863

Abstract

Introduction: Severe sepsis in children causes over 4,000 deaths & costs nearly $2B annually (Goldstein, 2005). Additionally, there exists significant variability in the recognition, evaluation & management of sepsis (Carcillo, 2006). 1st generation definitions & criteria for pediatric SIRS, sepsis(S), severe sepsis (SS), & septic shock (SSh) [2005 Sepsis] were published. SS was further parsed into cardiovascular (CV), respiratory (Resp) & Other (Oth). In 2009, an update further delineated pediatric practice parameters. As guidelines continue to be developed, the need for iterative derivation & validation is necessary to optimize efficiency. Hypothesis: We hypothesized that the efficacy of sepsis evaluation in the Pediatric Intensive Care Unit (PICU) can be quantified using the 2005 sepsis criteria. Methods: We reviewed all PICU admissions < 18 yrs old from Jul-Dec 2011 with length of stay > 3 days. Each SIRS event was analyzed independently. The 2005 criteria were used to define SIRS, S, & SS. Additionally, we defined Clinical Sepsis (CS) as treatment with IV antibiotics for > 2days. Multiple organ system failure (MOSF) was identified when 2+ SS domains were satisfied. Results: There were 384 evaluations in 150 patients & SIRS was identified in 239 (62%). SIRS eventuated in S & SS in 109(28%) & 140(37%) cases, respectively. Of the 140 cases of SS, CV, Resp & Oth accounted for 49%, 41%, & 10%, respectively. MOSF occurred in 47% of SS cases. CS occurred in 22% of evaluations. Of the 109 evaluations that met sepsis criteria, 56(51%) were treated for 2+ days. In contrast, 83 evaluations were treated for CS, 56 (67%) met sepsis criteria. Finally, WBC abnormality (vs temperature, HR & RR) was associated with the highest AUC-ROC (SEM) when S 0.68(0.027), MOSF 0.723(0.034) & CS 0.591(0.031) were outcomes. Conclusions: There is significant incoherence between the guidelines & clinical practice. Half of the evaluations meeting criteria were not treated. Conversely, one-third of those clinically treated did not meet criteria. Further efforts are underway to optimize the predictive value of the guidelines to treat those most at risk & minimize wasteful evaluations in PICU patients with non-infectious SIRS.