Abstract
Crisis stabilization units (CSUs), which offer a range of short-term psychiatric and psychological services, are one of several treatment programs that may create “alternative to arrest” options for law enforcement. Here, we examined the characteristics of the population who was referred to a newly established CSU in its first year of operation and examined the impact of the CSU on regional jail bookings. Administrative medical records and regional jail booking data were merged to form our study sample. Adults who had at least one jail booking and/or one CSU admission during our study period were included. We found that from September 1, 2018 through August 30, 2019, 458 people were admitted into the CSU. Approximately one-third (33.8%) had a jail booking during the study period. In the three months following CSU admission, 4.1% had an increase in jail bookings, 11.1% had a decrease, and 66.2% had no change. CSU patients self-reported high depressive and posttraumatic stress symptoms, while also reporting low quality of life scores overall. We conclude that CSUs may be promising components of jail diversion efforts, providing critical services to populations experiencing significant mental health symptoms and who are at risk for incarceration.
Keywords: Crisis Stabilization Unit, Mental Health, Jail, Incarceration
In March 2017, Arkansas passed the Criminal Justice Efficiency and Safety Act (Act 423) in response to rapid growth in the state’s prison population. Two of the foci of Act 423 centered on diverting people with mental illness away from the criminal justice system and in to community-based care. Specifically, Act 423: 1) required law enforcement officers to participate in training on how to identify and effectively respond to people in the midst of a mental health crisis; and 2) created local 16-bed Crisis Stabilization Units (CSUs) where law enforcement could divert people with mental illness who committed misdemeanor offences away from county jails. Research regarding whether and how much CSUs aid in overall diversionary efforts (e.g., whether there are upstream reductions in the number of people incarcerated) has been the subject of minimal empirical study. Patient-level data are especially limited. Here, we describe the results of a novel investigation into the short-term legal outcomes of people who received services at one of the four CSUs created by Act 423. We linked CSU patient and jail booking data to examine patterns of arrest and descriptive information on the demographic and psychological characteristics of patients served during the first year of this CSU’s operation.
Intersection of Mental Health and the Legal System
Federal-level shifts in funding from institutional to community-based settings that began in the 1960s rapidly transformed the landscape of mental health services in the United States (U.S.). Less than 20 years after President Kennedy signed the Community Mental Health Act of 1963, the U.S. reduced the number of people committed to inpatient psychiatric hospitals by 75.0% (Sheffield, 2013). However, despite eventual policy changes that financially incentivized community-based services, the public mental health system emerged underfunded and understaffed to meet the needs of the population living with mental illness in the U.S. (Lamb & Weinberger, 2005; Sharfstein, 2000).
Moreover, the 1970s brought a series of new policies that funneled an unprecedented number of Americans—especially Black Americans—into prisons and jails (Wildeman & Wang, 2017). Mass incarceration remains an ongoing public health crisis (Kinner & Wang, 2014; Wildeman & Wang, 2017), with costs exceeding $50 billion annually (Kyckelhahn, 2012) and people living with mental illness disproportionately incarcerated. Over half of people who are incarcerated have a mental health condition aside from substance use disorder (James & Glaze, 2006), while the Bureau of Justice Statistics also estimates that approximately sixty percent of people in state prisons and who have received adjudication in jails meet diagnostic criteria for substance use disorder (Bronson & Carson, 2019).
Crisis Stabilization Units: Services and Existing Research
CSUs provide short-term hospitalization for people experiencing acute mental health crises. Although specific services and program parameters vary by location, CSUs generally offer combinations of assessment, pharmacotherapy, case management, respite care, and counseling. CSU staff are multidisciplinary, and may include doctoral-level providers (i.e., psychiatrists, psychologists), nurses, social workers, and paraprofessionals (e.g., psychiatric technicians). Patients receive 24/7 nursing care, including medication administration and basic medical care. CSUs primarily aim to: (1) reduce psychological symptoms and/or distress to a level that can be managed in outpatient settings and (2) connect people with needed outpatient services (e.g., mental health centers, substance use treatment facilities). Since law enforcement officers are typically first responders to mental health crises, it is possible that CSUs can reduce arrests by diverting persons from jail and providing an alternative to the emergency department (ED).
