BACKGROUND
Capsule endoscopy is a common, pain-free diagnostic procedure for the small bowel. However, interpretation of the whole video recording is a time-consuming and costly procedure that can take up to 2 hours. The aim of the present study is two-fold: first to study the accuracy of capsule endoscopy analysis between a trained endoscopy nurse and a physician and secondly to determine if pre-evaluation by nursing staff might be time-effective for capsule reading. This study is especially important given the increasing financial pressure on current health-care systems.
METHODS
A long-standing experienced endoscopy nurse, who was trained to read capsule endoscopy, and a physician, both blinded to the patient diagnosis and the other clinical findings reviewed 48 consecutive capsule endoscopy videos. The analyses of both the nurse and the physician were re-evaluated by an independent doctor regarding the agreement of the marked findings.
RESULTS
Total time to read capsule endoscopy was significantly longer for the nurse's interpretation (63 +/- 26 min) as compared to the physician's interpretation (54 +/- 18 min, p < 0.01). The endoscopy nurse marked 236 thumbnails, whereas the doctor only marked 132 thumbnails. The nurse overlooked 4 of 64 relevant lesions (6 %), which had been detected by the physician. These overlooked lesions were not single important lesions, they were overlooked only in patients with multiple angiectasias of the small intestine, and thus the misdiagnosis was without clinical relevance. The physician overlooked 6 of 68 lesions detected by the nurse (9 %), also in patients with multiple angiodysplastic lesions and therefore without clinical relevance. On post-hoc analysis of the capsule video recordings the time needed by the physician to interpret the thumbnails marked by the nurse was 10 +/- 12 min. While there was no difference with respect to the estimated gastric emptying time (nurse 27 +/- 13 min vs. physician 28 +/- 14 min, n. s.), the estimated time of capsule passage through the ileocaecal valve was longer when interpreted by the endoscopy nurse (nurse 347 +/- 89 min vs. physician 326 +/- 74 min, n. s.). Nevertheless, the total cost for capsule pre-evaluation by the nurse was lower (13.23 euro vs. physician 17.82 euro).
CONCLUSION
The endoscopy nurse detected 94 % of the significant lesions seen by the physician and no clinically relevant findings were overlooked. A pre-evaluation of the capsule video by trained staff is an accurate method and might be time effective.