Benchmarking Pediatric Trauma Care in Mixed Trauma Centers: Adult Risk-Adjusted Mortality is not a Reliable Indicator of Pediatric Outcomes

Abstract

Background: Trauma center benchmarking has become standard practice for assessing quality. American College of Surgeons (ACS) adult trauma center verification standards do not specifically require participation in a pediatric-specific benchmarking program. Centers that treat adults and children may therefore rely solely on adult benchmarking metrics as a surrogate for pediatric quality. This study assessed discordance between adult and pediatric mortality within mixed trauma centers to determine the need to independently report pediatric-specific quality metrics. Study Design: A cohort of trauma centers (n=493, including 347 adult-only, 44 pediatric-only, and 102 mixed) that participated in the ACS Trauma Quality Improvement Program (TQIP) in 2017-2018 was analyzed. Center-specific observed-to-expected mortality estimates were calculated using TQIP adult inclusion criteria for 449 centers treating adults (16-65 years) and using TQIP pediatric inclusion criteria for 146 centers treating children (0-15 years). We then correlated risk-adjusted mortality estimates for pediatric and adult patients within mixed centers and evaluated concordance of their outlier status between adults and children. Results: The cohort included 394,075 adults and 97,698 children. Unadjusted mortality was 6.1% in adults and 1.2% in children. Mortality estimates had only moderate correlation (r=0.41) between adult and pediatric cohorts within individual mixed centers. Mortality outlier status for adult and pediatric cohorts was discordant in 31% (32/102) of mixed centers (weighted kappa-statistic 0.06 [-0.11, 0.22]), with 78% (n=23/32) of discordant centers having higher odds of mortality for children than for adults (6 centers with average adult mortality and high pediatric mortality and 17 centers with low adult mortality and average pediatric mortality, p<0.01). Conclusions: Adult mortality is not a reliable surrogate for pediatric mortality in mixed trauma centers. Incorporation of pediatric-specific benchmarks should be required for centers that admit children.