Ribotyping provides efficient differentiation of nosocomial Serratia marcescens isolates in a pediatric hospital.

Ribotyping with a nonradioactive probing system was used for the epidemiological evaluation of 15 Serratia marcescens nosocomial strains isolated from the stools of 12 children with no apparent illness in five different hospital wards over a 20-day period. Our results indicate that the occurrence of S. marcescens colonization was the result of the spread of a single epidemiological strain in the hematology ward, the oncology ward, and the gastroenterology ward and in two neonates in the neonatology ward, suggesting cross-contamination between the patients in these four wards. This isolate was genotypically unrelated to the bacterial strain found in the three other patients in the neonatology ward. Interestingly, one patient in the neonatology ward harbored these two genotypically different strains. Finally, the patient in the intensive care unit was colonized with a different strain. We find ribotyping to be a more reliable technique than biochemical typing. The results of ribotyping are more easily interpreted than are those of total DNA analysis, with an equivalent degree of discrimination.