BACKGROUND:

Methicillin-resistant Staphylococcus aureus (MRSA) has recently emerged as a common cause of infection in children in many parts of the world. The epidemiology of community-acquired MRSA (CA-MRSA) among healthy children has been recently described. However, little is known about CA-MRSA in children with underlying medical conditions.

OBJECTIVE:

To compare the clinical and molecular epidemiology of CA-MRSA in children with and without risk factors for health care-associated infections (RF-HAI).

METHODS:

We conducted a 3-year retrospective cohort study of children with CA-MRSA infection. RF-HAI, including hospitalization within the past year, indwelling medical devices or chronic medical condition, were identified by chart review. Genetic relatedness of CA-MRSA strains was assessed by pulsed field gel electrophoresis. Polymerase chain reaction was used to detect Panton-Valentine leukocidin and determine staphylococcal chromosomal cassette carrying the mecA methicillin-resistant gene (SCCmec) type.

RESULTS:

We identified 446 episodes of community-acquired S. aureus infections, of which 134 (30%) were caused by MRSA. During the 3-year study period, the proportion of S. aureus infections caused by MRSA rose from 15% (12 of 80) to 40% (93 of 235) (P < 0.001) with the increase noted predominately in children with skin and soft tissue infections. RF-HAI were identified in 56 (42%) patients with CA-MRSA. Among subjects with CA-MRSA, children with RF-HAI were more likely to have had an invasive infection than healthy children (32% versus 5%; P < 0.001). CA-MRSA isolates from children with RF-HAI were similar to those without RF-HAI; all laboratory-retained CA-MRSA isolates harbored the SCCmec type IV cassette, and almost all isolates were susceptible to trimethoprim-sulfamethoxazole and clindamycin. However, pulsed field gel electrophoresis revealed greater molecular diversity among CA-MRSA isolates recovered from children with RF-HAI compared with those from otherwise healthy children (P = 0.001). Additionally CA-MRSA isolates from children with RF-HAI were less likely to contain sequences for Panton-Valentine leukocidin (P < 0.001) and more likely to be resistant to 3 or more classes of antibiotics (P = 0.033).

CONCLUSION:

CA-MRSA strains recovered from children with RF-HAI were phenotypically similar to those recovered from healthy children The absence of SCCmec type II or III MRSA among children with RF-HAI suggests that CA-MRSA strains might have become endemic within pediatric health care facilities.