Although research on CSUs is limited and primarily done at the systems-level, existing data are promising. An extensive report which examined the impact of CSUs in Minnesota found that CSU decreased emergency health care utilization and increased outpatient mental health service utilization (Leite Bennett & Diaz, 2013). Also, some evidence shows that CSUs are a cost-effective alternative to inpatient treatment in psychiatric hospitals or EDs (Adams & El-Mallakh, 2009), result in economic savings (Leite Bennett & Diaz, 2013), and generate a net positive return on investment (Leite Bennett & Diaz, 2013).
The Current Study
Few studies have examined the criminal justice involvement of people who receive services in CSUs. There is also limited knowledge of the potential impact that CSUs have on arrests. Our study aims were to: 1) Describe characteristics of adults who were booked in to a regional jail and/or regional CSU during the CSU’s first year of operation, 2) Examine differences between CSU utilizers who did and did not have a recent jail booking, and 3) Examine factors associated with reduced bookings after CSU admission.
Methods
Setting
The Pulaski County Regional Crisis Stabilization Unit (PCRCSU), located in Little Rock, Arkansas was the second CSU to open in Arkansas as a result of Act 423. Pulaski County is the most populous county in Arkansas and was chosen as the site for our study because the Pulaski County Regional Detention Facility (PCRDF) is also the largest jail in Arkansas, booking over 15,000 unique individuals annually. In the years leading up to this study, the jail rate in Pulaski County was approximately 416 per 100,000 population, well above the national average of 326 per 100,000 population (Vera Institute of Justice, n.d.).
The University of Arkansas for Medical Sciences (UAMS) partnered with Pulaski County government stakeholders to create the PCRCSU, which became operational on August 20, 2018. PCRCSU receives referrals from six central Arkansas counties via law enforcement officers encountering people in the community experiencing mental health crises, including chemical dependency issues. Referrals also come from the community mental health centers and local hospitals. The PCRCSU clinical director is a licensed clinical psychologist (the fourth author). Services provided on the unit include nursing and psychiatric measures, such as medication initiation and detox services, and brief psychological interventions including group and individual psychotherapy. Extensive case management and discharge planning are also offered.
Data Sources
Administrative Data on Jail Bookings.
Data were obtained and merged through data-use agreements between UAMS and Pulaski County Government. The PCRDF provided information about all individuals with a jail booking that occurred in calendar year of 2018 and 2019 (i.e., name, date of birth, sex, race, booking date, and release date when applicable). Because one purpose of the study was to examine changes in jail booking frequency following CSU admission, we retained all bookings that occurred between June 1, 2018 and November 30, 2019 (i.e., 3-months before and 3-months after our 1-year study period of September 1, 2018 to August 31, 2019).
Administrative Data from CSU Medical Records.
We merged administratively-acquired jail data with patient-level data from the PCRCSU’s first year of operation that was compiled during the program evaluation mandated by the state of Arkansas. Program evaluation data included descriptive information about the patient population and their stay on the unit, as well as responses to self-reported symptoms and quality of life surveys administered to a subsample of patients (N = 209) during admission.1 The self-report surveys assessed depressive symptoms (Patient Health Questionnaire [PHQ-9]; Kroenke et al., 2001), trauma exposure (Life Events Checklist [LEC-5]; Weathers et al., 2013), posttraumatic stress disorder symptoms (PTSD Checklist [PCL-5]; Blevins et al., 2015), substance use (TCU Drug Screen-5; Knight et al., 2002), and quality of life (Quality of Life Scale [QOLS]; Burckhardt & Anderson, 2003). More information on each of these measures is described below.
Self-reported depressive symptoms.
The PHQ-9 (Kroenke et al., 2001) is a 9-item self-report measure that assesses depression symptoms. Participants rate how often they have experienced the listed symptoms in the past two weeks from 0 (Not at all) to 3 (Nearly every day). Total scores range from 0-27, with higher scores indicating more severe depressive symptoms.
