Inter‐observer and intra‐observer variability in the diagnosis of dysplasia in patients with inflammatory bowel disease: correlation of pathological and endoscopic findings

Abstract

Dysplasia is reliably diagnosed by expert gastrointestinal pathologists but has poor correlation with an endoscopic diagnosis of dysplasia.

Low interobserver agreement among endoscopists in differentiating dysplastic from non-dysplastic lesions during inflammatory bowel disease colitis surveillance

“…The accuracy of the overall differentiation described in our study (overall 62%, expert 68%, and nonexpert 58%) was similar to the accuracy observed in the study from Farraye where standard white light endoscopy was used (overall 65%, expert 75%, and nonexpert 56%) [25]. Although comparing the results of two image-based studies is not the optimal study approach, this suggests that the additional value of pCE in differentiating lesions encountered during IBD surveillance is smaller than previously perceived.…”

“…A recent study by Allende and colleagues looked at the correlation between endoscopic diagnosis and pathology and showed a poor correlation. In over half of the cases, dysplasia was seen histologically in which the endoscopic diagnosis was negative [25]. Although this study was performed from a pathological perspective, the poor correlation between endoscopy and pathology is comparable to our results and reflects the difficulties in making a diagnosis of lesions found during IBD surveillance colonoscopy solely based on the endoscopic image [26].…”

DNA flow cytometric and interobserver study of crypt cell atypia in inflammatory bowel disease

“…The lack of agreement in the grading of CCA was expected, as significant interobserver variability, particularly at the lower end of the dysplasia spectrum (NEG versus IND versus LGD), has been reported in IBD . Although the earlier study by Riddell et al .…”

“…suggested that good to modest interobserver agreement among pathologists could be achieved for the diagnosis of IBD‐related dysplasia, others have demonstrated that interobserver agreement is only fair for HGD (kappa = 0.54–0.61) and poor for LGD (kappa = 0.18–0.23) among both general and GI pathologists . Similarly, a high level of disagreement in the diagnosis of IND has been reported (kappa = 0.01–0.25) . Recognising that it would be difficult to establish objective criteria to diagnose and/or grade CCA in a consistent and reliable manner based on morphology alone, we believe that additional criteria should be sought.…”

Clinical Aspects of Idiopathic Inflammatory Bowel Disease: A Review for Pathologists

“…This system consists of 4 categories-negative for dysplasia, IND, LGD, and high-grade dysplasia-with an overall fair and modest interobserver agreement for negative for dysplasia and high-grade dysplasia, and poorer agreement (higher variability) for LGD and IND. [97][98][99] The IND category is used when the epithelial changes cannot be classified as either negative or positive for dysplasia with certainty. For example, IND may be rendered when coexisting florid acute inflammation, lack of the surface epithelium for evaluation, or processing/technical issues preclude definitive assessment of the epithelial alterations.…”