Development and validation of an ultrasound scoring system for children with suspected acute appendicitis.

Abstract

To facilitate consistent, reliable communication among providers, we developed a scoring system (Appy-Score) for reporting limited right lower quadrant ultrasound (US) exams performed for suspected pediatric appendicitis. The purpose of this study was to evaluate implementation of this scoring system and its ability to risk-stratify children with suspected appendicitis. In this HIPAA compliant, Institutional Review Board-approved study, the Appy-Score was applied retrospectively to all limited abdominal US exams ordered for suspected pediatric appendicitis through our emergency department during a 5-month pre-implementation period (Jan 1, 2013, to May 31, 2013), and Appy-Score use was tracked prospectively post-implementation (July 1, 2013, to Sept. 30,2013). Appy-Score strata were: 1 = normal completely visualized appendix; 2 = normal partially visualized appendix; 3 = non-visualized appendix, 4 = equivocal, 5a = non-perforated appendicitis and 5b = perforated appendicitis. Appy-Score use, frequency of appendicitis by Appy-Score stratum, and diagnostic performance measures of US exams were computed using operative and clinical finding as reference standards. Secondary outcome measures included rates of CT imaging following US exams and negative appendectomy rates. We identified 1,235 patients in the pre-implementation and 686 patients in the post-implementation groups. Appy-Score use increased from 24% (37/155) in July to 89% (226/254) in September (P < 0.001). Appendicitis frequency by Appy-Score stratum post-implementation was: 1 = 0.5%, 2 = 0%, 3 = 9.5%, 4 = 44%, 5a = 92.3%, and 5b = 100%. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were 96.3% (287/298), 93.9% (880/937), 83.4% (287/344), and 98.8% (880/891) pre-implementation and 93.0% (200/215), 92.6% (436/471), 85.1% (200/235), and 96.7% (436/451) post-implementation - only NPV was statistically different (P = 0.012). CT imaging after US decreased by 31% between pre- and post-implementation, 8.6% (106/1235) vs. 6.0% (41/686); P = 0.048). Negative appendectomy rates did not change (4.4% vs. 4.1%, P = 0.8). A scoring system and structured template for reporting US exam results for suspected pediatric appendicitis was successfully adopted by a pediatric radiology department at a large tertiary children's hospital and stratifies risk for children based on their likelihood of appendicitis.