Trauma exposure.
The LEC-5 (Weathers et al., 2013) is a 17-item self-report checklist that assesses traumatic event exposure in a manner consistent with how trauma is operationalized in the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5; American Psychiatric Association, 2013). Participants indicate if they have experienced each type of event inquired about by classifying their exposure as: 1) Happened to me 2) Witnessed it 3) Learned about it 4) Not Sure 5) Doesn’t apply. For the purpose of this study, we included only items endorsed as “happened to me” or “witnessed it” as affirmative responses for that type of trauma exposure.2
Posttraumatic Stress Disorder (PTSD) Symptoms.
The PCL-5 (Blevins et al., 2015) is a 20-item self-report measure that assesses PTSD symptom severity. Participants rate the extent of how much they have experienced each item on a five-point Likert scale, from 0 (Not at all) to 4 (Extremely). Total scores range from 0-80, with higher scores indicating more severe PTSD symptoms.
Substance Use.
The TCU-5 (D. K. Knight et al., 2018) is a self-report measure that screens for mild to severe substance use disorder in a manner that is consistent with symptom criteria according to the DSM-5. On the measure, substance use severity scores range from 0-11, with each 1-point increase indicating one additional substance use disorder criteria that has been endorsed (i.e., a score of 11 would indicate that the respondent has endorsed all DSM-5 substance use disorder criteria). Higher scores thus indicate more severe substance use disorder symptomology.
Quality of Life.
The QOLS (Burckhardt & Anderson, 2003) is a 16-item self-report measure that assesses one’s perceived quality of life. Participants rate the extent of how satisfied they are with areas in their life such as relationships, work, recreation, and health on a 7-point scale from 1 (Terrible) to 7 (Delighted). Total scores range from 16-112, with higher scores indicating greater perceived quality of life.
The EQ-5D-5L (Herdman et al., 2011) is a 6-item self-report measure that assesses health-related quality of life. The measure contains 5 Likert type items assessing functioning in various dimensions of health (e.g., mobility, self-care) and 1 visual analog scale item asking the respondent to self-report their current health on a scale from 0 (The worst health you can imagine) to 100 (The best health you can imagine). We used the latter item to capture participants’ perceptions of their health-related quality of life in this study.
Data Cleaning and Linking
Each dataset was cleaned to remove duplicate and incomplete records. Individuals in the jail data set were assigned a unique enrollment number based on name, date of birth, race and sex. The final cohort included individuals aged 18 and over. Those who completed an intake and were admitted to the CSU were labeled as “CSU utilizers”; those without an encounter were labeled as “non-utilizers.” Once the dataset was cleaned, all names were replaced with a unique identifier. For computing internal consistency on self-report measures, people who were missing more than two items for any measure were excluded.3
Analytic Strategy
We performed a series of retrospective analyses to address our study aims. First, we descriptively analyzed the administrative data from the PCRCSU medical records from the first year of operation to characterize the utilizer population’s sociodemographic characteristics, trauma history, mental health symptoms and quality of life during intake, and discharge disposition. We also: 1) compared PCRCSU utilizers with and without a jail booking on these variables using t-tests for continuous variables and chi-square tests for nominal and dichotomous variables and 2) summarized known information about the broader population of individuals booked in to the PCRDF over the study period (September 1, 2018 and August 31, 2019). Internal consistency for patient self-report measures were examined using Cronbach’s alpha. Mental health self-report measures with up to two missing values were scored, replacing missing values with the average score of the completed items.
Next, we sought to examine factors associated with reduced bookings after CSU admission. We analyzed those who had increases in jail bookings in the 3 months following their PCRCSU admission compared to the 3 months prior. For individuals with more than one admission to the PCRCSU during this period, we used their first visit to determine this time period. We used descriptive statistics to characterize the population and to examine differences, with respect to demographic characteristics between those who were jail “increasers,” jail “decreasers,” and those who had “no change” in their number of jail bookings. We dichotomized variables and analyzed them using Wald χ2 tests or Fisher’s exact test where appropriate, with statistical significance defined as p < .05. All analyses were conducted using SAS version 9.4.
