The impact of experience on polyp detection and sizing accuracy at capsule endoscopy: implications for training from an animal model study.

BACKGROUND AND STUDY AIMS: Indications for capsule endoscopy include polyp surveillance in Peutz-Jeghers syndrome and potentially colorectal examination and bowel cancer screening. The ability to detect and accurately size polyps associated with these conditions is critical when deciding which patients require further investigation or therapy. Inexperience may affect the ability of capsule endoscopists to perform these skills resulting in sub-optimal patient care. We assessed the performance of volunteers with different levels of endoscopy and capsule endoscopy experience using an animal-based polyp model.

PATIENTS AND METHODS: Thirty-six artificial polyps were sewn onto the luminal surface of porcine small bowel through which a capsule was propelled. Video images generated by the capsule were then shown to five expert capsule endoscopists, seven consultant gastroenterologists, seven trainee gastroenterologists, six endoscopy nurses, six gastrointestinal radiologists, and six nonmedical participants. Polyp detection rates and sizing accuracy were calculated according to each group and compared.

RESULTS: Expert capsule endoscopists had the highest polyp detection rate (91%) but only noncapsule endoscopy experience was independently correlated with improved polyp detection and sizing accuracy (OR missing polyp--endoscopy experience 1.0, no experience 1.96 [95% CI 1.29 - 2.97], P = 0.002; ratio of estimated to actual size--endoscopy experience 1.0, no experience 1.43 [95% CI 1.22 - 1.66], P < 0.001). Both expert capsule endoscopists and experienced endoscopists tended to underestimate polyp size more than novices, particularly for large polyps (OR capsule expert to non-expert: 2.39 vs. 1.0 [95% CI 1.73 - 3.29], P < 0.001).

CONCLUSIONS: Polyp detection rates and sizing accuracy during capsule endoscopy improve with endoscopic experience. However large polyps, which are the most clinically relevant, are least-accurately sized, and capsule endoscopy experts and experienced endoscopists are the most likely to underestimate the size of these polyps. Training to improve performance in these measures for capsule endoscopy novices as well as experts is required.