Nosocomial Bacteremia in Perspective: A Community-Wide Study

Abstract

Prospective surveillance of 1,527 episodes of bacteremia in the four major hospitals serving a community of 300,000 revealed striking differences among the hospitals. The rate of community-acquired bacteremia ranged from 1.9 to 6.8 per 1000 discharges; the rate of nosocomial bacteremia ranged from 0.9 to 9.8 per 1000 discharges. Death rates attributed to nosocomial bacteremias varied among the hospitals by as much as 40-fold. Nosocomial bacteremias were less frequent at two community non-teaching hospitals than at two teaching hospitals. Although fewer blood cultures were obtained at the nonteaching hospitals, the ratio of blood cultures obtained to deaths attributed to bacteremic infection was highest at these hospitals, suggesting that physicians recognized the importance of obtaining blood cultures in critically ill patients. The potential usefulness of bacteremia surveillance is reviewed. It is suggested that hospitals serving a single community might constitute an appropriate unit for such surveillance [Infect Control 1981: 2(3):219-226.]