Results
Characteristics of CSU Utilizers
Over the one-year study period (September 1, 2018 – August 31, 2019), 461 unique people were admitted to the CSU.4 Approximately two-thirds of those admitted were male (62.4%) and the overwhelming majority were non-Hispanic (95.0%). Most were White (57.2%) or Black (40.2%), with less than 3.0% self-reporting other race categories. The average length of stay was approximately 2.5 days (i.e., M = 59.12 hours, SD = 48.59 hours; Mdn =47 hours). More than two-thirds (69.7%) were referred by community mental health centers, 28.4% by law enforcement, and 2.0% by other (e.g., self, family, hospital). Of these admissions, over half (52.8%) were discharged somewhere other than a home or residence (e.g., to a homeless shelter, substance use treatment center, unknown location). Table 1 contains a complete description of sociodemographic characteristics of CSU utilizers.
Table 1:
Characteristics of CSU utilizers with and without jail bookings in the 3-months before and/or after admission between September 1, 2018 – August 30, 2019
| All CSU Utilizers N=458 | CSU Utilizers with at least 1 jail booking N=155 | CSU Utilizers without a jail booking N=303 | ||
|---|---|---|---|---|
| M or n (SD or %) | M or n (SD or %) | M or n (SD or %) | p * | |
| Demographics | ||||
|
| ||||
| Age | 37.1 (11.7) | 37.8 (±11.5) | 36.6 (±11.8) | 0.37 |
| Sex | ||||
| Male | 286 (62.4%) | 110 (71.0%) | 176 (58.1%) | <0.01 |
| Female | 172 (37.6%) | 45 (29.0%) | 127 (41.9%) | |
| Race | ||||
| White | 262 (57.2%) | 72 (46.5%) | 190 (62.7%) | <0.01 |
| Black | 184 (40.2%) | 81 (52.3%) | 103 (34.0%) | |
| Other | 12 (2.6%) | 2 (1.3%) | 10 (3.3%) | |
| Ethnicity | ||||
| Non-Hispanic | 435 (95.0%) | 152 (98.1%) | 283 (93.4%) | 0.03 |
| Hispanic | 23 (5.0%) | 3 (1.9%) | 20 (6.6%) | |
| Mental Health | ||||
| Depression (PHQ-9) | 15.20 (7.28)a | 14.81 (±6.65) | 15.39 (±7.59) | 0.61 |
| PTSD (PCL-5) | 42.20 (46.47)b | 35.85 (±65.62) | 45.08 (±34.52) | 0.36 |
| Substance Use (TCU-5) | 4.93 (4.19)c | 5.37 (±4.01) | 4.73 (±4.27) | 0.33 |
| Quality of Life | ||||
| General (QOLS) | 64.80 (24.58)d | 63.42 (±25.39) | 65.35 (±24.33) | 0.62 |
| Health-related (EQ-5D-5L) | 61.97 (25.70)e | 58.67 (±28.79) | 63.15 (±24.50) | 0.29 |
| Discharge | ||||
| Disposition | ||||
| Home/residence | 216 (47.2%) | 58 (37.4%) | 158 (52.2%) | <0.01 |
| Homeless shelter | 105 (22.9%) | 41 (26.5%) | 64 (21.1%) | |
| Substance use treatment center | 61 (13.3%) | 15 (9.7%) | 46 (15.2%) | |
| Other | 76 (16.6%) | 41 (26.5%) | 35 (11.6%) | |
p is for the comparison between CSU utilizers with and without recent jail bookings.
PHQ-9: 189 individuals had data, range 0-27, and α = 0.86.
PCL-5: 157 individuals had data, range 0-80, and α = 0.93.
TCU-5: 192 individuals had data, range 0-11, and α = 0.93.
QOLS: 209 individuals had data, range 16-112, and α = 0.94.
EQ-5D-5L: 197 individuals had data and alpha is not computed because the measure is a single-item scale ranging from 0-100.
