Does the American College of Surgeons NSQIP-Pediatric Accurately Represent Overall Patient Outcomes?

BACKGROUND: The National Surgical Quality Improvement Program-Pediatrics (NSQIP-P) collects data for institutional quality benchmarking of surgery performed on children using a sampling algorithm. The Pediatric and Infant Case Log and Outcomes (PICaLO) is a database of all general and thoracic pediatric surgery (GTPS) procedures performed at our institution with the attendant complications. This study compared postsurgical occurrences in a NSQIP-P sample with all postoperative occurrences at a single institution to test the hypothesis that a sample of higher risk procedures represents the actual event rate for all higher risk procedures. STUDY DESIGN: The definitions of postoperative occurrences used in PICaLO are derived from NSQIP-P but tracked past 30 days postoperatively and include additional occurrences (ie, anastomotic leak). The number and types of occurrences and number of deaths from PICaLO and NSQIP-P databases were compared for procedures specific to pediatric GTPS procedures during 2012 to 2013. A chi-square test evaluated the proportion of occurrences and deaths in PICaLO to NSQIP-P. RESULTS: The NSQIP-P sampled 37.7% of eligible GTS procedures recorded in PICaLO during the study period. The proportion of cases with 1 or more occurrences was significantly higher in the NSQIP-P dataset when compared with all cases in PICaLO (p < 0.0001). When NSQIP-P and PICaLO were compared based on specific CPT codes, NSQIP-P still had a higher event rate (p = 0.004). CONCLUSIONS: In focused comparisons, the data demonstrate that the NSQIP-P sampling algorithm successfully identifies CPT codes with higher postoperative event rates than the overall cohort of pediatric GTPS patients, but may not be reflective of the total experience for procedures with those CPT codes.