1 / 39

Thomas L. Moore, Ph.D., CEO, Herbert & Louis LLC

Matching HMIS and Mainstream Databases: Possibilities and Challenges of Using Multiple Datasets for Research. Thomas L. Moore, Ph.D., CEO, Herbert & Louis LLC Stephen Metraux, Ph.D., Department of Health Policy & Public Health,University of the Sciences in Philadelphia

mhoss
Download Presentation

Thomas L. Moore, Ph.D., CEO, Herbert & Louis LLC

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Matching HMIS and Mainstream Databases: Possibilities and Challenges of Using Multiple Datasets for Research Thomas L. Moore, Ph.D., CEO, Herbert & Louis LLC Stephen Metraux, Ph.D., Department of Health Policy & Public Health,University of the Sciences in Philadelphia Diana M Tester, SC Budget and Control Board, Office of Research and Statistics

  2. Overview of Session • Learning Objectives • Three “Ascending” Cases: • Inaccessible Databases: “A Workaround When Databases Were Inaccessible” • Tom Moore • Ad Hoc Data Merging: “Putting Together a Cost Study – Experiences from the PADOC - Gaudenzia FIR-St. Study” • Steve Metraux • A Permanent Research Database: “Linking HMIS Data to Mainstream Service Administrative Files: The South Carolina Experience” • Diana Tester September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  3. Learning Objectives • To understand the range of potential data sources that might be linked to HMIS data and what can be learned from them. • To learn about the administrative procedures, agreements, and technical challenges involved in matching data from diverse data sources. • To share findings from recent efforts in multiple communities that linked HMIS data with other datasets. September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  4. A Workaround When Databases Were Inaccessible Thomas L. Moore, PhD CEO Herbert & Louis LLC Wilsonville, OR 97070 tlmoore@herblou.com

  5. Background • Agency: Central City Concern, Portland Oregon • Three Assertive Community Treatment Teams (modified) • Serving chronically homeless, dually diagnosed • Study opportunity sample size n = 34 September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  6. The Problem • Need to document cost/benefit in face of funding cuts • Myriad of un-linked, multiple owner databases • Critical timeframe for implementation • Very limited study funding September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  7. Initial (Desired) Approach • Attempt to identify existing relevant data sources • Map potential access points • Multiple private hospitals/ERs & charitable agencies • Multiple public Mental Health/Addictions Agencies with various funding streams • Myriad of access points to housing and support services • City/County/State law enforcement • Attempt to gain access to data • Nearly everyone supported the idea/need • Complex agency specific approval processes • Difficult to compete with other priorities September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  8. Fall Back Approach • Identify and access other data sources for evidence-based averaged costs of generalized services • Rely on self-report for historical utilization through semi-structured multiple-session interviews • Conduct detailed clinical records review to identify first year post-enrollment utilization • Compare pre and post utilization using averaged service costs September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  9. Cost Domains • Inpatient Physical Health Hospitalization • Emergency Room Visits • Outpatient Physical Health Visits • Alcohol/Drug Inpatient Nights • Alcohol/Drug Outpatient Visits • Mental Health Inpatient Nights • Mental Health Outpatient Visits • Incarceration Days September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  10. Findings (Estimates) • Ave estimated pre-enrollment annual cost: $42,075 • Ave estimated post-enrollment annual cost: $17,199 • Ave treatment + housing investment: $9,870 • Post-enroll average savings: 35.7% • Savings for three teams (n=293): $4,396,758 • Estimated cost if no treatment: $12,327,975 September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  11. Limitations and Weaknesses • Reliability of self-report • Non-randomized sample - underestimated utilization • Utilization of “average” per encounter costs • Does not include benefits of treatment – e.g. employment, etc. September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  12. Putting Together a Cost Study –Experiences from the PADOC - Gaudenzia FIR-St. Study Stephen Metraux, Ph.D. Department of Health Policy & Public Health University of the Sciences in Philadelphia

  13. PADOC-Gaudenzia FIR-St. Study Primary Research Questions Among persons released from PADOC prisons to Philadelphia County locations, is mental illness a factor associated with an increased likelihood of: • “maxing out” and serving longer prison episodes (prior to release)? • rearrest and reincarceration among released prisoners? • increased frequency of homeless shelter use? To what extent do persons treated for MI while incarcerated receive public MH services after release? September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  14. PADOC-Gaudenzia FIR-St. Study Primary Research Questions con’t Is participation in the Gaudenzia FIR-St. program associated with: • serving less prison time? • getting arrested and reincarcerated (jail and prison) less? • staying in homeless shelters less? and • differing patterns of behavioral health services? Do the differences in services utilization translate into cost savings? September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  15. Data Sources • PA Department of Corrections • Gaudenzia FIR-St. • Philadelphia Prison System • Philadelphia Community Behavioral Health • Philadelphia Police Department • Philadelphia Office of Emergency Shelter Services (HMIS) • PA Department of Probation & Parole September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  16. Putting Together a Cost Study – 7 Steps • Human and Financial Resources • Data Availability • Data Access • Institutional Review Board • Database Integration • Data Analysis • Dissemination of Findings September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  17. Step 1 – Determine Human & Financial Resources • Source & purpose of study • Funding – amount and sources • Scope of work • Costs • Personnel • Data acquisition • Consulting September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  18. Step 2 – Data Availability • Identify target population • Is there a control population? • What types of services are used? • What types of data are available that document this services use? September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  19. Step 3 – Data Access • Political considerations • Confidentiality considerations • Anxiety about control over data • Legal considerations • Data quality issues • Cost issues • Bureaucracy September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  20. Step 4 – Institutional Review Board • Ensures that the interests of human subjects are protected • Particularly interested in safeguards for maintaining confidentiality and anonymity • Often will expedite review because no contact with research subjects is involved • HIPAA compliance • Special populations are often involved September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  21. Step 5 – Database Integration & Matching • Matching databases • Data management issues • Confidentiality and security September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  22. Step 6 – Data Analysis • Know the capabilities of your data • Have expertise available • Methodological rigor vs. “user friendliness” September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  23. Step 7 – Dissemination of Findings • Advocacy • Creating a body of research • Integrating findings with other related messages • Preliminary Findings from Gaudenzia FIR-St. Study September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  24. Parting Advice • Be Flexible • Be Creative • Be Patient • Expect the Unexpected • Ask (and give) help • Know your audience September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  25. Linking HMIS Data to Mainstream Service Administrative files: The South Carolina Experience Diana M Tester SC Budget and Control Board Office of Research and Statistics

