Representativeness of Emergency Department Data Reported to the BioSense System

Representativeness of Emergency Department Data Reported to the BioSense System Background The CDC BioSense System receives data from 461 non-federal hospitals based on their capability and willingness to supply electronic data. The National Hospital Ambulatory Medical Care Survey (NHAMCS) is an annual national probability sample survey of hospitals that collects data on U.S. emergency department (ED) visits. We assessed the representativeness of BioSense ED data by comparing it with the NHAMCS results. Methods For 2004, NHAMCS estimated a total of 110.2 million ED visits nationally and a yearly increase of 1.5 million visits; therefore we estimated 114.7 million ED visits in 2007. We compared 2004 NHAMCS data to 2007 BioSense data regarding total visits, patient ages, selected diagnosis categories, and geographic area. Results In 2007, BioSense received 13.4 million ED visits, 11.7% of the estimated U.S. total. By agegroup, percent of total visits for NHAMCS vs BioSense respectively was: <15 years, 20.8% vs 18.2%; 15-44 years 45.4% vs 43.5%; 45-64 years 19.5% vs 20.4%; and >64 years 14.3% vs 14.7%. Among selected diagnosis groups, percent of total visits for NHAMCS vs BioSense were: infections (ICD-9 codes 001-139) 3.1% vs 2.6%; respiratory (codes 460-519) 11.2% vs 9.3%; digestive (codes 520-579) 6.5% vs 6.0%; injury and poisoning (codes 800-999) 28.6% vs 14.7%; and symptoms, signs, and ill-defined conditions (codes 780-999) 20.4% vs 20.4%. By geographic area, percent of total visits for NHAMCS vs BioSense were: Northeast, 20.2% vs 1.3%; Midwest, 24.3% vs 64.7%; South 37.3% vs 26.8%; and West 18.1% vs 7.3%. Discussion Using PHIN standards, BioSense receives nearly 12% of total U.S. ED visits and automatically analyzes and displays them within hours of receipt. BioSense data are representative of age and most diagnosis groups but not of geographic area. Initiatives to increase recruitment and improve representativeness are underway.