Application of a biotyping system and DNA restriction fragment analysis to the study of Serratia marcescens from hospitalized patients

Use of pulsed-field gel electrophoresis to investigate an outbreak of Serratia marcescens infection in a neonatal intensive care unit

Serratia marcescens is a well-recognized hospital-acquired pathogen, which has been associated with a number of specific outbreaks, particularly in critically ill neonates. We used pulsed-field gel electrophoresis (PEGE) typing to analyse an outbreak in a neonatal intensive care unit (NICU). We included samples from nine patients, three handwashes and ten environmental isolates from an outbreak (February to August 1999) in addition to four patient isolates from different wards of our hospital during the same time period. The clinical presentations of the outbreak included bacteraemia (four cases), pneumonia (three cases), umbilical wound infection (one case) and conjunctivitis (one case). Nine outbreak isolates exhibited an identical PFGE fingerprint, while the epidemiologically unrelated strains demonstrated distinct patterns. Epidemiological investigation failed to reveal a common source of the outbreak, although the epidemic S. marcescens strain was isolated from hand-washes and doors of incubators. We concluded that cross-transmission via transient contamination of hands was the major route for this outbreak. Strict handwashing practices, the cohorting and isolation of colonized and infected patients, and the regular dis-infection of incubators are crucial steps for preventing the transmission of S. marcescens in an NICU. This PFGE method is highly discriminatory for the thorough epidemiological investigation of an outbreak of S. marcescens.

Role of molecular epidemiology in infection control

Molecular typing methods have enabled infection control personnel to investigate outbreaks and endemic nosocomial infections more quickly and thoroughly than they could have with basic epidemiologic and microbiologic methods. This article reviews molecular typing methods that have been used successfully in the practice of hospital epidemiology. Included is an explanation of the basic principles of these methods and a description of their strengths and weaknesses.

Use of pulsed-field gel electrophoresis typing to study an outbreak of infection due to Serratia marcescens in a neonatal intensive care unit

Serratia marcescens is a well-known cause of nosocomial infections and outbreaks, particularly in critically ill neonates and immunocompromised patients. Numerous methods have been proposed for typing. We used pulsed-field gel electrophoresis (PFGE) typing to analyze an outbreak in a neonatal intensive care unit (NICU). We included 23 patient isolates from an outbreak (March to July 1995), and 10 patient isolates from different wards during the same time period. PFGE of whole-cell DNA digested by SpeI was used as a marker of strain identity. The most common presentation of the infection was sepsis in 18 of 23 (78%) neonates. Only four different biotypes were identified; biotype A8d accounted for 84% of the strains. PFGE typing revealed two clones responsible for two different clonal strain dissemination outbreaks from March to July, with 24 patient isolates being pattern A and 4 patient isolates being pattern E. PFGE typing suggests cross transmission between patients in the NICU and other wards. The isolates from 5 other patients showed distinct PFGE patterns. Extensive investigation and cultures failed to identify any environmental or staff reservoir of S. marcescens. This is one of the first reports applying PFGE to the study of S. marcescens, and this method was a useful marker of strain identity. PFGE typing distinguished strains which appeared to be the same by biotyping.

Typing of nosocomial strains of Serratia marcescens: comparison of restriction enzyme cleaved genomic DNA fragment (PFGE) analysis with bacteriocin typing, biochemical profiles and serotyping

Eighty-eight selected clinical isolates of Serratia marcescens, representing 27 putative outbreaks of nosocomial cross-infection encountered during 1980-1995, were tested comparatively by bacteriocin typing, carbon source assimilation tests, serotyping (O and H antigens), and restriction pattern (RFLP) analysis of restriction cleaved (SpeI, XbaI) genomic DNA fragments after pulsed-field gel electrophoresis (PFGE). Serotyping served as the "gold standard" of the phenotypic methods. One pseudo-outbreak (bacteriocin typing incriminated type 26) was uncovered through serotyping as well as the biochemical profile and confirmed by PFGE analysis of genomic DNA. Bacteriocin typing and determination of biochemical profiles disclosed several instances of phenotypic variation; serotyping revealed two episodes of shifts from motility (H12) to nonmotility. Resolution of restricted genomic DNA fragments with the PFGE procedure permitted detection of 27 PFGE patterns (A-M, N-1-N-3, O-1, O-2, P-1-P-3, Q-1-Q-3, R-1-R-3, S-1, and T). Based on the analysis of PFGE patterns against the background of epidemiological data, the number of nosocomically significant strains of S. marcescens could be reduced to 16 (PFGE patterns A-M, N-2, O-1, P-2, and T). It was concluded that PFGE analysis of restricted genomic DNA of S. marcescens was superior to the three phenotypic methods.

Phenotypic variation of clinical Serratia marcescens isolates repeatedly recovered from individual patients

A total of 129 selected isolates of Serratia marcescens which had been recovered from 50 patients during the 1980-1995 period and which revealed phenotypic variation in terms of bacteriocin (phage tail) susceptibility, carbon source assimilation, or serotype, were reexamined with these three phenotypic methods. Seven isolates (5.4%) were bacteriocin nontypable; all 129 isolates utilized carbon sources and could be serotyped. Fourty-eight isolates from 20 patients yielded unambiguous results with these 3 phenotypic methods and were excluded from further analysis. Among the remaining 81 isolates from 30 patients, isolates from 2 patients revealed phenotypic variation in bacteriocin susceptibility only, whereas isolates from 6 patients showed variant bacteriocin types and variant biochemical profiles, but were of identical serotype. Isolates from 20 patients revealed variant biochemical profiles only. Three patients had become superinfected with strains of S. marcescens of different phenotype and genotype. In 4 patients, previously motile (H12) isolates had become nonmotile (H-). PFGE analysis of XbaI and SpeI-restricted genomic DNA of the 81 isolates of the 30 patients demonstrated the isolates of 22 patients to be genotypically identical. The isolates from 3 patients were closely related by genotype, and those from an additional patient proved to be possibly related. PFGE analysis demonstrated one patient to have become infected by two genotypically different strains of S. marcescens of identical serotype, which, however, differed in bacteriocin type and biochemical profile. It was concluded that PFGE analysis of restricted genomic S. marcescens DNA was superior to the three phenotypic methods examined comparatively. Serotyping was more reliable than bacteriocin typing, and the latter technique yielded fewer phenotypic variants than determination of biochemical profiles among consecutively recovered isolates from patients with long-lasting S. marcescens infection.

Epidemiological survey of an outbreak of multiresistant Serratia marcescens by PCR-fingerprinting

During an outbreak of Serratia marcescens from May to November 1993 43 strains obtained from 27 ICU patients infected or colonized with multiresistant S. marcescens were genotypically characterized with random amplified polymerase chain reaction (RAPD-PCR)-fingerprinting. In addition, 43 epidemiologically unrelated control isolates were selected. PCR-fingerprinting identified ten different genotypes of S. marcescens among the outbreak related strains. One predominant genotype was demonstrated in 21/43 isolates of 11/27 patients. A cluster of this genotype was found in seven/eight patients on the cardiosurgical ICU. The epidemiologically unrelated strains all showed different genotypes as compared to the predominant type. This survey proved RAPD-PCR to be a highly discriminatory and reproducible method for epidemiological studies of S. marcescens strains in nosocomial outbreaks.