In December 1989, 11 cases of Serratia marcescens septicemia occurred in a neonatal Intensive Care Unit (NICU). All isolates from blood of the 11 cases had the same antibiogram. In the preceeding year no S. marcescens septicemias were diagnosed in the same NICU. Medical records for all admissions to the NICU in December were reviewed. A case-control study was done to compare exposures between infants with S. marcescens septicemia (cases) and unaffected infants (controls). The case-control study showed that 91% of the cases had a stopcock on the intravenous infusion line compared with 18% of controls (OR 47, 95% CI 5-2104). Several drugs commonly given via the stopcock were also associated with cases. The external surfaces of 2 stopcocks were culture-positive for S. marcescens with the same antibiogram. Nurses were instructed to always wash hands before using the stopcock, not to give drugs through stopcock and never to leave syringes with heparin flush solutions attached to the stopcock. For the following four months no new S. marcescens septicemias were detected in the NICU. This outbreak probably resulted from contamination of the stopcock surfaces and the heparin flush syringe.
Conclusion: