Abstract
IntroductionBarrett’s oesophagus (BO) neoplasia surveillance guidelines can be poorly adhered to with endoscopists routinely under-biopsying longer segments. Dedicated endoscopy services run by those trained in BO may offer better outcomes.MethodsThis is a 5-year retrospective audit of a dedicated BO service at Wrightington Wigan and Leigh NHS trust compared with endoscopies performed on non-dedicated slots. Statistical significance between groups was calculated with Chi squared on SPSS.Results1023 Barrett’s surveillance procedures were included (733 patients). 658 (64%) endoscopies were on a dedicated BO list vs 365(36%) on non-dedicated. There was no difference in male to female ratio (p=0.235), median age (67 years (26-91 dedicated) vs 67years (32-91)) or median Prague M length (3cms (0-17dedicated) vs 3cms (0-15)).Dysplasia detection rate (DDR) was significantly higher in the dedicated cohort 9% (59/651) vs 3.7% (13/351) p=0.002. Significance was sustained when cases with known prior dysplasia were excluded: 6.1% 38/622 versus 3.2% 11/347 (p=0.045). Narrow band imaging (NBI) (p=<0.001) and acetic acid (p=<0.001) were more frequently used on the dedicated list. Targeted biopsies were more frequently obtained on the dedicated cohort (p=<0.001). Guideline adherence was significantly better with dedicated endoscopy (table 1).Across all cases factors associated with improved DDR included visible lesion documentation (p=<0.001), use of targeted biopsies (p=<0.001), acetic acid use (DDR 14.4% vs 5.6% p=<0.001) and use of narrow band imaging (DDR 10% vs 3% p=0.001). There was no association with documented Seattle protocol adherence (p=0.074)ConclusionsA dedicated Barrett’s service shows improved dysplasia detection and guideline adherence. As these data show lesion recognition, acetic acid and NBI use may improve outcomes, it is imperative surveillance is performed by adequately trained clinicians, skilled in neoplasia detection, across the tertiary and non-tertiary setting.Abstract P193 Table 1 BSG standard Dedicated (n=658) Non-dedicated (n=365) P Value Prague Classification documented98% (n=641/657) 87%(n=317/364) P=<0.001 Barrett’s island description 64% (n=417/656) 10% (n=35/363) P=<0.001 Hiatus hernia delineation 86% (n=565/655)57%(n=205/362) P=<0.001 Visible lesions documented 88% (n=577/656) 24% (n=86/363) P=<0.001 Visible lesions described with Paris classification 23% (n=57/252) 2% (n=2/128) P=<0.001 Seattle protocol adherence documented 80% (n=516/639) 31% (n=105/344) P=<0.001 Surveillance interval appropriate 93% (n=517/554) 67% (n=166/249) P=<0.001