RELATIVE EFFECTIVENESS OF DIFFERENT APPROACHES TO SURVEILLANCE OF INFECTIONS AMONG HOSPITALIZED PATIENTS

A method for estimating the effectiveness of systems designed to monitor the occurrence of events within a population is described. Specific application of this model lo the analysis of a control program for the surveillance of infectious disease in a university hospital is presented. A reporting system with three sources—physicians, nurses, and bacteriology laboratory—was instituted. The effectiveness of the system ranged from 61 to 85 per cent over a period of 12 months. There was considerable variation among the three sources in terms of their relative effectiveness in identification of patients with infections. Under the circumstances described, the most effective two-source reporting system would have been the physicians* discharge reports and bacteriology laboratory records. © Lippincott-Raven Publishers.