204 Community-Level Hospital Statistics from the Healthcare Cost and Utilization Project (HCUP): An Exploration of Septicemia - CANCELLED

Wednesday, June 25, 2014: 10:00 AM-10:30 AM
East Exhibit Hall, Nashville Convention Center
Halcyon G Skinner , Truven Health Analytics, Bethesda, MD
Rachel E Henke , Truven Health Analytics, Bethesda, MD
Gary Pickens , Truven Health Analytics, Bethesda, MD
Lorena Lopez-Gonzalez , Truven Health Analytics, Bethesda, MD
Anne Elixhauser , Agency for Healthcare Research and Quality, Rockville, MD
Roxanne Andrews , Agency for Healthcare Research and Quality, Rockville, MD
Laurel Holmquist , Truven Health Analytics, Bethesda, MD

BACKGROUND: To monitor the health of populations, it is important to consider health information across multiple domains, including risk factors, incidence and mortality, health outcomes, and health care. The Agency for Health Care Research and Quality’s (AHRQ) Healthcare Cost and Utilization Project (HCUP) is a collection of databases and software tools developed through a Federal-State-Industry partnership to create National and State level information about health care. HCUPnet (http://hcupnet.ahrq.gov/), an online data query tool, now features community-level statistics. This tool provides comprehensive information about hospital discharges and inpatient costs at the county level for participating states (23 states as of January 2014). All statistics can be further stratified by age and sex. To explore the utility of HCUP community-level data for epidemiology, we examined discharge rates and length of stay for septicemia across counties in one of the participating states, Florida.

METHODS:  Hospital discharge data are derived from State Inpatient Databases, which include information representing 97% of all US discharges. Discharge data are harmonized to permit easy comparison across States. County of residence for discharges is assigned by ZIP code with reference to the Nielsen ZIP code database, as are county population size estimates. Data from cells with fewer than 11 discharges are suppressed. Results from counties with incomplete data due to non-reporting hospitals are also suppressed. For each county in Florida, we used the HCUPnet Community-level Statistics path, to obtain the rates of hospital discharges for a principal diagnosis of septicemia and the mean length of stay for 2011 septicemia hospitalizations. 

RESULTS: The rate of septicemia discharges ranged from 159.9 per 100,000 (Collier county) to 712.9 per 100,000 (Highlands county), compared to 297.3 per 100,000 for Florida as a whole and 348.6 for the U.S. The mean length of stay ranged from 6.1 days (Taylor county) to 12.9 days (Madison county), compared to 9 days and 8 days for Florida and the U.S. respectively.

CONCLUSIONS: Examining hospital discharges for septicemia across Florida counties, we observed greater than 4 fold differences in hospital discharge rates and two fold differences in length of stay. The Community-level statistics available through HCUPnet provide a unique source of county-level data about hospital utilization and inpatient costs derived from the most complete all-payer inpatient datasets in the United States. These data are a valuable resource for epidemiologists to measure and monitor the health of populations. Future enhancements will include age and sex adjustments and a wider range of clinical conditions and procedures.