A sub-sample of those admitted to the PCRCSU completed self-report measures during their admission. We were only able to analyze this subsample because some patients were not administered self-report measures during their stay. The CSU staff were not able to administer measures if the patients had a very brief length of stay or if the patients were unable or unwilling to complete them during their admission. Staffing patterns also contributed to some missed measures. In total, 45.0% of people admitted at least once completed a self-report measure. There were no differences between those who completed a self-report and those who did not based on gender (p = 0.38), race (p = 0.40), or ethnicity (p = 0.85). Those who completed a self-report had longer length of stay (median 69.5 hours, IQR 44-104) than those who did not (median 33 hours, IQR 16-75) (p < 0.05).
Nearly all patients who completed the measures (92.9%) reported exposure to at least one traumatic event in their lifetime. The mean number of trauma categories experienced or witnessed was 6.57 (SD = 3.94), where 66.7% of the respondents reported 5 or more traumas. Table 1 also displays descriptive statistics for patients’ responses to the self-report measures.
Characteristics of Local Jail Population and Intersection with CSU Utilizers
Jail Population Characteristics
Over the study period, the total jail population was 15,857. The majority were male (71.1%) and the average age was 35.6 (SD = 14.13). With respect to race/ethnicity, which were only available as a combined variable in jail records, 37.0% were White, 58.5% were Black, were 3.6% Hispanic, and less than 1.0% were another racial/ethnic category. The average number of jail bookings was 1.88 (SD = 2.56).
Intersection between Jail and CSU Populations
After linking jail and PCRCSU records, administrative data revealed that 155 people who had utilized the CSU during the one-year study period (33.8%) also had a jail booking between June 1, 2018 and November 30, 2019 (i.e., within the period from 3 months before to 3 months after the target evaluation period).
Comparison between CSU utilizers with and without a jail booking.
CSU utilizers with and without a jail booking in the 3-months before and/or after their CSU admission in the selected period (between September 1, 2018 – August 30, 2019) differed on many of the variables that we examined. There was a statistically significant association between sex and jail booking status such that men admitted to the CSU were more frequently found to have had a jail booking than women (χ2 = 7.26, p < 0.01). There were also statistically significant associations between race and ethnicity and jail booking status such that Black CSU utilizers and non-Hispanic CSU utilizers were more frequently found to have had a jail booking (χ2=14.83, p > 0.01 and (χ2= 4.68, p < 0.01, respectively). Age did not significantly differ between PCRCSU utilizers with and without a jail booking. On the clinical variables, there were no statistically significant differences among PCRCSU utilizers with and without a jail booking on any self-report measure. However, there was a statistically significant difference between discharge disposition and jail booking status such that individuals with a jail booking were more frequently discharged to a homeless shelter or other/unknown location than those without a jail booking. CSU utilizers without a jail booking were more frequently discharged to a home/residence or to a substance use treatment center than those with a jail booking.
Factors Associated with Changes in Jail Bookings for CSU Utilizers
Characteristics of utilizers of the PCRCSU in the first year of operation were observed between those who had an increase in jail booking (n = 19) after their first CSU admit date and those who had either a decrease (n = 51) or who had no change (n = 388) (see Table 2). The two groups did not significantly differ in regard to gender, age, race, ethnicity or across all mental health measures. However, those who had an increase in jail bookings after their first PCRCSU admission (increasers) had a significantly higher number of bookings (p < 0.01) in the 3 months after their initial CSU admission (M = 2.42) than the decreasers (M = 0.20) and those who had no change in their bookings (M = 0.01). For overall jail bookings, increasers also had a significantly higher average number of total bookings than those in the no-change group during the entire study period. Lastly, there was a statistically significant (p < 0.01) difference in distribution of discharge disposition across increasers, decreasers, and individuals who had no change in their number of bookings.
Table 2.
