[영문]objective. The purpose of this study was to provide a national estimate of thenumber of healthcare-associated infections (HAI) and deaths in United Stateshospitals.Methods. No single source of nationally representative data on HAIs is currentlyavailable. The authors used a multi-step approach and three datasources. The main source of data was the National Nosocomial InfectionsSurveillance (NNIS) system, data from 1990?2002, conducted by the Centersfor Disease Control and Prevention. Data from the National Hospital DischargeSurvey (for 2002) and the American Hospital Association Survey (for 2000) wereused to supplement NNIS data. The percentage of patients with an HAI whosedeath was determined to be caused or associated with the HAI from NNISdata was used to estimate the number of deaths.Results. In 2002, the estimated number of HAIs in U.S. hospitals, adjusted toinclude federal facilities, was approximately 1.7 million: 33,269 HAIs amongnewborns in high-risk nurseries, 19,059 among newborns in well-baby nurseries,417,946 among adults and children in ICUs, and 1,266,851 among adultsand children outside of ICUs. The estimated deaths associated with HAIs inU.S. hospitals were 98,987: of these, 35,967 were for pneumonia, 30,665 forbloodstream infections, 13,088 for urinary tract infections, 8,205 for surgicalsite infections, and 11,062 for infections of other sites.Conclusion. HAIs in hospitals are a significant cause of morbidity and mortalityin the United States. The method described for estimating the number of HAIsmakes the best use of existing data at the national level.