Estimating the extra charges and prolongation of hospitalization due to nosocomial infections: a comparison of methods.

Estimates of the economic consequences of nosocomial infections were derived by studying the same patient population with two different methods: physician's assessment and comparisons of patients with nosocomial infection and those without, matched on five characteristics. Estimates of extra days and extra routine charges obtained by the comparison were about 2 1/2 times greater than those obtained by the physician's assessment (P less than 0.0001). Even when the match of patients was exact and measures were taken to avoid confounding, patients with nosocomial infection had more discharge diagnoses recorded (P = 0.02) and experienced more episodes of pulmonary embolism, renal failure, and death in the hospital than did their counterparts. These differences suggest that, despite careful matching, the members of the pairs were not comparable in their intrinsic predisposition to prolonged hospitalization. Unless patients can be matched on this prediposition, the estimates from a comparison study will be exaggerated.