Utility of CT after sonography for suspected appendicitis in children: integration of a clinical scoring system with a staged imaging protocol.

Abstract

To improve diagnosis of pediatric appendicitis, many institutions have implemented a staged imaging protocol utilizing ultrasonography (US) first and then computed tomography (CT). A substantial number of children with suspected appendicitis undergo CT after US, and the efficient and accurate diagnosis of pediatric appendicitis continues to be challenging. The objective of the study is to characterize the utility of CT following US for diagnosis of pediatric appendicitis, in conjunction with a clinical appendicitis score (AS). Imaging studies of children with suspected appendicitis who underwent CT after US in an imaging protocol were retrospectively reviewed by three radiologists in consensus. Chart review derived the AS (range 0-10) and obtained the patient diagnosis and disposition, and an AS was applied to each patient. Clinical and radiologic data were analyzed to assess the yield of CT after US. Studies of 211 children (mean age 11.3 years) were included. The positive threshold for AS was determined to be 6 out of 10. When AS and US were concordant (N = 140), the sensitivity and specificity of US were similar to CT. When AS and US were discordant (N = 71) and also when AS ≥ 6 (N = 84), subsequent CT showed superior sensitivity and specificity to US alone. In the subset where US showed neither the appendix nor inflammatory change in the right lower quadrant (126/211, 60 % of scans), when AS < 6 (N = 83), the negative predictive value (NPV) of US was 0.98. However, when AS ≥ 6 (N = 43), NPV of US was 0.58, and the positive predictive value of subsequent CT was 1. There was a significant decrease in depiction of the appendix on US with patient weight-to-age ratio of >6 (kg/year, P < 0.001) and after-hours (1700 -0730 hours) performance of US (P < 0.001). Results suggest that the appendicitis score has utility in guiding an imaging protocol and support the contention that non-visualization of the appendix on US is not intrinsically non-diagnostic. There was little benefit to additional CT when AS < 6 and US did not show the appendix or evidence of inflammation; this would have avoided CT in 140/211 (66 %) patients. CT demonstrated benefit when AS ≥ 6, suggesting that cases with AS ≥ 6 and features that limit depiction of the appendix on US may be triaged to CT.