Measuring Outcomes Over Time

1
Measuring Outcomes Over Time

• Data Types
• Data Sources
• Data Sharing Agreements
• Data Matching
• Lessons Learned

2
Data Types

• Survey Data
• Self-report
• Non-restrictive
• Resource Intensive
• Low Response Rate
• Point in Time
• IRB Approval

3
Data Types

• Administrative Data
• Ongoing Sustainable
• Empirically-Based Objective
• Less Resource Intensive
• Longitudinal
• Incremental for Populations Time Frames
• Unobtrusive
• Limited by Availability and Collaboration

4
Data Sources

• DMHSAS
• State-funded SA Mental Health
• Department of Public Safety
• DUI Convictions
• Public Intoxication
• Department of Corrections
• Incarcerations, Probation Parole
• Employment Security Commission
• Wage Unemployment Benefits

5
Data Sources

• Oklahoma State Department of Health
• Mortality, Hospital Discharge
• Oklahoma Health Care Authority
• Medicaid Claims
• Oklahoma State Bureau of Investigation
• Felony Arrests/Dispositions
• Department of Human Services
• TANF
• Food Stamps
• Child Welfare

6
Data Sharing Agreements

• Who is matching the data
• Confidentiality Concerns
• 42 CFR
• HIPAA
• Security
• Business Associate
• Qualified Service Organization

7
Data Matching

• Receiving the Data
• Cleaning and Preparing the Data
• Matching Methods
• All-Or-Nothing
• Probabilistic
• Combination of the Two
• Overlap Procedure (PPE)

8
Data Matching

• ODMHSAS Methodology
• Block by Sex
• Join by DOB Components
• Assign Weights to All Identifying Variables
• Determine Threshold of Likelihood
• Unduplicate

9
Lessons Learned

• Allow Much More Time Than You Think At Each Step
• Be Flexible Willing to Start Over Or Live With
  It
• Meet With Other MIS Staff
• Know What Is Needed From The Data Beforehand

10
CD

• Matching Algorithm
• SQL code
• Data Linking Reference List
• Presentations