Pseudoepidemics in hospital.

20 (11%) of 181 nosocomial epidemics investigated by the Center for Disease Control during 1956-75 were actually false outbreaks. In each pseudoepidemic, on-site investigation showed that errors in specimen processing (11), surveillance artefacts (6), or clinical misdiagnosis (3) had resulted in real clusters of false infections or artefactual clusters of real infections. Specimen-processing errors included contamination during collection or transport, cross-contamination in the laboratory, and incorrect identification of organisms. Prevention of pseudoepidemics depends on quality control aimed at avoiding such errors and on close cooperation between laboratory workers and clinicians, to ensure systematic review of incongruous laboratory results.