Stepping Up Technical Assistance

Data Outcomes and Technical Assistance Processes Across Five Stepping Up Counties

May 2020

Download PDF

Executive Summary

In 2015, the National Association of Counties, the American Psychiatric Association Foundation and The Council of State Governments Justice Center introduced Stepping Up (SU), a national initiative focused on reducing the number of individuals with mental illness and co-occurring disorders in county jails. The SU initiative is executed at a county level and requires a commitment from community systems and partners, including the county board of commissioners. Through support of the Michigan Mental Health Diversion Council (MMHDC) and the Michigan Department of Health and Human Services, the Wayne State University School of Social Work Center for Behavioral Health and Justice (CBHJ) provided technical assistance to five Michigan counties navigating the Stepping Up framework during the 2018-2019 fiscal year: Charlevoix, Genesee, Jackson, Muskegon, and Washtenaw.

This report describes SU technical assistance services provided and highlights data outcomes across the first five counties engaged with the CBHJ. These outcomes were analyzed from the results of standardized screening instruments collected on a sample of individuals as they booked into each county jail. The screening instrument detected symptoms of severe mental illness (SMI) using the Kessler-6 (K6), substance use disorder (SUD), co-occurring mental illness and substance use disorders (COD), recidivism, and housing stability. Results of current jail processes for identifying SMI are also presented, along with the process for providing follow-up referral and services to those with SMI.

Aggregate data from the SU counties were compared with aggregate data from ten Jail Diversion (JD) counties in 2017 and 2019. The JD counties are those counties who participated in the mental health jail diversion pilot program sponsored by the MMHDC from 2015 2020. Rates of SMI identification and mental health service referral and provision in the five SU counties were similar to the ten JD counties in 20171. It is noted that the JD counties implemented data-driven improvements to county practices between 2017 and 2019 collection periods and, as such, have improved identification, referral, and service practices; the data for SU counties represents the first look at these indicators for these jails.

Findings

  • Of all individuals booked into the five SU county jails, 24% show symptoms of SMI based on the K6 screening tool; 55% screened positive for symptoms of SUD; and 17% screened positive for COD.
  • SU counties identified fewer individuals with SMI (14%) than the need identified by the K6 (24%).
  • K6 positive individuals not identified by the jail report significantly greater mental health symptoms on the K6 and have higher proportions of SUD, compared to those identified with SMI through the jails' standard practices.
  • 73% of those identified by the jail as having an SMI were referred to services, with 43% receiving services.
  • Length of stay averaged 13 days, with individuals with SMI staying significantly longer (21 days) than those without SMI (11 days).
  • Two SU counties completed data integration technical assistance from the CBHJ for ongoing systems monitoring.
  • Two SU counties used their data to secure federal funding for improved mental health services in the jail, with a total of $2.25 million in additional resources to the state.

Recommendations

  • Encourage more counties to pass a SU resolution and create a committee to focus on the SU framework.
  • Encourage jails to implement standardized mental health and substance use screening tools at bookingideally into their jail management systemin order to effectively identify those in need of services. Officer training in identification should continue to be utilized.
  • Modify and formalize protocols for identifying, referring, and serving individuals with behavioral health concerns to ensure timely and appropriate services. Interventions should consider SMI, SUD, and additional risk factors such as homelessness and recidivism.
  • Identify and reduce barriers to data integration and collection among county jails and behavioral health providers.

Next Steps

  • The CBHJ will continue providing technical assistance to additional SU counties.
  • The CBHJ will work with the MMHDC to initiate statewide strategies to improve system barriers.
  • The CBHJ will develop policy briefs in an effort to inform legislators and state administrators about this work, and support needed to continue this work county-by-county.

^ Back to top 


Introduction

Background

Stepping Up (SU) is a national initiative to reduce the number of people with mental illnesses in jails and is the result of a collaboration between the National Association of Counties, The Council of State Governments Justice Center and the American Psychiatric Association Foundation. Since 2015, over 500 counties across the country have committed to SU to reduce the number of individuals with mental illness in jails.

