P16 Transient elastography is underutilised in London hospitals and results in unnecessary endoscopic screening for oesophageal varices: a multi-site audit of practice

IntroductionThe Baveno VI consensus provides guidance on using non-invasive methods to identify patients with compensated advanced chronic liver disease (cACLD) who are unlikely to have clinically significant portal hypertension (CSPH) Patients with a platelet count of 150,000/Litre and a liver stiffness of 20kPa, assessed using transient elastography (TE), have a sufficiently low risk of variceal bleeding that they do not require variceal screening endoscopy to examine for oesophageal varices (OV) costing approximately £342 per procedure This identifies potential substantial cost savings to healthcare systems and reduces risk to patients from unnecessary investigations However, concordance with these guidelines, availability of TE and number of avoidable endoscopies is unknown MethodRetrospective data collection from 10 sites across London, 6 teaching hospitals and 4 district general hospitals (DGH), over a 6 month period from 1st January to 30th June 2019 by reviewing oesophagogastroduodenoscopy (OGD) requests and analysing those with indications of ‘variceal screening’, ‘cirrhosis’, ‘liver disease’ or ‘variceal surveillance’ Patient platelet count and TE result within a year of OGD was recorded ResultsData was collected for 353 endoscopies, 7 were excluded due to incomplete data and 89 due to decompensation at the time of endoscopy 141 screening procedures were included Endoscopic findings included: 74 5% no OV, 16 3% grade I OV and 9 2% ≥grade II OV or high risk stigmata 49 7% did not have a recent TE (48 5% in teaching hospitals vs 52 4% in DGH) Of those who did have a recent TE result, 54 (76 1%) met the Baveno criteria for absence of CSPH, of whom 5 (9 3%) were found to have clinically significant varices Median follow-up was 350 5 days and 0 of these patients subsequently bled The performance of the Baveno criteria in this study was: sensitivity 64 3%, specificity 85 9%, positive predictive value 52 9% and negative predictive value 90 7% Avoiding OGD in patients meeting Baveno criteria in this cohort would have potentially saved over £18000 DiscussionOur study shows that TE is not widely used for risk stratifying patients with cACLD across London prior to screening OGD These simple non-invasive markers can achieve substantial cost savings, avoid exposing patients to unnecessary investigations and relieve pressure on endoscopy departments under increased strain due to the Coronavirus pandemic Whilst a small proportion of OV will be missed, the bleeding risk in these is low with adequate follow-up Availability and utilisation of TE for risk stratification in cACLD should be improved