Hospital Variation in Mortality after Inpatient Pediatric Surgery

Objective: The aim was to determine the association between risk adjusted hospital perioperative mortality rates, postoperative complications, and failure to rescue (FTR) after inpatient pediatric surgery. Background: FTR has been identified as a possible explanatory factor for hospital variation in perioperative mortality in adults. However, the extent to which this may be the case for hospitals that perform pediatric surgery is unclear. Methods: The Pediatric Health Information System database (2012-2020) was used to identify patients who underwent one of 57 high-risk operations associated with significant perioperative mortality (n=203,242). Academic, pediatric hospitals (n=48) were stratified into quintiles based on risk adjusted inpatient mortality [lower than average, quintile 1 (Q1); higher than average, quintile 5 (Q5)]. Multivariable hierarchical regression was used to evaluate the association between hospital mortality rates, complications, and FTR. Results: Inpatient mortality, complication, and FTR rates were 2.3%, 8.8%, and 8.8%, respectively. Among all patients who died after surgery, only 34.1% had a preceding complication (Q1, 36.1%; Q2, 31.5%; Q3, 34.7%; Q4, 35.7%; Q5, 32.2%; trend test, P=0.49). The rates of observed mortality significantly increased across hospital quintiles, but the difference was <1% (Q1, 1.9%; Q5; 2.6%; trend test, P<0.01). Relative to Q1 hospitals, the odds of complications were not significantly increased at Q5 hospitals [odds ratio (OR): 1.02 (0.87-1.20)]. By comparison, the odds of FTR was significantly increased at Q5 hospitals [OR: 1.60 (1.30-1.96)] with a dose-response relationship across hospital quintiles [Q2 - OR: 0.99 (0.80-1.22); Q3 - OR: 1.26 (1.03-1.55); Q4 - OR: 1.33 (1.09-1.63)]. Conclusions: The minority of pediatric surgical deaths are preceded by a postoperative complication, but variation in risk adjusted mortality across academic, pediatric hospitals may be partially explained by differences in the recognition and management of postoperative complications. Additional work is needed to identify children at greatest risk of postoperative death from perioperative complications as opposed to those at risk from pre-existing chronic conditions.