Risk-Adjusted Incidence Of Prolonged Postoperative Critical Care Varies Widely Across Congenital Heart Surgery Centers

Introduction: Prolonged critical care (PCC) after congenital heart surgery (CHS) is associated with poor outcomes, but previous literature has failed to provide actionable insights for clinicians. We aimed to determine the clinical predictors of PCC after CHS specific to neonates and non-neonates, and explore variation in PCC across Pediatric Cardiac Critical Care Consortium (PC4) centers. Methods: Observational analysis of the PC4 clinical registry; all patients admitted to the cardiac intensive care unit (CICU) after index CHS from 8/2014-12/2016 were included and stratified into neonates (≤28 days) and non-neonates. PCC was defined as critical care duration ≥ 90 th percentile within strata. Multivariable logistic regression identified independent predictors of PCC (p 1 respectively, with a 95% confidence interval not including 1. Results: There were 2419 neonates and 10687 non-neonates from 23 centers; the PCC cutoff for each stratum was ≥35 and ≥10 days, respectively. Complications were more common in the PCC group (93% vs 58% in neonates and 77% vs 21% in non-neonates, respectively; p Conclusions: The incidence of PCC after CHS varies widely across centers. While many PCC predictors are not modifiable, we identified several complications, particularly in non-neonates, that may be targets for quality improvement. Identifying practices at centers with lower-than-expected PCC could lead to initiatives that reduce complications and shorten the critical illness period.