Back to Annual Meeting

Back to Annual Meeting

APHA Scientific Session and Event Listing

Jim E. Banta, PhD, MPH, Seth Wiafe, MPH, Sam Soret, PhD, and David Dyjack, DrPH. School of Public Health, Loma Linda University, 24951 North Circle Drive, Loma Linda, CA 92350, (909) 558-7753, jbanta@llu.edu

Much effort has been spent preparing for a possible epidemic of Avian flu. Although it is impossible to predict if, when, and where such an epidemic might occur, hospital discharge data may be used to identify areas most likely to be negatively impacted. One can determine which areas have historically experienced higher proportions of hospital discharges with upper respiratory infections.

Publicly-available California discharge data for all community-based acute care hospitals was obtained for 1999–2003. SAS 8.0 software was used to scan the twenty-five diagnosis fields available for each discharge. Definitions from the Agency for Healthcare Research and Quality's Clinical Classification Software for ICD-9-CM were used to identify discharge records with diagnoses of influenza (5,319 cases in 2003 alone), pneumonia (234,136 cases in 2003), and acute bronchitis (34,374 cases in 2003). ArcView 9.1 was used to geocode the inpatient data to 2,648 valid zipcodes in California (3.6 to 3.8 million discharges per year). SaTScan 6.0 was used to calculate relative risk rates for each zip code and year by dividing the number of discharges for each type of respiratory infection by the total number of discharges. Raw rates were computed, as well as rates adjusted by age, race, ethnicity, and gender. Krieging methods in ArcView were applied to the relative risks to create smoothed maps.

Resulting maps clearly identify areas that may be more susceptible to negative health outcomes in the event of a flu epidemic. Such information may assist local and state-level planning for response to an outbreak.

Learning Objectives:

• Articulate advantages of inpatient discharge data for population risk assessment.
• List procedures for creating relative risk maps.
• Discuss benefits of smoothed versus raw relative risk maps for planning purposes.

Keywords: Geographic Information Systems, Risk Assessment

Presenting author's disclosure statement:

Any relevant financial relationships? No