Within Michigan, at least 24 counties endorsed the initiative as of May 2020. While local support for SU is strong, communities often lack sufficient data and technical expertise to effectively engage in the cross-system, outcome-oriented planning necessary to move forward. Specifically, local partners often struggle with insufficient data to properly identify the target population and implement the appropriate system-wide response, to select and implement appropriate evidence-based interventions, and to develop a sustainable system to track and monitor impact.

To assist Michigan counties in achieving the objectives of SU, Michigan Department of Health and Human Services is providing the technical assistance services and expertise of the Wayne State University School of Social Work Center for Behavioral Health and Justice (CBHJ) to counties that have endorsed the SU initiative.

Engagement Process

Technical assistance provided to SU communities is based on the needs of the individual county, but can include:

  • Observation and assessment of county advisory board. 
  • Documentation of SU goals & objectives.
  • Review and mapping of current jail process.
  • Assessment of system efficacy.
  • Review of current screening and assessment tools.
  • Recommendation of evidence-based tools.
  • Data review and validation.
  • Development of baseline data.
  • Development/implementation of data collection protocol.
  • Data collection, cleaning, coding, merging, and analysis.
  • Establish prevalence of SMI in jail.
  • Create context by comparing county with others across the state.
  • Partner action planning session.
  • Presentation of findings and recommendations to partners.
  • Development and/or implementation of ongoing, sustained data collection and analysis.
  • Data system integration consultation and recommendations.
  • Identification of sustainability strategies.

The table below lists the six questions that make up the SU framework and outlines the typical CBHJ technical assistance services provided to address each question, as well as an estimated timeline. The data in this report is derived from the steps outlined under questions two six, although not presented in this order.

Table 1: Center for Behavioral Health and Justice Stepping Up Technical Assistance Framework
Phase 1: Assess and Observe

Question 1: Is Your Leadership Committed?
Introductory Meeting (1-2 Hours)

  • How did Stepping Up evolve in the community?
  • What steps have been taken to date?
  • What agencies are represented on the diversion council? 
  • What are the goals and objectives for Stepping Up?

Questions 2-3: Do You Have Timely Screening and Assessment? Do You Have Baseline Data?
Site visit and jail tour (4-6 hours)

  • Map jail processes from booking to release.
  • Review existing data/data validation.
  • Review or establish baseline data.Development of technical assistance plan.
  • Observe Diversion Council meeting (when applicable).
  • Review current screening & assessment tools.
Phase 2: Assist & Plan
Question 4: Have You Conducted a Comprehensive Process Analysis?
Regular Data Calls & Data Collection (3-6 months)
  • Execute technical assistance plan.
  • Establish data collection protocol.
  • Conduct regular data collection calls with key partners.
  • Assess current system efficacy.
Questions 5-6: Have you Prioritized Policy, Practice & Funding? Do You Track Progress?
Action Planning (3-6 months)
  • Presentation of findings & recommendations to partners.
  • Identify strategies for sustainability.
  • Identify short- and long-term goals.
  • Create context: compare data with other counties across the state.
  • Provide recommendations for sustained data collection and analysis.
  • Develop integrated report to track progress.

 

^ Back to top 


Do You Have Baseline Data?
Prevalence of Serious Mental Illness, Substance Use Disorder, and Co-Occurring Disorder

Prevalence and Demographics of Individuals who are K6 Positive

Overall, SU counties have a similar rate of individuals who scored positive on the Kessler-6 (K6) (24%) as Jail Diversion (JD) 2019 Counties (23%), as seen in Figure 1. The rate of individuals who are K6 positive2 ranges from 17% in County E to 28% in County B. County E has a significantly lower rate of K6 positive individuals (17%) than other SU Counties (23%).

Bar chart showing K6 positive rates by project and county, as described in the previous paragraph.
Figure 1: K6 Positive Rates by Project and County

In SU Counties, females (31%) were significantly more likely to be K6 positive than males (21%) and white individuals (27%) were more likely to be K6 positive than People of Color (20%). No relationship was found between age and being K6 positive (17 to 24: 23%; 25 to 30: 21%; 31+: 26%). The demographics for K6 positive individuals seen in SU Counties largely follow the trends seen in JD 2019 Counties: individuals who are K6 positive are more likely to be female and White, with no particular age group at higher or lower risk.

