Background:
‘Incidental’ BE, defined as BE diagnosed during exams for reasons other than BE screening or surveillance, has unknown incidence and outcomes. Patient characteristics, detection of dysplasia, and adherence to established quality indicators in this patient population has not been described.
Aims:
Using a national benchmarking clinical registry with matched endoscopy and pathology data, to compare (i) proportion of BE patients diagnosed with dysplasia and (iii) adherence to quality indicators – sampling using the Seattle biopsy protocol and surveillance endoscopy recommendations in non-dysplastic BE (NDBE) based on indication.
Methods
We analyzed data from the GI Quality Improvement Consortium (GIQuIC) Registry, a national repository of endoscopy quality measures. Procedure indication, demographics, endoscopy findings, pathology results, and recommendations for further endoscopy, were assessed from 1/2015 – 7/2022. Patients with endoscopic and histological findings of BE were included and were categorized as: (i) BE screening, (ii) BE surveillance and (iii) ‘incidental' BE’. Outcomes included dysplasia detection rate (DDR, defined as any dysplasia), adherence to Seattle biopsy protocol and 3-5 year surveillance interval in NDBE patients based on indication. Adherence to Seattle protocol was defined by dividing BE length by number of pathology jars submitted, with a ratio of ≤2.0 considered adherent. Patient, procedure characteristics and study outcomes were compared across groups using one and two-way ANOVA and Chi-square tests.
Results
Among 3,046,072 EGDs assessed, 173,010 (5.7%) met inclusion criteria [BE screening 42,985 (25%), BE surveillance 91,435 (53%), ‘incidental’ BE 38,590 (22%)] (Table 1). Mean BE length was 2.3 (2.5) cm and distribution based on histology was NDBE 88.1%, low-grade dysplasia (LGD) 2.1%, indefinite for dysplasia (IND) 2.8%, high-grade dysplasia (HGD) 1.1%, and unknown 5.9%. Table 2 highlights study outcomes in the overall cohort and stratified by indication. Overall, DDR was 3.1% and detection of LGD/HGD was 1.6%. Nearly 20% were not adherent to Seattle biopsy protocol and 33% were not adherent to appropriate surveillance intervals. ‘Incidental’ BE patients were similar in number to the screening group, and significantly more likely to demonstrate dysplasia compared to the screening group despite having lower adherence to the Seattle biopsy protocol (p<0.01). Nearly 50% of incidental BE were not adherent to appropriate surveillance intervals.
Discussion
‘Incidental’ BE is associated with higher rates of DDR than BE found on screening exams. These results highlight the importance of a high-quality endoscopic exam in BE patients regardless of procedure indication. Despite being enriched for dysplasia, ‘incidental’ BE exams had lower adherence rates to BE quality indicators when compared to screening/surveillance exams.

Jeffrey Williams

Appearances