CT in Differentiating Complicated From Uncomplicated Appendicitis: Presence of Any of 10 CT Features Versus Radiologists' Gestalt Assessment.

Abstract

OBJECTIVE. The purpose of this study is to propose a sensitive CT criterion (the presence of any of 10 CT features) for complicated appendicitis that could be used in the nonoperative management of appendicitis and to compare the diagnostic performance of this sensitive CT criterion with that of gestalt assessment. MATERIALS AND METHODS. This retrospective study, which was conducted in a tertiary teaching hospital, included 100 patients with suspected appendicitis on CT. Complicated appendicitis, defined as gangrenous or perforated appendicitis, was pathologically or surgically confirmed in 32 patients. Six radiologists independently determined the presence of 10 previously reported CT features of complicated appendicitis (contrast enhancement defect of the appendiceal wall, abscess, extraluminal air, intraluminal air, extraluminal appendicolith, intraluminal appendicolith, moderate-to-severe periappendiceal fat stranding, periappendiceal fluid, ileus, and ascites) and rated the likelihood score for complicated appendicitis using gestalt assessment. The sensitivity and specificity of CT for complicated appendicitis were measured by the presence of any of 10 CT features (the any-of-10-features criterion) and by the radiologists' gestalt assessment. Pooled sensitivity and specificity were compared using a generalized linear mixed model. RESULTS. The pooled sensitivity of the presence of any of 10 CT features was higher than that of gestalt assessment (92% vs 64%; difference, 28% [95% CI, 10-46%]; p < 0.001), although the pooled specificity was lower (43% vs 76%; difference, -33% [95% CI, -48% to -17%]; p < 0.001). CONCLUSION. The pooled sensitivity of the presence of any of 10 CT features was higher than that of gestalt assessment, at the cost of lower specificity. For prudent selection of patients who should receive nonoperative treatment of appendicitis, the any-of-10-features criterion may be used to decrease treatment failure associated with a false-negative diagnosis of complication.