Figure showing K6 positive rates across demographics, as described in the paragraph above.
Figure 2: K6 Positive Demographics for Stepping Up Counties

Prevalence and Demographics of Individuals Reporting Substance Misuse

Overall, a similar proportion of individuals reported substance misuse in the JD 2019 Counties (55%) and the SU Counties (56%), as seen in Figure 3. The rate of substance misuse ranges from 35% in County E  to 67% in County B. County3 (35%) has a significantly lower rate of substance misuse than SU Counties overall (56%). County B (67%) and County C (63%) both have a significantly higher rate of substance misuse than SU Counties overall (56%). 

Bar chart showing substance misuse by project and county, as described in the paragraph above.
Figure 3: Substance Misuse by Project and County

Individuals in SU Counties that are K6 positive report significantly more substance misuse (68% report either drinking or drugs, or drinking and drugs)those who screen positive for co-occurring mental health and substance use disorders (COD) as compared to those who are K6 negative (51%)4. The relationship between K6 status and substance misuse in SU Counties is similar to the relationship seen in JD 2019 Counties (K6-postive: 69%; K6-negative: 49%). 

Chart showing substance misuse groupings by K6 status as described in the paragraph above.
Figure 4: Substance Misuse Groupings by K6 Status by Project

In SU Counties, individuals who are K6 positive are significantly more likely to report alcohol misuse (49%) and drug misuse (43%) than individuals who are K6 negative (38%; 25%). Older individuals are significantly more likely to report alcohol misuse (31+: 43%) as compared to younger individuals (17 to 24: 36%; 25 to 30: 39%). There was no relationship between age and drug misuse. Males were significantly more likely to report alcohol misuse (42%) than females (36%). However, there was no relationship between sex and drug misuse. White individuals were significantly more likely to report alcohol misuse (45%) and drug misuse (32%) than People of Color (36%; 26%). 

Chart showing substance misuse demographics for Stepping Up Counties, as described in the paragraph above.
Figure 5: Substance Misuse Demographics for Stepping Up Counties

Overall, the demographics for individuals with substance misuse in SU Counties are similar to those that are seen in JD 2019 Counties. In both SU and JD 2019 counties, individuals who are K6 positive, older, male, or White are more likely to report alcohol misuse, and individuals who are K6 positive or White are more likely to report drug misuse. In JD 2019 Counties, females were more likely to report drug misuse than males, while there was no relationship between sex and reported drug use in SU Counties.

Prevalence and Demographics of Individuals with Co-Occurring Mental Health and Substance Use Disorder

Overall, a similar proportion of individuals screened positive for co-occurring mental health and substance misuse (COD) in JD 2019 counties (16%) as in SU counties (17%), as seen in Figure 6. The rate of COD in County E5 (15%) was significantly lower than the rate of COD in SU Counties overall (17%), while the rate of COD in County C (21%) was significantly higher than SU Counties overall (17%). 

Chart showing co-occurring disorder by project and county as described in the paragraph above.
Figure 6: Co-Occurring Disorder by Project and County

In SU Counties, females were significantly more likely to screen positive for COD (22%) than males (15%), and White individuals were significantly more likely to screen positive for COD (20%) than People of Color (12%). There was no significant relationship between age and COD (17 to 24: 16%; 25 to 30: 14%; 31+: 18%).

Figure showing co-occurring demographics for Stepping Up Counties as described in the paragraph above.
Figure 7: Co-Occurring Demographics for Stepping Up Counties

The demographics of individuals who screen positive for COD in SU Counties were similar to those seen in JD 2019 Counties: females and White individuals were more likely to screen positive for COD. However, older individuals in JD 2019 Counties were trending more likely to screen positive for CODa pattern not seen in SU Counties.

^ Back to top 


Other Risk Factors

Prevalence of Recidivism in the Past Year

Overall, a similar proportion of individuals in SU Counties reported past year recidivism (41%) as individuals in JD 2019 Counties (42%), as seen in Figure 8. Recidivism in the past year rates varied in SU Counties. The rate of past year recidivism in County B (32%) is significantly lower than the rate in SU Counties overall (41%), while the rate of past year recidivism in County C (50%) is significantly higher than the rate overall (41%).

