Detection and False-Referral Rates of 2-mSv CT Relative to Standard-Dose CT for Appendiceal Perforation: Pragmatic Multicenter Randomized Controlled Trial.

The objective of our study was to compare diagnostic performance of 2-mSv CT and standard-dose CT (SDCT) for the diagnosis of perforated appendicitis in adolescents and young adults. We used the intention-to-treat analysis set of a pragmatic randomized controlled trial involving 3074 patients (age range, 15-44 years) with suspected appendicitis and 161 radiologists from 20 hospitals. The patients were randomized to undergo either 2-mSv CT or SDCT. Predefined endpoints were sensitivity and specificity. Considering potential verification bias caused by the difference in diagnostic interventions (2-mSv CT vs SDCT), we added endpoints of detection rate (DR) and false-referral rate. The reference standards were surgical or pathologic findings. We used Fisher exact tests. Sensitivity analyses included the following: first, a per-protocol analysis; second, an analysis of a surgical reference standard but not a pathologic reference standard; and, third, an analysis to adjust for site clustering. We tested for heterogeneity in DR and false-referral rate across various patient and hospital characteristics. The 2-mSv CT and SDCT groups were comparable regarding DR (5.1% [78/1535] vs 4.9% [76/1539]; 95% CI for the difference, -1.4 to 1.7 percentage points; = 0.87), false-referral rate (3.1% [48/1535] vs 3.1% [47/1539]; 95% CI for the difference, -1.2 to 1.3 percentage points; = 0.92), sensitivity (42.9% [78/182] vs 43.2% [76/176]; 95% CI for the difference, -10.6 to 9.9 percentage points; > 0.99), and specificity (89.2% [305/342] vs 91.2% [354/388]; 95% CI for the difference, -6.4 to 2.3 percentage points; = 0.38). Sensitivity analyses showed similar results. We found no significant subgroup heterogeneity. The performance of 2-mSv CT, with limited sensitivity but high specificity, is comparable to that of SDCT for the diagnosis of appendiceal perforation.