Recurrent epidemics caused by a single strain of erythromycin-resistant Staphylococcus aureus. The importance of molecular epidemiology.

OBJECTIVE--To compare the accuracy of clinical epidemiologic findings with results of molecular epidemiologic analysis in identifying the source of recurrent epidemics of erythromycin-resistant Staphylococcus aureus (ERSA) infections in a well-baby nursery. DESIGN--Epidemic investigations with retrospective and prospective microbiological surveillance. Staphylococcus aureus isolates were evaluated by plasmid analysis and genotyping. SETTING--A well-baby nursery in a 700-bed university teaching hospital with approximately 250 deliveries per month. PATIENTS--Newborn infants who developed ERSA infections during 1990 and 1991. INTERVENTION--Traditional infection control measures, including cohorting of infected infants and hand washing, were implemented. Personnel were cultured for nasal carriage of S aureus during both epidemics. Employees carrying ERSA were removed from the nursery and treated with mupirocin. RESULTS--In the first epidemic, 15 infants were infected with ERSA. A nursing assistant who cared for most of the infants was found to be a carrier of ERSA. She was removed from the nursery and the epidemic resolved. Fifteen months after the first epidemic ended, an epidemic that involved 11 infants began. The attending physician had a facial furuncle and was found to be a carrier of ERSA. The physician was treated and the epidemic resolved. Plasmid and genotyping showed the ERSA organisms from both epidemics were the same. The employee implicated in the first epidemic did not have the epidemic strain, but the physician who attended during both epidemics did. CONCLUSIONS--Traditional epidemic investigations may lead to false conclusions that can only be recognized with molecular epidemiologic techniques. For these techniques to be useful in the control of outbreaks, plasmid analysis and/or genotyping must be readily available.