Bar chart showing Past Year Recidivism by Project and County as described in the paragraph above.
Figure 8: Past Year Recidivism by Project and County

Prevalence of Housing Insecurity

Overall, significantly more individuals in SU Counties reported being housing insecure (45%) as compared to JD Counties (34%), as seen in Figure 9. Housing insecurity in SU Counties range from a low of 43% in County C to a high of 47% in County D, with no significant differences between any county and SU counties overall (45%). 

Figure of Housing Insecurity by Project and County as described in the paragraph above.
Figure 9: Housing Insecurity by Project and County

Number of Risk Factors by K6 Status

The risk factors included in this analysis are: (1) being K6 positive, (2) screening positive for substance misuse, (3) reporting recidivism in the past year, and (4) being housing insecure. For each individual, these risk factors have been added together. Overall, in SU Counties, individuals who are K6 positive have a significantly higher average number of risks (2.6) than individuals who are not K6 positive (1.4). Relatedly, over half of individuals who are K6 positive have three to four risk factors, as compared to only 11% of those who are not K6 positive. This relationship is similar to the one we see in JD Counties, where K6 positive individuals have a significantly higher average number of risks (2.6) than those who are not K6 positive (1.2). On the whole, the SU Counties are significantly riskier (average number of risks: 1.7) than the JD 2019 Counties (average number of risks: 1.5).

Figure showing Number of Risks by K6 Status by Project as described in the paragraph above.
Figure 10: Number of Risks by K6 Status by Project

^ Back to top 


Have You Conducted a Comprehensive Process Analysis? 

Jail Based Services: Mental Health Identification, Referral, and Services

K6 Status and Jail Identification Status Groupings for Stepping Up Counties

Overall for SU counties, 28% of individuals who are K6 positive were also identified by the jail process-as-usual as having SMI, as seen in Figure 11. Approximately 10% of individuals who were K6 negative were identified by jail as having SMI. 

Flowchart showing K6 Status and Jail Identification Status for Stepping Up Counties as described in the paragraph above.
Figure 11: K6 Status and Jail Identification Status for Stepping Up Counties

Prevalence of K6 Positive and Jail Identification

SU Counties identified significantly fewer individuals as SMI using jail process-as-usual (14%) as compared to JD 2019 Counties using process-as-usual (22%). Alternatively, there was no significant difference in the proportion of individuals who were K6 positive in SU Counties (24%) as compared to JD 2019 Counties (23%).

Bar graph of K6 Positive and Jail ID* Identification by Project as described in the paragraph above.
Figure 12: K6 Positive and Jail ID* Identification by Project

SU Counties varied in the identification of SMI using jail process-as-usual ranging from 2% in County C to 25% in County E where a standardized mental health screen was previously implemented at booking.

Layered bar graph K6 Only and Jail ID* Identification by Project and County as described in the paragraph above.
Figure 13: K6 Only and Jail ID* Identification by Project and County

Characteristics of Jail-ID and K6 Only Groups

As seen in Table 2, individuals who were K6 positive but were not identified by using jail process-as-usual had significantly higher scores on the K6 and significantly more self-reported drug misuse, but were significantly less likely to be referred to services or to receive services. Finally, they were also less likely to have been released from jail at the time data collection ended. There was no significant difference in these groups for age, sex, race, alcohol misuse, housing insecurity, or past year recidivism.

Table 2: Characteristics of Jail ID and K6 Only Groups
  Average K6 Score Drug Misuse Mental Health Referral Mental Health Services Released during data collection period
K6 Only 14.7 45% 8% 13% 70%
Jail ID 8.2 31% 73% 43% 86%

Data Source: Five SU Counties; Jail Identified SMI and K6 Only SMI Identified subsample; Jail Based Services sample: Stepping Up n=514; County A n=72; County B n=116; County C n=103; County D n=95; County E n=128.

