Development and Evaluation of the American College of Surgeons NSQIP Pediatric Surgical Risk Calculator

J Am Coll Surg. 2016 Nov;223(5):685-693. doi: 10.1016/j.jamcollsurg.2016.08.542. Epub 2016 Sep 22.

Abstract

Background: There is an increased desire among patients and families to be involved in the surgical decision-making process. A surgeon's ability to provide patients and families with patient-specific estimates of postoperative complications is critical for shared decision making and informed consent. Surgeons can also use patient-specific risk estimates to decide whether or not to operate and what options to offer patients. Our objective was to develop and evaluate a publicly available risk estimation tool that would cover many common pediatric surgical procedures across all specialties.

Study design: American College of Surgeons NSQIP Pediatric standardized data from 67 hospitals were used to develop a risk estimation tool. Surgeons enter 18 preoperative variables (demographics, comorbidities, procedure) that are used in a logistic regression model to predict 9 postoperative outcomes. A surgeon adjustment score is also incorporated to adjust for any additional risk not accounted for in the 18 risk factors.

Results: A pediatric surgical risk calculator was developed based on 181,353 cases covering 382 CPT codes across all specialties. It had excellent discrimination for mortality (c-statistic = 0.98), morbidity (c-statistic = 0.81), and 7 additional complications (c-statistic > 0.77). The Hosmer-Lemeshow statistic and graphic representations also showed excellent calibration.

Conclusions: The ACS NSQIP Pediatric Surgical Risk Calculator was developed using standardized and audited multi-institutional data from the ACS NSQIP Pediatric, and it provides empirically derived, patient-specific postoperative risks. It can be used as a tool in the shared decision-making process by providing clinicians, families, and patients with useful information for many of the most common operations performed on pediatric patients in the US.

Publication types

  • Evaluation Study

MeSH terms

  • Adolescent
  • Child
  • Child, Preschool
  • Decision Support Techniques*
  • Female
  • Health Status Indicators*
  • Humans
  • Infant
  • Infant, Newborn
  • Logistic Models
  • Male
  • Patient Participation
  • Pediatrics
  • Postoperative Complications / epidemiology
  • Postoperative Complications / etiology*
  • Preoperative Care / methods*
  • Risk Adjustment
  • Risk Assessment / methods
  • Risk Factors
  • Societies, Medical
  • Specialties, Surgical
  • Surgical Procedures, Operative / mortality*
  • United States