The epidemiology and attributable outcomes of candidemia in adults and children hospitalized in the United States: a propensity analysis. Clin Infect Dis 41

(See the editorial commentary by Fridkin on pages 1240–1) Background. Candida species are the fourth most common cause of bloodstream infection and are the leading cause of invasive fungal infection among hospitalized patients in the United States. However, the frequency and outcomes attributable to the infection are uncertain. This retrospective study set out to estimate the incidence of candidemia in hospitalized adults and children in the United States and to determine attributable mortality, length of hospital stay, and hospital charges related to candidemia. Methods. We used the Nationwide Inpatient Sample 2000 for adult patients and the Kids ’ Inpatient Database 2000 for pediatric patients. We matched candidemia-exposed and candidemia-unexposed patients by the propensity scores for the probability of candidemia exposure, which were derived from patient characteristics. Attributable outcomes were calculated as the differences in estimates of outcomes between propensity score–matched patients with and without candidemia. Results. In the United States in 2000, candidemia was diagnosed in an estimated 1118 hospital admissions of pediatric patients and 8949 hospital admissions of adult patients, yielding a frequency of 43 cases per 100,000 pediatric admissions (95 % confidence interval [CI], 35–52 cases per 100,000 pediatric admissions) and 30 cases per 100,000 adult admissions (95 % CI, 26–34 cases per 100,000 adult admissions). In pediatric patients, candidemi