Prevalence of Jail Identification

As seen in Figure 14, SU Counties overall identified significantly fewer (14%) individuals using the jail process-as-usual than JD 2019 Counties (22%). For SU Counties, identification rates using the jail process-as-usual varied from 2% in County C to 26% in County A. Given the wide range of identification rates, the mean (14%) does not appear to accurately represent most of the counties well. Only County B (15%) was not significantly different from the overall SU average. County A (26%) and County E (25%) both had significantly higher rates of identification using jail process-as-usual identification than SU Counties overall (14%). County C (2%) and County D (9%) both had significantly lower rates of identification using jail process-as-usual than SU Counties overall (14%).

Bar chart of Jail Identification Rate by Project and County as described in the paragraph above.
Figure 14: Jail Identification Rate by Project and County

Prevalence of Referral for Jail Identified Population

Overall, SU Counties referred significantly fewer of the Jail ID population to services (73%) than JD 2019 Counties (92%), as seen in Figure 15. Given the wide variation in referral rates, the overall SU referral rate does not appear to represent the counties wellonly County B (69%) is not significantly different or trending significantly different from the overall SU Counties average (73%).  County A referred significantly fewer (11%) Jail ID individuals than SU Counties overall (73%), while County C (100%) and County E referred significantly more (98%) than SU Counties (73%) overall. County D's referral rate (87%) was trending significantly higher (p=.07) than SU Counties (73%) overall.

Bar chart of Rate of Referral for Jail Identified Individuals by Project and County as described in the paragraph above.
Figure 15: Rate of Referral for Jail Identified Individuals by Project and County

Prevalence of Service Provision for Jail Identified Population

SU Counties provided services to significantly fewer of their jail identified population (43%) than JD 2019 Counties (57%), as seen in Figure 16. Given the wide variation in services provision rates to the jail identified population, the average SU rate (43%) doesn't appear to represent the counties well. Only County D (57%) was not significantly different from the SU County average. County A (15%) and County E (34%) both had significantly lower rates of service provision to the jail identified population than the SU County average (43%) overall. County B (65%) and County C (100%) both had significantly higher rates of service provision to the jail identified population than the SU County average (43%).

Bar graph of Rate of Services for Jail Identified Individuals as described in the paragraph above,
Figure 16: Rate of Services for Jail Identified Individuals

Jail Identified, Referred, and Served Populations for Stepping Up and Jail Diversion 2019 Project

Bar chart showing 932 individuals were identified as K6 positive in Jail Diversion counties, 92% of which were referred to mental health services and 57% of which were served.
Figure 17: Number of Individuals Identified, Referred, and Served for Jail Diversion 2019 Counties
Bar chart showing 236 individuals were identified as K6 positive in Stepping Up counties, 73% of which were referred to mental health services and 43% of which were served.
Figure 18: Number of Individuals Identified, Referred, and Served for Stepping Up Counties

Length of Jail Stay

Overall, SU Counties had a similar length of stay (13 days) as JD 2019 Counties (12 days). The Jail ID population in SU Counties had significantly longer stays (21 days) than individuals who were not jail identified as SMI (11 days). This is similar to the pattern seen in JD 2019 Counties where individuals who were jail identified as SMI had significantly longer lengths of stays (17 days) than individuals who were not jail identified as SMI (11 days). Individuals in SU Counties who were jail identified as SMI had significantly longer lengths of stays (21 days) than individuals in JD 2019 counties who were jail identified as SMI (17 days).

Bar graph of Length of Stay by Jail Identification in Days by Project as described in the paragraph above.
Figure 19: Length of Stay by Jail Identification in Days by Project

^ Back to top 


Do You Have Timely Screening and Assessment?

Length of Time Between Identification, Referral, and Services for Jail Identified Population

As shown in Figure 20, individuals in SU Counties who were jail identified as SMI had an average length of stay that was 21 days. It took an average of one day from booking for an individual to be identified by the jail as having SMI, an average of eight days from identification for an individual to be referred to mental health services, and an average of seven days from referral for an individual to receive mental health services. Overall, it took an average of 16 days from booking for an individual that is identified as having SMI by the jail to receive services.

Graphic representing Timing of Jail Based Services in Stepping Up Counties as described in the paragraph above.
Figure 20: Timing of Jail Based Services in Stepping Up Counties

^ Back to top 


Do You Track Progress? 

