Two nosocomial outbreaks of sepsis caused by Serratia marcescens, which occurred in July 1999 and January 2002--Tokyo

Abstract

Two nosocomial outbreaks of sepsis caused by Serratia marcescens, which occurred in Tokyo were the following cases. CASE A: In July 1999, 10 inpatients admitted to the third floor ward of the General Hospital A, developed sudden onset of high fever, coagulation disorders (disseminated intravascular coagulation), and acute renal failure, of which 5 died. Twenty-one strains of Serratia marcescens were isolated from the inpatient's blood and urine, nurse fingers and environmental samples from floor and cooling tower. Serratia infection was strongly suspected as the cause of sepsis. These cases were defined as "inpatients who developed fever 38 degrees C or more during July 26 to 29 and from whom S. marcescens was isolated by blood culture". Ten isolates were detected from the blood. In order to investigate the background of S. marcescens isolation in the hospital and to compare molecular and biochemical characteristics of S. marcescens, cultures were attempted from samples of other inpatients and staffs and hospital environment. Those were classified into 9 groups by various different typings: biotyping with Api Rapid 20; susceptibility typing of antimicrobial agents tested; pulsed-field gel electrophoresis (PFGE) typing of SpeI- or Xba I-restricted chromosome. All 10 isolates causing sepsis were found to be in the same group. CASE B: In January 2002, 24 inpatients, admitted to Neurosurgical Hospital B, developed sudden onset of high fever, of which 7 died. S. marcescens was isolated from a towel, environmental samples and inpatients. These cases were defined as "inpatients who developed fever of 38.5 degrees C and S. marcescens isolated by blood culture". Twelve strains were isolated from the blood samples in 12 cases. In order to investigate the background of S. marcescens isolation in the hospital, cultures were attempted from other inpatient's urine and environmental samples from medical tape, Tshake and a towel. These isolates were classified into 3 groups by the previous typings; biotyping with Api Rapid 20; susceptibility typing of antimicrobial agents tested; and PFGE typing. All 12 isolates in 12 cases were found to be in the same group. These cases of 2 nosocomial outbreaks of sepsis were defined as "in-patient who developed high fever and S. marcescens isolated by blood culture". However in both cases transmission routes of Serratia infection remain unknown by field investigation.