Characteristics of CSU utilizers by jail booking status
| No-change n = 388 (84.7%) |
Decreasers n = 51 (11.1%) |
Increasers* n = 19 (4.2%) |
||
|---|---|---|---|---|
| M or n (SD or %) | M or n (SD or %) | M or n (SD or %) | p | |
| Age | 36.81 (+11.63) | 40.12 (±12.25) | 35.47 (±10.30) | 0.12 |
| Sex | ||||
| Male | 236 (60.8%) | 37 (72.6%) | 13 (68.4%) | 0.23 |
| Female | 152 (39.2%) | 14 (27.5%) | 6 (31.6%) | |
| Race | ||||
| White | 229 (59.0%) | 25 (49.0%) | 8 (42.1%) | 0.16 |
| Black | 147 (37.9%) | 26 (51.0%) | 11 (57.9%) | |
| Other | 12 (3.1%) | 0 (0.0%) | 0 (0.0%) | |
| Ethnicity | ||||
| Non-Hispanic | 366 (94.3%) | 50 (98.0%) | 19 (100.0%) | 0.56 |
| Hispanic | 22 (5.7%) | 1 (2.0%) | 0 (0.0%) | |
| Mental Health | ||||
| Depression (PHQ-9) | 15.51 (±7.43) | 13.62 (±6.05) | 12.25 (±6.16) | 0.30 |
| PTSD (PCL-5) | 42.24 (±48.67) | 41.25 (±23.72) | 43.20 (±6.91) | 0.99 |
| Substance Use (TCU-5) | 4.82 (±4.20) | 5.60 (±3.94) | 5.78 (±4.60) | 0.65 |
| Quality of Life | ||||
| General (QOLS) | 65.43 (±24.55) | 61.6 (±23.19) | 56.11 (±27.47) | 0.50 |
| Health-related (EQ-5D-5L) | 62.53 (±25.72) | 52.00 (±26.38) | 66.60 (±22.86) | 0.36 |
| Jail Bookings | ||||
| Total bookings | 0.10 (±0.44) a | 2.51 (±2.77) | 2.95 (+2.57) a | 0.03 |
| 3 months before CSU admit | 0.01 (±0.07) a | 1.71 (±1.75) b | 0.53 (±1.17) a b | < 0.01 |
| 3 months after CSU admit | 0.01 (±0.07) a | 0.20 (±0.72) b | 2.42 (1.68) a b | < 0.01 |
Increasers are the reference group
No-change group and increasers differ
Decreasers and increasers differ
Discussion
We examined the impact of a newly opened CSU on jail diversion in Arkansas’s most populous county. We found that individuals admitted to the PCRCSU in its first year of operation reported high levels of depressive symptoms, posttraumatic stress symptoms, and substance use severity, as well as low overall quality of life. Nearly all were trauma exposed, many to multiple events. Many had not been booked into the jail during the overall study period (66.2%).
Importantly, the CSU population was reflective of the racial makeup of its county (i.e., according to the U.S. Census Bureau 2019, Pulaski County is 57.2% White and 37.9% Black (QuickFacts United States, 2019), which nearly matches the CSU utilizer population during the study period. However, for the jail booking population, Black individuals were disproportionally booked at 52.3%, while White individuals were at 46.5%. This is consistent with existing evidence that Black Americans are disproportionately incarcerated.
Follow-up analyses examining patterns of jail booking before and after PCRCSU utilization revealed that most PCRCSU utilizers maintained or decreased the number of jail bookings following admission (95.9% of sample examined, including people with no bookings either before or after PCRCSU admission). However, a small sub-sample who had significantly more jail bookings both before and after their PCRCSU admissions than their counterparts did significantly increase their jail bookings following PCRCSU admission. People who increased jail bookings and those who did not increase following their PCRCSU admission did not differ on any variables we were able to compare. Additional services or strategies—particularly those geared toward individuals who are unhoused—may be required to address factors leading to jail booking for PCRCSU utilizers that are at risk for continuing interactions with law enforcement.
