Data Sharing and FUSE: The National Landscape

Presentation By: Kim Keaton, Director of Data and Analytics/CSH
Prepared for Hamilton County 3/18/2019

There are a lot of reasons to share data across housing and health, jail and other sectors. These are some typical ones seen that are supportive of FUSE efforts.

1. It may be that a data sharing exercise can help both systems understand the complexity and medical and social needs of the target population, which can be helpful from a policy and champion-engagement standpoint.

2. Identification of the costliest subset of high utilizers for housing (this is the typical FUSE effort)
3. Shared data can also support improved coordination between justice, health and housing systems, and improved coordination can help improve outcomes for clients.
4. Shared data across the whole population of systems can help make the business case for a supportive housing intervention. FUSE evaluations have shown often the costs avoided by reduced systems utilization can offset the cost of housing and services.

Transcript

1. Data Sharing and FUSE: The National Landscape Kim Keaton, Director of Data and Analytics Prepared for Hamilton County 3/18/2019
2. Frequent Users Systems Engagement (FUSE): More than 30 communities have implemented supportive housing for frequent users using the FUSE planning approach, combining data driven targeting, multi-stakeholder engagement, and targeted supportive housing csh.org/fuse
3. © All rights reserved. No utilization or reproduction of this material is allowed without the written permission of CSH. FUSE Evaluations Show Success: Examples • 40% reduction jail days • 91% fewer shelter days • 50% reduction in psych. inpt. • 86% housed after 2 years NYC FUSE • 67% reduction in total public costs after 2 years • 60-80% reduction in ambulance, ER, hospitalizations, arrests, jail days • Net savings of over 200% after paying for housing+services San Diego Project 25 • 50% fewer arrests • 87% fewer shelter days • 24% less ambulance service charges • 43% less hospital charges MeckFUSE (Charlotte)
4. © All rights reserved. No utilization or reproduction of this material is allowed without the written permission of CSH. FUSE Action Planning Steps Planning Phase •Step 1: Identify a Champion and Project Manager •Step 2: Assemble a Cross Systems Planning Team •Step 3: Execute a Cross Systems Data Match •Step 4: Create SH Pipeline •Step 5: Secure Evaluation Resources Pilot Implementation Phase •Step 6: Create Referral Process •Step 7: Inreach/Outreach, Lease up •Step 8: Implementation Monitoring and Support Scaling & Replication Phase: •Step 9: Determine Scaling Needs •Step 10: Identify Financing for Scaled FUSE
5. 5 Benefits of Using Data for FUSE 1. To understand the complexities of the FUSE target population, both medically and socially, and can help address policy concerns such as rising health care costs 2. To identify the costliest and most vulnerable subset of high utilizers to prioritize them for housing 3. To improve coordination between justice, health and homeless and housing systems, which can in turn improve outcomes for patients/clients. 4. To make the business case for a supportive housing intervention, and with the right data on utilization costs and costs of housing and services, can often show a potential ROI, or return on investment.
6. © All rights reserved. No utilization or reproduction of this material is allowed without the written permission of CSH. FUSE is an individual and systems approach Jan-Dec Year 2 Jan-Dec Year 1 6 2 4 13 1 4 1 5 1 6 17 18 19 20 21 2 2 2 3 1 2 3 4 5 7 8 9 1 0 1 1 1 2 Frequent User Case Study Shelter Jail Unknown Frequent Use Across System Populations
7. © All rights reserved. No utilization or reproduction of this material is allowed without the written permission of CSH. Why do all this? What’s the added value? Public Systems – Uncoordinated Response Maximized Resources, Systems Change
8. © All rights reserved. No utilization or reproduction of this material is allowed without the written permission of CSH. Examples of FUSE communities changing systems with data Community Boise, ID Washtenaw Co., MI Miami, FL Data Matching: Instituting a scalable data matching process to support data management in HMIS for prioritization and evaluation of 40 unit PSH for frequent utilizers Executed BAA with U of MI hospital system and HMIS to identify highest users of hospitals in county Matched frequent user lists from hospital, jail, HMIS, police outreach to identify top 300 frequent users. Instituted care coordination teams to find people on the list. Systems Change: Opened New Path supportive housing residence for high utilizers of hospitals and jail, which is co-funded by hospitals and Ada County After demonstrating promising early evaluation results, hospital dedicated funding for services for PSH program. Evaluation found significant cost reductions for high utilizers Created community wide prioritization for top users; built out HMIS because found that highest users were not hitting homeless system. Placed into Coalition LIFT, funded by state housing finance
9. Aren’t we concerned about privacy? What about HIPAA?? 42 CFR????
10. © All rights reserved. No utilization or reproduction of this material is allowed without the written permission of CSH. Data sharing and HIPAA: A brighter future?  HIPAA already allows for sharing for the purposes of care coordination  In September 2016, HHS issued an FAQ on how covered entities could collect criminal justice data that included care coordination for the specific purpose of identifying a patient’s need for supportive housing:  “For example, a provider may disclose PHI about a patient needing health care supportive housing to a service agency that arranges such services for individuals.”  https://www.hhs.gov/hipaa/for-professionals/faq/2073/may-covered-entity- collect-use-disclose-criminal-data-under-hipaa.html  In December 2018, HHS’s Office of Civil Rights issued a Request for Information on Modifying HIPAA Rules to Improve Coordinated Care  Explicitly sought input around homeless populations which need care coordination  Specifically, they were seeking input on whether and how to encourage HIPAA covered entities to share PHI with non-covered entities when needed to coordinate care and provide related health care services  Indicates that HHS OCR may be close to issuing new and supportive guidance on data sharing  https://www.federalregister.gov/documents/2018/12/14/2018-27162/request- for-information-on-modifying-hipaa-rules-to-improve-coordinated-care#p-35
11. © All rights reserved. No utilization or reproduction of this material is allowed without the written permission of CSH.