The costs associated with nosocomial bloodstream infections in the pediatric intensive care unit

Abstract

OBJECTIVE: To assess the operational and subsidiary costs and length of stay (LOS) attributable to nosocomial bloodstream infections (BSI) in a pediatric intensive care unit (PICU). DESIGN: Matched case-control study. SETTING: Sixteen bed PICU in a 250-bed tertiary-care pediatric hospital. PATIENTS: Cases with BSI were prospectively identified from PICU patients who developed a nosocomial BSI from August 1996 to July 1998. Controls were PICU patients who were matched for age, severity of illness, diagnosis, and admission date who did not develop a nosocomial BSI. RESULTS: A total of 38 cases and 38 controls form the basis for this study. The cases and controls were similar with respect to the matching criteria. In addition, the cases and controls did not differ with respect to demographic characteristics or PICU survival. There was a trend toward increased hospital mortality among cases (23.7% vs. 10.5%, p =.084). Significant differences were encountered in the utilization of PICU therapeutic modalities. Cases were significantly less likely to be managed care plan enrollees (36.8% vs. 60.5%, p =.043). Total operational and subsidiary costs for radiology, pharmacy, and laboratory services were significantly higher for cases than controls ($78,272 vs. $35,005, $3,622 vs. $1,432, $8,635 vs. $4,630, and $8,648 vs. $3,971, respectively; all p <.001). The PICU and hospital LOS were significantly higher for cases than controls (19.3 vs. 4.6 days for PICU and 46.7 vs. 24.4 days for hospital; both p <.001). The operational costs attributable to nosocomial infection were $46,133. Radiology, pharmacy, and laboratory costs attributable to nosocomial infection were $2,364, $4,691, and $5,156, respectively. The PICU and hospital LOS attributable to a nosocomial BSI were 14.6 days and 21.1 days, respectively. The attributable mortality rate was 13.1%. CONCLUSIONS: The costs and LOS associated with nosocomial BSI in patients admitted to the PICU were significantly higher than controls.