Initial Performance of Radiologists and Radiology Residents in Interpreting Low-Dose (2-mSv) Appendiceal CT

Abstract

OBJECTIVE. The objective of our study was to prospectively evaluate the initial diagnostic performance and learning curve of a community of radiologists and residents in interpreting 2-mSv appendiceal CT. SUBJECTS AND METHODS. We included 46 attending radiologists and 153 radiology residents from 22 hospitals who completed an online training course of 30 2-mSv CT cases. Appendicitis was confirmed in 14 cases. Most of the readers had limited (<= 10 cases, n = 32) or no (n = 118) prior experience with low-dose appendiceal CT. The order of cases was randomized for each reader. A multireader multicase ROC analysis was performed. Generalized estimating equations were used to model the learning curves in diagnostic performance. RESULTS. Diagnostic performance gradually improved with years of training. The average AUC was 0.94 (95% CI, 0.90-0.98), 0.92 (0.88-0.96), 0.90 (0.85-0.96), and 0.86 (0.80-0.92) for the attending radiologists, senior residents, 2nd-year residents, and 1st-year residents, respectively. We did not observe any notable intrareader learning curves over the training course of the 30 cases except a decrease in reading time. Diagnostic accuracy and sensitivity were significantly affected by the reader training level and prior overall experience with appendiceal CT but not by the prior specific experience with low-dose appendiceal CT. CONCLUSION. The learning curve is likely prolonged and forms gradually over years by overall radiology training and clinical experience in general rather than by experience with low-dose appendiceal CT specifically.