Identifying and mapping preventable admissions and costs: A collaboration between DoD and AHRQ

Background: The project is a collaboration of the TRICARE Management agency (TMA) and the Agency for Healthcare Quality Research (AHRQ). The TMA mined several sources of Military Health System (MHS) administrative data for hospital admissions and beneficiary information. The 2007 national continuously enrolled population is more than 6 million in the HMO portion of the MHS. The AHRQ is applying software tools analyzing and mapping geographic variation in rates of preventable hospital admissions (e.g., Adult asthma, diabetes complications, urinary tract infections...) and cost. Objectives: This project will determine unexpectedly high or low rates and cost, by county, for active and retired military members and families enrolled in the TRICARE in 5 states: CA, CO, FL, TX, and VA. Methods: AHRQ Prevention Quality Indicator software documentation was first downloaded from the AHRQ website to identify the files and variables required from the source data. Nine files from the MHS Data Repository (MDR) with 22,498,699 records containing encounter and claims records of admissions to military and civilian hospitals, as well as population data were analyzed and processed to produce the four files required for use with the AHRQ software. Findings: Four files were produced for this project; a file of 106,742 military hospital admissions; a file of 130,520 civilian hospital admissions; a file of military hospital identifiers; and a population file of continuous MHS enrollees in 2007. AHRQ has begun processing the information to produce the rates of preventable admissions and cost mapping information for reports on each state. Conclusions: The work is in progress, and results will be available for the annual meeting. In the interim, the project demonstrated the feasibility of tracking quality of care and cost differences across the MHS or other enrolled health plans by using data mining of administrative medical records, standardized measures, and a multidisciplinary team.