) participants completing interviews, respectively.
While progressively lower follow-up response rates were
of concern, an additional problem emerged that further limited the amount of data we had for examining CIT’s impact
on calls over time. At baseline, only 130 (60.2%) participants
indicated they had responded to any calls involving a person
with mental illness in the prior month. At 1, 3, and 6 months
respectively only 65 (46.8%), 52 (48.1%), and 52 (53.1%)
reported any past month calls. Those reporting any past month
calls were fairly evening distributed between CIT and nonCIT officers. One explanation is that we sampled across all
assignments, thus our sample included officers on desk and
lock-up duty that may be less likely to respond to mental disturbance calls.
Qualitative sample. In total, 20 officers participating in the
quantitative study also completed qualitative interviews that
explored perceptions of the CIT program, barriers to fully
implementing CIT, and their recent experiences on calls
involving persons with mental illness. These data are allowing us to explore CIT implementation issues as well as differences between CIT certified and non-CIT–certified police
personnel.
Dispatch records and CIT call forms. We obtained dispatch
records on the number of mental disturbance and threaten/
attempt suicide calls in each of the study districts. Looking
across study districts during the baseline data collection period,
dispatch records indicate that there was a total of 97 mental
health–related calls. Our subsample of officers in those districts
reported well over 500 mental health calls during that period.
Even allowing for some duplication in calls reported, we question the reliability of the dispatch records for identifying the
calls of interest. To date, we have not obtained the CIT call
forms. Department personnel are in the process of redacting
subject identifying information contained in the reports before
forwarding copies to the research team.

540
Downloaded from rsw.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016

 Watson

541

Summary of Findings and Lessons Learned
We have conducted several sets of analysis to date that are
reported in detail elsewhere. Here, I briefly review these findings and the ongoing analysis. First, we have examined the
impact of CIT training on outcomes of mental disturbance calls
(Watson et al., 2009). We collapsed call outcomes into three
categories: direct to mental health services without arrest
(includes transport to the hospital for psychiatric evaluation,
transport to other facility, and making referrals), arrest, or contact only (officer resolved the situation informally). Results
indicated that controlling for district and officer and call characteristics, CIT officers directed a significantly greater proportion (18% more) of subjects to services than non-CIT officers.
Several factors moderated the CIT training effect. The CIT
effect for direction to services was largest when subject resistance was low and for officers that reported someone close to
them had a mental illness. There was not a significant direct
effect for CIT on contact only. However, among officers who
possessed a positive view of mental health resources, and those
with prior familiarity with mental illness, CIT training reduced
the use of contact only to resolve calls. Our findings did not
suggest CIT had an impact on arrest decisions. We are currently working on examining district characteristics such as
CIT saturation and mental health resource availability as moderators of the CIT effect on call outcomes.
In another set of analysis, we examined the effect of CIT on
use of force (Morabito et al., 2009). We did not find an effect
for officer CIT training. However, officers (CIT and non-CIT)
working in Districts A and B, both with high levels of saturation, reported using less force than officers in District C, where
CIT saturation was limited. The more highly saturated districts
had more CIT trained Sgts, who supervise CIT and non-CIT
trained officers in the field. Perhaps, all officers in these districts experienced greater support for using de-escalation techniques and responding in a manner that reduced the need for
force.
Additional analyses of the data from this study are underway, which examine elements of our model of CIT effectiveness. These include exploration of the influence of attitudinal
variables and presence of champions, examination of districtlevel effects, and where possible, changes in attitudes and outcomes over time.

Lessons Learned
Although we continue with our analysis of the data from this
study, we are in the initial stages of developing our next project, a multi-city study of CIT that examines the influence of
different mental health system arrangements (availability,
access, and collaboration) on CIT outcomes. This study will
build on our current work in Chicago and will be very mindful
of the lessons we have learned.
First, our approach recruiting officers and administering
baseline interviews worked well. Meeting with district commanders and announcing the study at roll calls ensured officers
knew about the study and that they were allowed to participate.

