P0607 - Purulent and non-purulent cellulitis in hospitalized children: Risk factors and outcomes

Background: Cellulitis is a growing burden to the pediatric population due to increased prevalence of community-acquired methicillin-resistant Staphylococcus aureus (MRSA). While cellulitis can be purulent and non-purulent, it is unknown if additional factors affect purulence risk and its effect on outcomes for hospitalized children.

Objective: To identify risk factors and outcomes associated with development of purulent cellulitis in children who require admission.
Design/

Methods: In this retrospective cohort study, cellulitis was identified by ICD-9 codes. Cases were reviewed for eligibility, including age less than 19 years and lack of serious chronic conditions. Standardized data extraction tool was used to collect relevant information. Wound and blood cultures, white blood cells (WBC) and absolute neutrophil count (ANC) as well as C-reactive protein (CRP) were analyzed. Purulent cellulitis was defined as presenting with abscess. Cellulitis cases were stratified as non-purulent (Group 1) and purulent (Group 2) for further statistical comparison. Descriptive and linear regression analysis were constructed to identify factors associated with purulent cellulitis and length of stay (LOS). Data were presented as mean, percentage, and 95% Confidence Interval (95%CI).

Results: Of 139 children included, 27.4% presented with abscess. Groups were comparable for demographics and medical history. Patients in Group 1 were on average 3 years younger and had one day shorter duration of illness than those in Group 2 (P<0.01). In Groups 1 and 2, 48.5% and 21.1% had cellulitis located on the head, respectively (P<0.001). Among wound cultures performed, MRSA and negative results were detected in 2.6% and 69.2% in Group 1 and 52.9% and 14.7% in Group 2, respectively (P<0.0001). Blood culture was negative in 99.2% of 125 patients tested. WBC, ANC and CRP showed no difference amongst the study groups. Regression models confirmed the role of older age, longer duration of illness, and non-head location on development of purulent cellulitis. LOS (days) in Group 1 was comparable to Group 2 (3.2, 95%CI 2.7-3.5 versus 3.4, 95% CI 2.8-4.0, P=0.5).
Conclusion(s): The likelihood of developing purulent cellulitis substantially increased in the presence of MRSA, especially in children with older age and prolonged duration of illness. No difference in level of inflammatory markers or LOS was found between patients with respect to cellulitis (purulent versus non-purulent) type.