  26. Who Are We? • SC Budget and Control Board is a Service Agency in South Carolina • Neutral, no programmatic responsibilities • Statistical and Research-focused September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  27. Integrated Data System LEGEND Elder Services & Assessments Disabilities & Special Needs Vocational Rehabilitation Law Enforcement Legal/Safety Services Juvenile Justice Disease Registries Public Safety Social Services Education Probation, Pardon & Parole Claims Systems Health Department Corrections* All Payer Health Care Databases Environmental Conditions Child Care Behavioral Health Social Services Alcohol & Drug Services Medicare Health Department Mental Health Medicaid Services Education Health Professions State Employee Health Services Other State Support Agencies Hospitalizations Community Health Centers Disease Registries Free Clinic Visits Emergency Room Visits Outpatient Surgeries Home Health Care *Still in contract negotiations September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  28. SC Data Warehouse • Build off of existing systems (legacy systems from state agencies and private sector) • Create a Unique ID for each person (not related to any other number) • Identifiers are pulled off of the statistical data. Use only the statistical data • Data is always “owned” by the originating agency. Must have permissions to use and/or link any data September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  29. Rules for Building a Data Warehouse • The integrated system should include data from Public, Private and Not-for-Profits with all being treated equal. (No Politics!) • Do not disturb the balance of power – run from anyone who wants your data in their computers for them to control September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  30. Rules (cont) • The issue is access to data for research – fair and equal access – no organization should have more access than another – public, private and not-for-profit • The agency that houses the data should preferably not be a programmatic agency – this could bias analysis, access, etc… September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  31. Agencies Using the Data Warehouse: • Evaluate programs • Look at outcomes • Understand better how their programs interact with other agency programs • Study Health, Human Service, Education, and Law Enforcement Issues • Statistical – Aggregate Information September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  32. How Does This Relate to Homelessness? • South Carolina Council on Homelessness represents key leadership from 20 agencies that provide mainstream and other services and funds to homeless individuals, programs or organizations. • Development of a common vision – including for data • In Nov. 2004, the Council released its Blueprint to End Homelessness in South Carolina. • Goal 5: “Develop a comprehensive, ongoing statewide homeless data collection and analysis system” September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  33. How Can We Make This Vision Happen? • We needed a volunteer! • Upstate Coalition steps up to the plate • A Win-Win … The Coalition needed data • Drafting an MOU • Extracting a file September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  34. Fast Facts • 3310 (2,657 adults) homeless persons in our cohort, in 2003: • 1223 (37%) received Food Stamps • 238 (7%) received TANF • 1014 (30%) received Medicaid – including behavioral health services. • 414 (13%) had an inpatient hospitalization • 1565 (47%) had an ER visit • 1765 (66%) of adults had an arrest • 984 (30%) had been under supervision with Probation, Parole, and Pardon services September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  35. FINDINGS • In 2003 alone, homeless persons linking to Medicaid and Social Services consumed over $6 million in services • Further analysis could help determine if the lives of these people have improved because of these expenditures September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  36. Where are we now? • South Carolina Homeless Coalition along with the South Carolina Council on Homelessness support the development of this database and have been active in pursuing it. • Development of a new MOA template • Another Coalition has signed a MOA and developed an extract. • Another Coalition is about to sign on (There are only 10 Coalitions in SC) September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  37. Barriers • Lack of Resources • It takes time & resources to develop relationships and MOAs • It takes time & resources to develop extracts and link appropriate populations to relevant administrative files • It takes time & resources to analyze and understand what the data tells us about homelessness in SC September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  38. Parting Advice (I stole this slide!) • Develop a Vision • Be Persistent (even if no resources!) • Be Patient • Be Creative • Continue the Vision September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

  39. For Further Information Walter “Pete” Bailey 803-898-9949 pbailey@drss.state.sc.us Diana M Tester 803-898-9962 dtester@drss.state.sc.us SC Budget & Control Board, Office of Research and Statistics September 18-19, 2006 - Denver, Colorado Sponsored by the U.S. Department of Housing and Urban Development

More Related