Officers were generally agreeable to participating in the
interview while on duty. The department was not comfortable
with the manpower burden presented by our conducting all three
follow-up interviews in person with on duty officers. Thus, we
decided to follow-up with officers by phone during their off duty
time. This did not go as well. Clearly, officers were less willing
to give up their personal time to participate. Limiting the number
of follow-ups and conducting them with officers during duty
hours is likely a better strategy. In terms of sampling, it may
be more efficient to exclude officers with assignments that make
them less likely to respond to mental disturbance calls.
At least in Chicago, dispatch documentation does not appear
particularly useful for answering questions about CIT. We are
still working on obtaining CIT form data. However, we have
been advised by the CIT training unit that the forms are not
consistently completed, particularly in the busier districts.
Hence, their usefulness for examining CIT outcomes and triangulation may also be limited. Efforts are underway at CPD to
improve documentation of mental health related calls, so in the
future these may be useful data sources.

Conclusions
As we discovered in our study of CIT, conducting research on
interventions in complex settings is challenging but possible.
As social work researchers, we were particularly well suited
to study an intervention that was designed to address complex
issues involving multiple systems and actors and affecting a
particularly vulnerable population. We developed a conceptual
model of CIT effectiveness and designed a study to begin testing the model and approaches to measuring key components.
With some creativity, skill, and luck, we were able to negotiate
the research process within the law enforcement setting.
To date, our findings suggest that CIT in Chicago is increasing
linkage to mental health services and improving safety in calls
involving persons with mental illness. The strategies and measures we developed and the lessons we learned provide a solid
foundation for future studies. Overtime, we hope to incrementally add to the evidence base on the effectiveness of CIT.
Our findings suggest that CIT may improve outcomes of
police encounters with persons with mental illness, particularly
among officers with more positive views of mental health
resource responsiveness. Social workers, as mental health service providers, can support this process by reaching out to law
enforcement to assist with training and develop cross-system
collaborations to improve outcomes for persons with mental illness, who come in contact with police.
The importance of continuing this work became tragically
clear between the 3 and 6 month follow-up periods of our
study. A veteran police officer from one of our comparison districts was shot and killed with his own gun by a woman with
serious mental illness (Main & Sweeney, 2008). He was not
CIT trained. The woman was shot six times by officers arriving
on the scene as backup. This tragic event underlined the need to
continue our efforts to understand the effectiveness of interventions such as CIT.
541

Downloaded from rsw.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016

 542

Research on Social Work Practice 20(5)

Acknowledgements
This data was provided by and belongs to the Chicago Police
Department. Any further use of this data must be approved by the
Chicago Police Department. Points of view or opinions contained
within this document are those of the author and do not necessarily represent the official position or policies of the Chicago Police Department.
The author would like to acknowledge her co-investigators that contributed significantly to the development and implementation of this studyBeth Angell, Jeff Draine, Melissa Morabito and Victor Ottati.

Declaration of Conflicting Interests
The author(s) declared no conflicts of interest with respect to the
authorship and/or publication of this article.

Funding
This work was supported by NIMH R34 MH081558. The contents are
the sole responsibility of the authors and do not necessarily represent
the official views of the NIH.

References
Chicago Department of Public Health. (2007). Profile of Chicago’s
Mental Health System 2003. Retrieved October 17, 2007, from
http://www.cchsd.org/reports.html
Chicago Police Department. (2008). Annual report 2007: A year in
review. Retrieved October 10, 2008, from www.CityofChicago.
org/Police
Compton, M., Esterberg, M., McGee, R., Kotwicki, R., & Oliva, J.
(2006). Crisis intervention team training: Changes in knowledge,
attitudes, and stigma related to schizophrenia. Psychiatric Services, 57, 1199-1202.
Compton, M. T., Bahora, M. A., Watson, A. C., & Oliva, J. (2008).
Comprehensive review of extant research on Crisis Intervention
Team (CIT) programs. Journal of the American Academy of Psychiatry and the Law, 36, 47-55.
Compton, M. T., Demir, B., Oliva, J. R., & Boyce, T. (2009). Crisis
Intervention Team training and special weapons and tactics callouts in an urban police department. Psychiatric Services, 60,
831-833.
Council of State Governments. (2002). Criminal Justice Mental Health
Consensus Project. Retrieved from http://consensusproject.org/
Deane, M. W., Steadman, H. J., Borum, R., Veysey, B., &
Morrissey, J. (1999). Emerging partnerships between mental health
and law enforcement. Psychiatric Services, 50, 99-101.
Ditton, P. (1999). Mental health and treatment for inmates and probationers. Washington, DC: Bureau of Justice Statistics.
Dupont, R., & Cochran, S. (2000). Police response to mental health
emergencies: Barriers to change. Journal of the American Academy of Psychiatry and Law, 28, 338-344.
Finn, M., & Stalans, L. (2002). Police handling of the mentally ill in
domestic violence situations. Criminal Justice and Behavior, 29,
278-307.
Hanafi, S., Bahora, M., Demir, B. N., & Compton, M. T. (2008).
Incorporating crisis intervention team (CIT) knowledge and skills
into the daily work of police officers: A focus group study. Community Mental Health Journal, 44, 427-432.