Limitations of the present study included that the administrative nature of the study data limited our ability to probe for factors that may differentiate CSU utilizers from the people who were booked into the jail and did not use the PCRCSU. Our reliance on jail bookings in a single regional jail means that more people than we recorded could have had criminal justice involvement outside of the catchment area. While this is important to consider, we believe this limitation is offset by the low rate of referrals from other counties outside of Pulaski County (where the PCRCSU and jail were both located). In fact, there were only 18 admissions that were referred from outside counties during the one-year study period. Additionally, jail bookings were our singular measure of criminal justice involvement and analyses were unable to examine factors such as incarceration duration. Furthermore, our pre- and post-CSU analysis was limited to only three months due to the duration of our jail data. Future work in this area should extend the pre- and post-periods to assess for stability of our findings. Finally, while using self-report symptom measures to conduct further sub-group analyses would have been ideal, completion rates were lower than expected and likely were confounded with other factors (e.g., length of stay on unit, symptomology), thus precluding their use.
Future Directions
Given the nascent nature of research on CSUs broadly and the limited scope of the current study specifically, there remains a need for additional research on the role(s) that CSUs can play in jail diversion and health service engagement. Studies that longitudinally evaluate post-discharge outcomes to determine if CSUs are effective are especially needed, both in 1) connecting individuals to long-term, community-based mental health treatment and 2) reducing future contact with law enforcement. Studies that seek to replicate the results of our study in additional jurisdictions would also be helpful, given the preliminary nature of this investigation and the need for national-level data that could inform policy efforts. Finally, studies on the factors related to the implementation of CSUs would also be helpful, given that it is likely that local policy on eligibility for diversion, admission requirements, and available services vary across geographic areas. Implementation-focused research could identify common challenges to initiating and maintaining CSU services, while providing information for other jurisdictions looking to implement similar programs in the future.
Conclusions
CSUs not only provide alternatives to incarceration, but can also support recovery and provide connection to community resources to those who are experiencing acute mental health crises. The rise of CSUs in communities could lower jail booking rates, while also providing treatment and consultation for individuals who need it. Such a movement would require strong partnerships between legal system stakeholders, including law enforcement, judiciaries, mental health providers, medical centers, and the state departments of health. Our results provide preliminary data for the promising nature of CSUs as an initiative that may provide benefit to all involved.
Impact Statement:
This study demonstrates that Crisis Stabilization Units (CSUs) may be promising components of jail diversion efforts. We provide data on the characteristics and criminal justice outcomes of CSU patients, which is lacking in the literature to date and which will be useful in informing policy and program decisions.
Disclosures and Acknowledgements:
The authors wish to thank the Pulaski County Regional Crisis Stabilization Unit team, patients, providers, and community partners. The authors have no conflicts of interest to report.
Financial Support:
This work was supported by the National Institute on Drug Abuse at the National Institutes of Health P30DA040500-04 (PI: Schackman), R25DA037190-05 (PI: Beckwith), K23DA048162 (PI: Zielinski), and K01DA051684 (PI: Barocas).
Footnotes
Although the goal was to administer self-report measures to all patients beginning in October 2019, limited staff time, brief lengths of stay, and occasional inability or refusal to complete the measures contributed to less than the full sample completing all measures. Our sense is that brief length of stay was the most common factor resulting in non-completion.
Events that are directly experienced or witnessed are universally consistent with the operationalization of trauma exposure in the DSM-5. Events that are “learned about” are only consistent under particular conditions (e.g., learning about the sudden violent death of a loved one, repeated exposure to aversive details of an event in one’s occupation). We did not include events that participants reported only as “learned about” given that this study utilized administrative data and thus further assessment of those events for consistency with DSM-5 criteria was not possible.
Two people were excluded from calculations for PHQ-9 and PCL-5, and seven people were excluded for the QOLS.
Three individuals were excluded from further analysis due to missing data on key variables. Some individuals were admitted to the unit multiple times. A total of 55 people had more than one admission to the PCRCSU during the study period: 42 had two admissions, 12 had three admissions, and one person had four admissions. However, we only included the information from each unique individuals’ first admission.
