Improving Outcomes through Coordinated Health and Justice Systems Adams County, Colo.

Adams County encompasses the northeast suburban area of Denver, Colo. The county’s proximity to a large metropolitan area leads to highly transient populations, especially individuals experiencing homelessness. Many of these individuals are familiar faces to county workers and service providers—high-needs residents with complex behavioral health conditions who frequently cycle through jails, homeless shelters, emergency departments and other crisis services. Recognizing that increases in both housing demand and homelessness have no boundaries, Adams County is crafting a regional approach to its housing and homelessness response.

The Adams County Department of Community Safety & Well-Being (CSWB) leverages affordable and open-source technologies to build and use a robust data warehouse for tracking people experiencing homelessness and analyzing the department’s program outcomes. The intention is to identify familiar faces within Adams County and neighboring counties and assess whether existing programs are effectively meeting the needs of vulnerable residents.

REGIONAL PARTNERSHIPS

To better understand the underlying drivers of homelessness and the needs of people experiencing it, CSWB built a data warehouse that incorporates data from all relevant municipal partners, such as shelters and city programs. The CSWB’s Poverty Reduction Unit convened participating partners at the end of 2021 to launch the data warehouse and formalize a collaborative relationship moving forward. Leaders from municipalities within the county signed memoranda of understanding (MOUs) to share their housing and housing services data with Adams County. Participating cities like Aurora, Westminster and the City and County of Broomfield straddle multiple county lines, creating data connections with neighboring counties. Under the MOUs, municipalities allow Adams County to extract data from their Homeless Management Information Systems (HMIS). A core team of staff from each jurisdiction meets monthly to review the data, conduct a rolling gaps analysis and identify changes in the population to determine who within this multi-jurisdiction partnership has the right services to meet the identified needs.

DATA MATCHING AND ANALYSIS

The entirety of CSWB’s warehouse is built with Microsoft SQL Server, FileZilla and R.

• Microsoft SQL Server hosts CSWB’s data warehouse, which includes data from homelessness outreach and HMIS data from three different Adams County municipalities
• FileZilla is the pipeline that funnels data into CSWB’s warehouse from three primary sources:
  • The Metro Denver Homelessness Initiative, Adams County’s continuum of care provider
  • Adams County Sheriff’s Office
  • Adams County’s Department of Human Services; and
• R is an open-source statistical software package that CWSB utilizes to combine row-level records from the disparate sources listed above with its own data warehouse records to identify familiar faces across service providers.

As the numerous datasets lack a consistent primary key identifier of individuals, CSWB’s data team apply R packages such as RecordLinkage, which has an underlying logistic regression engine, to score the probabilities that specific records are reflective of the same individual experiencing homelessness. This technique affords Adams County the ability to begin understanding the full cycle of homelessness that vulnerable residents experience in the county.

HOMELESSNESS OUTREACH

The CSWB’s Poverty Reduction Unit activates the data through its Homelessness Outreach team. The data team pushes the familiar faces matched in R into a dashboard that Homelessness Outreach uses to identify who to provide outreach to and what services they’re most in need of, including:

• Mobile first aid for mental health and/or chronic health services
• Severe Weather Activation Plan hotel/motel vouchers
• Mobile shower and laundry services
• Day Works, a low-to-no barrier stop-gap employment program that offers day labor opportunities for same-day cash pay
• Lyft concierge to provide transportation to and from court hearings or other service locations
• Temporary storage and mail collection, and
• Safe parking lot identification for people experiencing vehicular homelessness.

Homelessness Outreach first used the dashboard to target veterans experiencing homelessness and in the first year, Adams County’s homeless veteran population decreased by 82 percent. The Outreach team will next use the dashboard to target unaccompanied youth. Homelessness Outreach also has general outreach staff who visit encampments and ask everyone if they are interested in signing up for any of the above-listed services.

NEXT STEPS

Adams County is pulling data from local EMS and 211 call systems into the data warehouse to better understand who is using various services. Once the data warehouse is complete, the county plans to send the data back to its partner municipalities. With data specific to their municipality, each city or town will have better insights into their housing insecure and homeless populations, facilitating the creation and adaptation of targeted programs design to improve the lives of familiar faces in their jurisdiction.

NACo would like to thank Max Cercone, Senior Program Evaluator, Department of Community Safety & Well-Being for sharing information on Adams County. Mr. Cercone can be reached at MCercone@adcogov.org.

This case study was created with support from Arnold Ventures as part of the Familiar Faces Initiative, seeking better outcomes and lower incarceration rates for individuals who frequently cycle through jails, homeless shelters, emergency departments and other local crisis services.