Horvitz-Lennon, M., Donohue, J. M., Domino, M. E., & Normand, S.
T. (2009). Improving quality and diffusing best practices: The case
of schizophrenia. Health Affairs, 28, 701-712.
Klinger, D. (1997). Negotiating order in patrol work: An ecological
theory of police response to deviance. Criminology, 35,
277-306.
Main, F., & Sweeney, A. (2008). A final salute for a good cop - 27-yr.
veteran trying to resolve fight on a bus is shot to death with his gun
by woman said to be mentally ill. Chicago Sun Times. July, 3,
2008, p. 2.
Morabito, M. S., Kerr, A. N., Watson, A. C., Draine, J., Angell, B.
(in press). Crisis Intervention Teams and People with Mental
Illness: Exploring the Factors that Influence the Use of Force.
Crime & Delinquency.
Reuland, M., & Cheney, J. (2005). Enhancing success of police-based
diversion programs for people with mental illness. New York, NY:
GAINS Technical Assistance and Policy Analysis Center for Jail
Diversion.
Rogers, E. (2003). Diffusion of innovations (5th ed.). New York, NY:
The Free Press.
Sampson, R. J., & Laub, J. H. (1990). Crime and deviance over the life
course: The salience of adult social bonds. American Sociological
Review, 53, 766-769.
Skeem, J., & Bibeau, L. (2008). How does violence potential relate to
Crisis Intervention Team responses to emergencies? Psychiatric
Services, 59, 201-204.
Steadman, H., Deane, M., Borum, R., & Morrissey, J. (2000).
Comparing outcomes of major models of police responses
to mental health emergencies. Psychiatric Services, 51, 645649.
Strauss, G., Glenn, M., Reddi, P., Afaq, I., Podolskaya, A.,
Rybakova, T., & . . . El-Mallakh, R. (2005). Psychiatric
disposition of patients brought in by crisis intervention
team police officers. Community Mental Health Journal, 41,
223-228.
Teller, J., Munetz, M., Gil, K., & Ritter, C. (2006). Crisis intervention
team training for police officers responding to mental disturbance
calls. Psychiatric Services, 57, 232-237.
Teplin, L. A. (1990). The prevalence of severe mental disorder among
urban male detainees: Comparison with the epidemiologic catchment area program. American Journal of Public Health, 80,
663-669.
University of Memphis CIT Center National Directory. (n.d.).
Retrieved September 22, 2009, from http://cit.memphis.edu/
USA.htm
U.S. Census Bureau. (2001). ‘‘Census 2000 Summary File 3
(SF 3)—Sample Data.’’ Retrieved from http://www.census.gov/
Press-Release/www/2002/sumfile3.html
Watson, A. C., & Angell, B. (2007). Applying procedural justice theory to law enforcement’s response to persons with mental illness.
Psychiatric Services, 58, 787-793.
Watson, A. C., Morabito, M. S., Draine, J., & Ottati, V. (2008).
Improving police response to persons with mental illness:
A multi-level conceptualization of CIT. International Journal of
Law & Psychiatry, 31, 359-368.

542
Downloaded from rsw.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016

 Watson

543

Watson, A. C., Ottati, V. C., Morabito, M., Draine, J., Kerr, A. N., &
Angell, B. (2009). Outcomes of police contacts with persons with
mental illness: The impact of CIT. Administration and Policy in
Mental Health and Mental Health Services Research,
doi:10.107/s10488-009-0236-9.

Wells, W., & Schafer, J. A. (2006). Officer perceptions of police
responses to persons with a mental illness. Policing: International Journal of Police Strategies and Management, 29,
578-601.

543
Downloaded from rsw.sagepub.com at Univ of Illinois at Chicago Library on September 29, 2016