Evolution of Diagnosing Appendicitis in a Busy Pediatric Emergency Department—Has Ultrasound-First Become the Gold Standard?

Introduction:  Acute appendicitis remains a leading cause of pediatric non-traumatic abdominal pain. Our previous work has shown that the Pediatric Appendicitis Score with selective ultrasonography (PAS+US), has a high... [ view full abstract ]

Introduction: 

Acute appendicitis remains a leading cause of pediatric non-traumatic abdominal pain. Our previous work has shown that the Pediatric Appendicitis Score with selective ultrasonography (PAS+US), has a high diagnostic sensitivity and specificity. We sought to evaluate the sensitivity and specificity of US first compared to PAS+US combined and hypothesize that US first may have equal diagnostic accuracy for acute appendicitis in the emergency department (ED).

Methods:

An IRB approved retrospective review of patients with suspected appendicitis seen at our institution between 10/2010 and 12/2014 was performed. Patients who were transferred from outside hospital with prior imaging were excluded. Sensitivity and specificity were estimated for each group, as were length of ED stay and costs. The Clopper-Pearson method was used to compute 95% confidence intervals.

Results:

2645 patients were included (median age (IQR) 11.1years (7.4, 14.6)), 54.5% female, 42.4% Non-Hispanic Black or African American, 39.5% Non-Hispanic White, and 11.5% Hispanic. Based on the diagnostic pathway used, patients were grouped into 4 subpopulations (sensitivity, specificity): LeB-P (n=251; 82.1%, 96.1%), US first (n=1022; 84.7%, 96%), CT first (n=168; 87.9%, 98.2%) or no imaging (1204; 73.9%, 99.9%). Length of ED stay and hospital costs were comparable in the LeB-P and US-first groups but higher in the CT first group (Median LOS hours; 3.4, 3.8, 4.3, p<.0001. Median cost, dollars; 2075.06, 1713.85, 3422.46, respectively).

Conclusion: 

Compared to combined PAS+US, use of US first without use of the PAS may have comparable diagnostic accuracy, length of ED stay and costs, with appropriate patient selection.