References
- Adams CL, & El-Mallakh RS (2009). Patient outcome after treatment in a community-based crisis stabilization unit. Journal of Behavioral Health Services and Research, 36(3), 396–399. 10.1007/s11414-008-9141-3 [DOI] [PubMed] [Google Scholar]
- American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). 10.1176/appi.books.9780890425596 [DOI] [Google Scholar]
- Blevins CA, Weathers FW, Davis MT, Witte TK, & Domino JL (2015). The Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5): Development and initial psychometric evaluation. Journal of Traumatic Stress, 28(6), 489–498. 10.1002/jts.22059 [DOI] [PubMed] [Google Scholar]
- Bronson J, & Carson A (2019). Prisoners in 2017. Washington, DC: US Department of Justice, Office of Justice Programs, Bureau of Justice Statistics, 2019. [Google Scholar]
- Burckhardt CS, & Anderson KL (2003). The Quality of Life Scale (QOLS): Reliability, validity, and utilization. Health and Quality of Life Outcomes, 1, Article 60. 10.1186/1477-7525-1-60 [DOI] [PMC free article] [PubMed] [Google Scholar]
- James DJ, & Glaze LE (2006). Mental Health Problems of Prison and Jail Inmates. In Bureau of Justice Statistics Special Report. [Google Scholar]
- Herdman M, Gudex C, Lloyd A, Janssen M, Kind P, Parkin D, Bonsel G, & Badia X (2011). Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Qual Life Res, 20(10), 1727–1736. 10.1007/s11136-011-9903-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kinner SA, & Wang EA (2014). The case for improving the health of ex-prisoners. American Journal of Public Health, 104(8), 1352–1355. 10.2105/AJPH.2014.301883 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Knight DK, Blue TR, Flynn PM, & Knight K (2018). The TCU drug screen 5: Identifying justice-involved individuals with substance use disorders. Journal of Offender Rehabilitation, 57(8). 10.1080/10509674.2018.1549180 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Knight K, Simpson D, & Morey JT (2002). An evaluation of the TCU Drug Screen. In US Department of Justice Office of Justice Programs United States of America. [Google Scholar]
- Kroenke K, Spitzer RL, & Williams JBW (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613. 10.1046/j.1525-1497.2001.016009606.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kyckelhahn T (2012). State Corrections Expenditures, FY 1982-2010. Bureau of Justice Statistics, US Department of Justice, December. https://www.bjs.gov/content/pub/pdf/scefy8210.pdf [Google Scholar]
- Lamb HR, & Weinberger LE (2005). The shift of psychiatric inpatient care from hospitals to jails and prisons. In Journal of the American Academy of Psychiatry and the Law (Vol. 33, Issue 4, pp. 529–534). [PubMed] [Google Scholar]
- Leite Bennett A, & Diaz J (2013). Crisis stabilization claims analysis: Technical report Assessing the impact of crisis stabilization on utilization of healthcare services. www.wilderresearch.org [Google Scholar]
- QuickFacts United States. (2019). United States Census Bureau. [Google Scholar]
- Sharfstein SS (2000). Whatever Happened to Community Mental Health? Https://Doi.Org/10.1176/Appi.Ps.51.5.616, 51(5), 616–620. 10.1176/APPI.PS.51.5.616 [DOI] [PubMed] [Google Scholar]
- Sheffield W (2013). The Community Mental Health Act of 1963. www.ymadvocacy.org/the-community-mental-health-act-of-1963/ [Google Scholar]
- Vera Institute of Justice. (n.d.). Incarceration Trends. http://trends.vera.org/incarceration-rates?year=2015
- Weathers FW, Litz BT, Keane TM, Palmieri PA, Marx BP, & Schnurr PP (2013). The PTSD Checklist for DSM-5 (PCL-5). National Center for PTSD, 5(August), 2002. 10.1037/t02622-000 [DOI] [Google Scholar]
- Wildeman C, & Wang EA (2017). Mass incarceration, public health, and widening inequality in the USA. The Lancet, 389, 1464–1474. 10.1016/S0140-6736(17)30259-3 [DOI] [PubMed] [Google Scholar]