Blood Culture Improvement Guidelines and Diagnostic Stewardship for Antibiotic Reduction in Critically Ill Children

The Bright STAR Collaborative (BSC), or Blood Culture Improvement Guidelines and Diagnostic Stewardship for Antibiotic Reduction in Critically Ill Children Collaborative, is a multicenter quality improvement program to reduce blood culture use within pediatric intensive care units. Investigators will use data collected by participating sites to determine whether reliable implementation of clinical practice guidelines for evaluation of patients with signs and symptoms of sepsis can decrease antibiotic use in pediatric intensive care units. Investigators will perform a quasi-experimental study to compare outcome data in pre- and post- periods. Greater than or equal to 10 institutions will participate in this collaborative. Participating institutions will develop and implement an evidenced-based clinical decision-making tool as part of their quality improvement (QI) program in their pediatric intensive care unit (PICU). Aim 1: To determine if reliable implementation of clinical practice guidelines for evaluation of patients with signs and symptoms of sepsis can decrease blood culture use in pediatric intensive care units. Aim 2: To determine if reliable implementation of clinical practice guidelines for evaluation of patients with signs and symptoms of sepsis can decrease central line-associated bloodstream infections in pediatric intensive care units. Aim 3. To determine if reliable implementation of clinical practice guidelines for evaluation of patients with signs and symptoms of sepsis can reduce antibiotic use and Clostridium difficile infection. Aim 4. To determine whether a clinical practice guideline for evaluation of patients with signs and symptoms of sepsis in the PICU has an unintended consequence of patient harm. Aim 5. To evaluate the implementation of a multi-institutional quality improvement initiative and identify strategies for successful scale-up and adoption of similar practice guidelines in other clinical settings. Variables: blood cultures and central line-associated blood stream infections (CLABSIs), antibiotic use, , episodes of Clostridium difficile infection mortality, length of stay, ICU readmission, hospital readmission, episodes of sepsis, and episodes of septic shock. Analyses: The analytic approach equates to estimating and comparing the blood culture incidence during the "baseline/pre-implementation" and "post-implementation" periods, using a generalized linear mixed model (GLMM) assuming a Poisson distribution for the monthly number of blood cultures with the monthly number of patient days as an offset. Similar analyses will be conducted to evaluate the incidence of blood cultures drawn from central lines and CLABSIs. Due to the expected low incidence of CLABSIs, investigators will define time in quarters, not months, for that outcome. Similar analyses will be performed for secondary outcomes.