Brad Kaeter, Hennepin County, MN Criminal Justice Coordinating

Report
Session II: Convening
Stakeholders and
Conducting Cross System
Data Matches
Moderator: Kim Keaton, Corporation for Supportive Housing
Panelist: Brad Kaeter, Hennepin Co. Criminal Justice
Coordinating Committee
Panelist: Keith Corry, San Diego Housing Commission
June 7, 2012
Three Pillars, Nine Steps
Data-Driven ProblemSolving
Policy and Systems
Reform
Targeted Housing and
Services
Cross-system data match to
identify frequent users
Convene interagency and
multi-sector working group
Create supportive housing
and develop assertive
recruitment process
Track implementation
progress
Troubleshoot barriers to
housing placement and
retention
Recruit and place clients
into housing, and stabilize
with services
Measure outcomes/impact
and cost-effectiveness
Enlist policymakers to bring
FUSE to scale
Expand model and house
additional clients
Identifying Stakeholders
Stakeholder/
Partner
Data
match?
County leadership (commissioners,
managers, executives)
Workgroup?
Example Role

Policy implementation and support at county levels
County corrections department


Data matching, program oversight, policy advocacy, service
enhancement funding, facilitate jail in-reach
County department of social
services


Data matching, program oversight, policy advocacy, service
enhancement funding, facilitate shelter in-reach
Local or state behavioral health
agency (for frequent users of
mental health services)


Data matching, program oversight, policy advocacy, service
enhancement funding, facilitate hospital in-reach


Can commit city-specific resources such as vouchers, data
(e.g. police arrest data), and overall support

Provide slots in future or existing supportive housing sites,
perform outreach to potential tenants, service provision

Program design, assembled and coordinated funding,
program oversight and troubleshooting, TA/training,
State or local housing authority

Provide Section 8 or other housing vouchers
Foundation support

Provide funding for service enhancements and evaluation
City leadership partners – executive
leadership, police, housing
authorities
Supportive housing providers
CSH (where applicable)

Where are the frequent users of
county public systems?
• Street outreach
encounters
• Shelters
Homeless
System
Correction
System
• Jails
• Specialty courts
• Police arrests
•Hospital inpatient
•Emergency rooms
•Behavioral health
services
•Psychiatric hospitals
Health System
Substance
Abuse Services
• Detox centers
• Residential
rehabilitation
• Outpatient
treatment
Cross-system data matching



Where is the best data in terms of quality?
There may already be an agency with an
internal “frequent user” analysis, like a top
100 longest shelter stayers or other analysis
Where is the analytical capability?
◦ Staff who can receive data from other systems and
conduct match and analysis
◦ External researcher/organization (might cost $)

