Three-year surveillance of community onset health care-associated Staphylococcus aureus infections in children

Background: Staphylococcus aureus causes skin and soft tissue or invasive infections in children in the community, in the hospital or in other ways associated with the health care system (HCA). Methods: Prospective community-acquired S. aureus infection surveillance at Texas Children's Hospital was initiated on August 1, 2001. Community onset HCA (CO HCA) infections were identified. Demographic and clinical data were collected. Antibiotic susceptibilities were determined. Data were analyzed by chi(2) or Student's t test. CO HCO-isolates were characterized by pulsed field get electrophoresis and staphylococcal chromosomal cassette carrying the mecA methicillin-resistant gene (SCCmec) typing. Results: From August 1, 2001 to July 31, 2004, 61.5% of 322 in year 1, 62.9% of 259 in year 2 and 56.9% of 318 in year 3 of CO HCA isolates were methicillin-resistant S. aureus (MRSA). Among the CO HCA-MRSA isolates, 8.9% of 542 were from children with invasive infections compared with 24.1% of 357 CO HCA-methicillin-susceptible S. aureus (MSSA; P < 0.001). Sixty-six percent of children with CO HCA-S. aureus isolates were admitted to the hospital. Clindamycin resistance increased over the 3 years (CO HCA-MRSA, from 3.5% to 18.8%, P < 0.001; CO HCA-MSSA, from 3.2% to 10.2%, P = 0.053). Thirty-three of 35 (94.3%) CO HCA-MRSA carried SCCmeclV; 30 were USA300. Only 3 of 35 MSSA were related to USA300 by pulsed field gel electrophoresis. Conclusions: CO HCA-S. aureus infections remained steady over the 3-year study at Texas Children's Hospital. Clindamycin resistance increased > 4-fold for CO HCA-S. aureus isolates over the 3 years and is no longer appropriate for empiric treatment of invasive infections suspected to be caused by CO HCA-MRSA at our hospital. In our setting, CO HCA-MRSA infections are steady in number despite substantial increases in community-acquired MRSA infections and both being related to the same clone.