Data Integration Toolkits

The purpose of data integration technical assistance is for counties to be able to monitor progress towards SU framework into the future. There are four recommended measures to track among individuals flagged for behavioral health concerns: 1) serious mental illness (SMI) prevalence at booking, 2) length of stay, 3) connections to treatment, and 4) recidivism. CBHJ staff worked alongside the SU communities to integrate data across systems; between the county's Community Mental Health agency and the Sheriff's Office. A framework for the data integration process can be found in Figure 21.

Graphic of Data Integration Toolkit Framework, showing two agencies funneling datasets into one global data set, with a gate keeper organization integrating the datasets into one dataset. Please contact cbhj@wayne.edu with questions.
Figure 21: Data Integration Toolkit Framework

Each county that participated in data integration received a toolkit report, which included:

  1. A description of the logistical best practices involved in data sharing between disparate systems including decision guides related to informed consent, confidentiality, and bioethics
  2. An index of all pertinent data sources
  3. A detailed codebook of each pertinent variable, scoring instructions for standardized assessments (e.g., K6), and instruction for statistical analyses and reporting.

The integrated toolkit provides a means to improve connections to care and provides SU communities with the data-driven decision capabilities by bolstering reporting through ongoing surveillance of emergent SMI-related trends. The CBHJ has developed integrated report toolkits for Genesee and Charlevoix County Jails and community mental health providers, while continuing to consult with other SU and JD counties. Although progress has been impeded due to COVID-19, the following counties are currently engaged in data integration toolkit development: Barry, Jackson, Kalamazoo, Kent, Livingston, Oakland, and St. Joseph.

^ Back to top 


Have You Prioritized Policy, Practice, and Funding?

Strategic Planning, Development of Goals, and Utilization of Data to Improve Practices

In order to assist with sustainability and action planning, the CBHJ offered each community the opportunity to engage in strategic planning. The number and focus of strategic planning sessions varied depending on each county's needs, but services offered included the following:

  • Mapping services and gaps across the Sequential Intercept Model.
  • Guidance on developing a SU partner group and subcommittees.
  • Coaching on effective meeting facilitation and agenda creation.
  • Observation and feedback on partner group.
  • Development of mission, vision, and goals related to SU.
  • Development of short and long-term goals.

Each data presentation included tailored recommendations from the CBHJ on how the county may improve services for individuals with mental illness or COD. Based on the data and recommendations presented, SU counties prioritized the following:

  • Implementation of standardized screening tools at booking.
  • Enhanced collaboration between partner agencies.
  • Improved data collection and sharing practices.
  • Addition of mental health staff in the jail.
  • Utilizing data to seek funding for additional staffing and programming.

Finally, the technical assistance provided by the CBHJ equipped counties with the data, collaboration, and knowledge needed to seek funding related to the intersection of behavioral health and justice.

  • Two counties applied for and received federal Bureau of Justice Assistance grants using the data developed through the SU project.
  • The CBHJ remains in contact with all counties and because of the relationships and data developed, offers to partner on new projects. For example, two of the counties (Jackson and Muskegon) in this report are now involved in the CBHJ's Opioid Treatment Ecosystem project.

^ Back to top 


This report was made possible through support from the Michigan Mental Health Diversion Council and the Michigan Department of Health and Human Services. 


1  2017 baseline comparisons are not pictured in this report. ^ Back

2  An individual is K6 positive if they scored a nine or higher on the K6 standardized screen.  ^ Back

3  County E does not use the same two question screen for substance use; at booking, the county conducts the TCUDS (Texas Christian University Drug Screen). As such, the county's numbers are likely to be significantly smaller than other counties for potential substance misuse, as their screen has a higher threshold of problematic behavior to be counted as "positive". ^ Back

4  An individual is K6 positive if they scored a nine or higher on the K6 standardized screen. ^ Back

5  County E does not use the same two question screen for substance use; at booking, the county conducts the TCUDS (Texas Christian University Drug Screen). As such, the county's numbers are likely to be significantly smaller than other counties for potential substance misuse, as their screen has a higher threshold of problematic behavior to be counted as "positive".  ^ Back