Consider data sharing constraints, legal and
privacy issues
Corrections data
HMIS/Shelter data
Behavioral
health
services
data
Hospital/
ER data
Restrictiveness
Data Sharing Flow
Possible Results of a Cross-System
Data Match
Frequent User Case Study
5-Apr-02
8-Apr-02
2-Aug-02
31-Dec-02
9-Mar-02
16-Oct-01
12-Nov-01
21-Dec-01
DOC
DHS
DOC
DHS
3-Feb-01
21-Feb-01
9-Mar-01
22-Mar-01
14-Jun-01
18-Jun-01
17-Jul-01
23-Jul-01
4-Aug-01
7-Sep-01
8-Mar-02
DHS
DHS
DHS
26-Jan-01
Neither System
DOC
DOC
DOC
15-Jan-01
DOC
DHS
DHS
DHS
1-Jan-01
DHS
DOC
DOC
DOC
DOC
DHS
Example of FUSE Match,
Hennepin County
There are a number of considerations when starting a cross system
matching project. These include:
1. Meeting the business needs of your primary partners and
sponsors
2. Meeting the requirements of the funding streams
3. Finding available data to match
4. Creating a FUSE list
5. Privacy and confidentiality
Clients are served in multiple
systems
Hennepin County Jail demographics from the National 2010 Arrestee
Drug Abuse Monitoring Program (ADAM) report.
http://www.whitehouse.gov/sites/default/files/ondcp/policy-and-research/adam2010.pdf
Serving the Needs of Partners
Our goal is to make a positive impact in clients lives as a result of
offering homeless services. Some of our partner’s agency goals that
we may also affect include:
• A reduction in street homelessness
• A reduction in arrests/police contacts or recidivism
• A reduction in emergency room visits
• Additional re-entry resources for persons exiting jail/prison or
for those under community supervision.
• A increase in community resources for clients with chemical
dependency or mental health issues
• Leveraging previously untapped funding sources
Meeting Funding Stream and
Partner Needs
Identify available resources, both cash and in kind, that are available for start-up and
support sustainability. In the case of Hennepin FUSE, these included:
•
Remaining funds from an MN Office of Justice Program re-entry grant serving
offenders with mental health issues awarded to Corrections. We needed to get
approval to serve a similar population when using the remaining funding.
•
A CHS supportive housing grant was awarded to St. Stephens Human Services. We
used the basic CSH New York FUSE model criteria, booked in Jail at least five times
in four years AND at least four admits to county shelters in three years.
•
Scattered site supportive housing funding through the MN Group Residential
Housing (GRH) Program helps sustain the program after startup. To be qualified
essentially clients had to be disabled and eligible for Minnesota General Assistance.
•
Given Corrections was a partner and providing a probation officer, new clients
selected were on probation for at least a year.
MN GHR Program
http://www.dhs.state.mn.us/main/idcplg?IdcService=GET_DYNAMIC_CONVERSION&RevisionSelectionMethod=LatestRelea
sed&dDocName=id_002549
Compulsory vs. Volunteer Systems
As we move forward with partnerships addressing housing and other systems, it
is important to understand basic differences between the Criminal Justice and
Human Service Systems.
•
In Human Services, a competent adult may volunteer for services. Noncompliance with the program may mean that they can no longer participate.
•
In Criminal Justice, offenders must meet court conditions for the term of the
sentence (probation or parole). These conditions may include no new crime,
no alcohol or drugs, or getting a place to live. When the supervision ends, the
involvement of probation ends. The court or corrections no longer has
jurisdiction over the offender.
The trend is to use the authority of the Criminal Justice system to encourage
offenders into engaging in services/programs that reduce their involvement in
the CJ system. See Drug Court, DWI Court, Community Court, Housing Court,
Mental Health Court, etc.
Finding available data to match
Partners will often release their datasets or assign analysts for projects such as FUSE.
When possible, you should start with data that is publically available and focus on the
elements that are already posted online or considered public data. Examples include:
•
Jail Data - name, dob, charges, booking and release dates, arresting agency, charging
agency, city and state
•
Court data – name, dob, convictions, city, state and zip code
•
Police arrest/contact data – not posted online but often public available on request
•
Shelter data – It may be available through HMIS or through your county for countyfunded shelter.
•
Other human services or medical data – these may be available in summary format if
you provide the organization with a match file for program analysis.
It’s best to partner with agencies that provide data, then update them as the program
progresses.
Privacy and Confidentiality
The goal should be share a little information
as possible when identifying clients to
minimize risk of releasing private data.
I
•
•
•
•
recommend not matching on the Social Security Number
The SSN is private information
It is not intended to be a universal identifier
It is often self reported by the client, so may it may be wrong
There is danger of fraud or crime if numbers get inadvertently
released.
We also should be aware of HIPPA (Health Insurance Portability and
Accountability Act) Privacy Rule, especially when partnering with
health care providers.
HIPPA Privacy Rule http://www.hhs.gov/ocr/privacy/hipaa/administrative/privacyrule/index.html
Creating a Frequent User List
1. Start with organizing the most public data file first. In Hennepin, that meant
sorting the jail file to look for offenders that had more than 4 times in 5
years. Five years of jail data represent approximately 250,000 admits to the
jail.
• First, eliminate any duplicate booking records or information that is not
public or necessary for the match (race, home address, etc. - consult
with your attorney)
•
Then, concatenate the first name, last name and dob. The result should
look something like “BRADKAETER40938”
•
Aggregate or count the number of bookings per each concatenation
(individual) over the period you select. Sort out any that don’t meet the
criteria, ours was 4 bookings in 5 years)
•
In Hennepin the file was matched to who is on probation for at least a
year. Those that were not were eliminated. This was to serve Corrections
need for finding housing to help reduce recidivism.
•
Pass off this more “public” file off to the Shelter/Human Services
Analysts for the match on the more sensitive shelter data.
Creating a Frequent User List
2.
Follow a similar process for available shelter data.
•
Eliminate duplicate admits (same person, same date)
•
Then, concatenate the first name, last name and dob
(“BRADKAETER40938”)
•
Aggregate or count the number of admits per each concatenation over
the period you select. Sort out any that don’t meet the criteria (4 uses
of emergency shelter in years to be defined as chronically homeless).
Hennepin County has about 19,000 shelter units per month (one
night/one person)
•
Match the file on the concatenation. In Hennepin the list was
approximately 200 persons long after the jail/shelter/probation match
•
Share the short list of eligible names with the FUSE team, Shelter Staff
and corrections re-entry staff to offer services as people flow through
key points in the system.
Key Learnings from Hennepin
FUSE
To make a FUSE program work:
• Meet the Needs of the Clients
• Meet the Needs of your Partners
• Meet the Needs of your Funders

similar documents