OBSCURE BLEEDING USING WIRELESS CAPSULE ENDOSCOPY (WCE) CAN ANYBODY READ THEM?

R Enns,¹ M Brodie,¹ K Pluta,¹ C Lo¹, K Mergener²

Div. Gastroenterology, Dept. of Medicine, St. Paul's Hospital, University of British Columbia, Vancouver, British Columbia¹, and Section of Gastroenterology, Virginia Mason Medical Center, Seattle, Washington, USA²

BACKGROUND: Capsule endoscopy is a novel diagnostic method of imaging the GI tract using a wireless capsule which transmits images to a data recorder while the device traverses the small intestine. One of the drawbacks to this technology is that reviewing the images is time consuming (45-60 minutes). To avoid this, some capsular endoscopists are routinely having their staff (nursing/physician/technology assistants) interpret the results and provide a thumbnail sketch for review to the physician. This practise has not been well evaluated.
OBJECTIVE: To determine the inter-observer agreement between 'readers' of capsular endoscopy who are at varying levels of training.
METHODS: 100 capsular studies (with no accompanying information) were provided by a second investigator (KM). This ensured that all studies were 'blinded' to the investigators. The individuals reading the capsules all had observed at least several endoscopies including enteroscopies and had at least a basic training session of approx. 3 hours reviewing several capsular studies and their accompanying abnormalities. With the exception of the physician, none of the assistants had performed endoscopic procedures. 'Readers' of varying levels of training were included: summer research medical student (CL), full-time (non-nurse) research assistant (KP), capsular nurse (MB) and gastroenterologist with interest in obscure GI bleeding (RE). Only patients with obscure GI bleeding were included. The physician's interpretation was used as a 'gold-standard.' Endpoints in the study included: evaluation of the esophagus and lower esophageal sphincter (if visualized), time of exit through pylorus, small bowel lesions accounting for bleeding (definite and probable), incidental findings (i.e.lymphangiectasia), time to exit to the cecum.
RESULTS: To date 25 of the first 100 capsules have been reviewed. Definite lesions have been found on 52% of patients and recorded by all 'readers.' Kappa statistics appear reasonable in all areas that are being assessed. The time required for evaluation of each study appears to be longer by trainees than the physician.
CONCLUSION: Further data needs to be collected, however, this preliminary analysis suggests that nurses and other capsule endoscopy assistants can be used to 'screen' a capsule study, thereby decreasing the time required for review by the physician.