Quantifying the Burden of Interhospital Cost Variation in Pediatric Surgery Implications for the Prioritization of Comparative Effectiveness Research

IMPORTANCE Practice variation is believed to be a driver of excess health care spending, although few objective data exist to guide the prioritization of comparative effectiveness research (CER) in pediatric surgery. OBJECTIVE To identify high-priority general pediatric surgical procedures for CER on the basis of the following 2 complementary measures: the magnitude of interhospital cost variation as a surrogate for the need for and potential effect of CER at the patient level and the cumulative fiscal burden of this cost variation when considering the case volume from all hospitals as a surrogate for public health relevance. DESIGN, SETTING, AND PARTICIPANTS This was a cohort study of patients undergoing 1 of the 30 most costly pediatric surgical operations at 45 children's hospitals between January 1, 2014, and September 30, 2015. Cost data were extracted from the Pediatric Health Information System database and adjusted for differences in unit-based costing at the hospital level and for differences in case mix and disease severity at the patient level. MAIN OUTCOMES AND MEASURES First, the width of the interquartile range (WIQR) of the adjusted procedure-specific median cost across hospitals. Second, the procedure-specific cost variation burden, which was calculated as the aggregate sum of absolute cost differences between the overall adjusted median cost derived from all patients treated at all hospitals and the adjusted cost of each individual patient treated at all hospitals. RESULTS A total of 92 535 encounters were analyzed. The median number of encounters per hospital was 2011 (interquartile range [IQR], 1224-2619), and the median number of encounters per procedure was 610 (IQR, 442-2610). In the final cohort, 66.9%(n = 61 933) of the patients were male, and the median age was 7 years (IQR, 1.9-12.3 years). Cost variation at the hospital level was greatest for gastroschisis (WIQR, $48471; median, $111566 [IQR, $91195-$ 139 936]), congenital diaphragmatic hernia (WIQR, $43948; median, $154730 [IQR,$129764-$173712]), tracheoesophageal fistula/esophageal atresia (WIQR, $39206; median, $105259 [IQR, $87335-$126541]), and total colectomy for ulcerative colitis (WIQR, $24497; median, $34910 [IQR,$28815-$53312]). The following 5 diagnoses accounted for 52.5% of the cumulative cost variation burden from all 30 conditions: uncomplicated appendicitis (18.0%[$66205117]), complicated appendicitis (14.1%[$51702402]),gastroschisis (9.5%[$34940331]), gastrostomy (5.8% [$21227436]), and small-intestinal atresia (5.1%[$18840546]).Neonatal cases contributed 3.6% of the case volume and accounted for 26.8% of the cumulative cost variation burden from all 30 conditions. CONCLUSIONS AND RELEVANCE A small number of procedures account for most of the cost variation burden in pediatric surgery, with some demonstrating wide cost variation among hospitals. Gastroschisis and small-intestinal atresia may be particularly high-yield targets for multidisciplinary CER efforts, while the management of appendicitis and gastrostomy should be considered high-priority conditions among pediatric surgeons.