|Institution:||Iowa State University|
|Keywords:||Canine; Endoscopy; IBD; Veterinary Medicine|
|Full text PDF:||http://lib.dr.iastate.edu/etd/13758
Measures of disease activity are necessary when diagnosing and monitoring response to therapy in canine inflammatory bowel disease (IBD). Endoscopy has been described as the gold standard to evaluate the mucosal surface and obtain biopsies for a histopathologic diagnosis. Previous veterinary studies have failed to definitively determine the benefit of endoscopic gastrointestinal mucosal evaluation in defining disease severity in dogs with IBD. The aim of the following thesis was to evaluate the inter-observer agreement between trainee and expert endoscopists in the assessment of mucosal lesions in dogs with IBD, and to evaluate if trained operators can identify and agree upon most endoscopic lesions of mucosal inflammation using the proposed simplified endoscopic scoring index. Archived images from endoscopic procedures performed in dogs diagnosed with IBD at the Iowa State Lloyd Veterinary Medical Center from 2002-2012 were reviewed. In total, 95 images of inflammatory and normal mucosa from dogs with IBD were displayed to 3 expert and 5 trainee endoscopists (initial test). Each picture was assessed independently by the endoscopist for inflammatory changes using established indices or interpreted as normal mucosa from multiple areas of the GI tract (ie. stomach, duodenum, and colon). Agreement was measured between the trainee and expert endoscopists for each organ. The developed index was then applied to a prospective independent group of dogs (23 total) diagnosed with inflammatory bowel disease for a validation study. Comparisons were made between 2 expert endoscopists (JES & AEJ) to measure mucosal assessment agreement. Regression analysis showed a significant (p<0.01) difference between expert versus trainee endoscopy scores in duodenal evaluation trial 1, although repeat duodenal lesion evaluation aided by use of a visual template improved the overall scores of trainee endoscopists to near that of expert endoscopists (p=0.06). For the validation study, the expert endoscopists had substantial to almost perfect agreement for each lesion assessed in the stomach (k>0.8), had moderate to substantial agreement assessing the small intestine (k>0.6-0.84) and substantial agreement when assessing the colon (k>0.7-1). The conclusion is that trained operators can identify and agree upon most endoscopic lesions of mucosal inflammation.