Objective: Complete review of wireless capsule endoscopy (WCE) recordings by a physician is time-consuming and laborious and may be perceived as a limitation to perform WCE. The aim of this study was to evaluate the efficacy of a nurse in interpreting WCE.
Methods: A total of 102 WCE videos were evaluated by a single gastroenterologist and a nurse experienced as an assistant in diagnostic and interventional endoscopy and trained in WCE. After independently reviewing WCE videos, the two readers discussed their findings and came to a consensus.
Results: The mean capsule reading time was significantly longer for the nurse compared with the gastroenterologist (117.3+/-24.8 vs. 63.8+/-8.5 min, P<0.001). No statistical differences were observed regarding the correct recognition of first gastric, duodenal and caecal images between the two readers. For the gastroenterologist, both sensitivity and specificity in detecting abnormal findings were 100% except for angiodysplasia [sensitivity 88.5%, 95% confidence interval (CI): 70-97.4]. For the nurse, the lowest sensitivity rates were in detecting polyps (70%, 95% CI: 34.9-92.3) and angiodysplasias (92.3%, 95% CI: 74.8-98.9). The interobserver agreement as determined by Cohen's κ coefficient was excellent except for polyps (k=0.71, 95% CI: 0.46-0.96).
Conclusion: A trained nurse is highly accurate in detecting abnormal findings and interpreting WCE recordings. Physician's role could be limited to consider and confirm thumbnails created by a nurse.