Changing epidemiology of methicillin-resistant Staphylococcus aureus colonization in pediatric intensive care units over ten years

Changing Epidemiology of Methicillin-Resistant Staphylococcus aureus Colonization in Pediatric Intensive Care Units

Background:   Colonization with methicillin-resistant Staphylococcus aureus (MRSA) increases the risk for hospital-onset MRSA infection in children in intensive care units (ICUs). Community-associated MRSA (CA-MRSA) causes an increasing proportion of hospital-onset infections, but the prevalence of CA-MRSA colonization in pediatric ICUs, other than neonatal ICUs, has not been previously reported.

Objective:   To determine the prevalence of MRSA colonization and the prevalence of CA-MRSA colonization among children admitted to ICUs at a quaternary care pediatric hospital over the past decade. 

Methods:    A retrospective cohort study of MRSA-colonized neonatal, medical/surgical, and cardiac ICU patients at Children's Hospital Boston from 2001-09, identified by review of microbiology and infection control records was performed.  Frozen first isolates of MRSA obtained from admission and weekly active surveillance cultures from nares were characterized by spa type with polymerase chain reaction (PCR) and sequencing, staphylococcal cassette chromosome (SCC mec) type with multiplex PCR, and the presence of the genes encoding Panton-Valentine leukocidin (PVL) with PCR. Antibiotic susceptibilities of MRSA isolates and patient demographics were obtained by chart review. CA-MRSA was defined as the presence of spa type 1. Trends in the prevalence of spa types, SCC mec types, PVL and specific antibiotic susceptibilities were analyzed using the Cochran-Armitage trend test.

Results:   The proportion of patients colonized with MRSA remained stable (average=3.3%) over time, although the total number detected/screened increased from 21/746 in 2001 to 107/3470 in 2009. The proportion of isolates identified as CA-MRSA with spa type 1 increased significantly (p<0.001) over time, with a maximum of 38.2% in 2009. The proportion of MRSA isolates carrying SCC mec type IV and PVL also increased significantly, with maximums in 2009 of 56.6% (p=0.003) and 30.3% (p<0.001) respectively. Antibiotic susceptibility patterns of MRSA isolates changed significantly, with an increasing proportion of isolates susceptible to clindamycin, gentamicin, tetracycline and trimethoprim/sulfamethoxazole (all p-values <0.001). Specifically, the proportion of MRSA isolates susceptible to clindamycin and trimethoprim/sulfamethoxazole were 0% and 58.8% respectively in 2001 and 55.6% and 98.8% respectively in 2009. 

Conclusions: From 2001 to 2009 the prevalence of MRSA colonization in our pediatric ICUs remained low and stable, but an increasing proportion of colonizing isolates were CA-MRSA. Further research is needed to assess whether hospital-onset MRSA infections in pediatric ICU patients are increasingly caused by community strains.