Interobserver Agreement Between Experienced Endoscopist and Endoscopy Nurse in the Detection of Lesions in the Capsule Endoscopy Examination

Abstract

Aim: to evaluate the agreement between an experienced endoscopist and an endoscopy nurse in the detection of lesions in the capsule endoscopy (CE) examination. Patients and methods: we conducted a prospective study in which 32 consecutive CE examinations performed in a 6 month period with the Given™ M2A plus CE were included. Examinations were visualized first by a nurse and then by an experienced endoscopist blinded to the findings reported previously. Time of visualization of the recorded images and positive findings reported were compared between the two observers. Results: time needed for the visualization of the studies by the nurse was significantly longer (138 ± 6 vs 92 ± 9 min, P < 0.05). The endoscopist observed 11 significant lesions (5 small bowel stenosis and 6 haemorragic lesions), all of them were also visualized by the nurse, who also reported another significant lesion not detected by the endoscopist, and which was not confirmed in a subsequent review of the recorded images. Interobserver agreement was excellent (kappa coefficient 0.93). Conclusions: initial evaluation of CE examinations by an endoscopy nurse appears to be reliable and could be useful in terms of indicating an early review of the images by a physician, specially in cases needed of an urgent management such as active bleeding. Aim: to evaluate the agreement between an experienced endoscopist and an endoscopy nurse in the detection of lesions in the capsule endoscopy (CE) examination. Patients and methods: we conducted a prospective study in which 32 consecutive CE examinations performed in a 6 month period with the Given™ M2A plus CE were included. Examinations were visualized first by a nurse and then by an experienced endoscopist blinded to the findings reported previously. Time of visualization of the recorded images and positive findings reported were compared between the two observers. Results: time needed for the visualization of the studies by the nurse was significantly longer (138 ± 6 vs 92 ± 9 min, P < 0.05). The endoscopist observed 11 significant lesions (5 small bowel stenosis and 6 haemorragic lesions), all of them were also visualized by the nurse, who also reported another significant lesion not detected by the endoscopist, and which was not confirmed in a subsequent review of the recorded images. Interobserver agreement was excellent (kappa coefficient 0.93). Conclusions: initial evaluation of CE examinations by an endoscopy nurse appears to be reliable and could be useful in terms of indicating an early review of the images by a physician, specially in cases needed of an urgent management such as active bleeding.