P5 The use of the Bristol prognostic score in predicting 12-month mortality in patients with advanced chronic liver disease

<h3>Background</h3> Liver disease in the UK is the third commonest cause of premature death. Patients with advanced chronic liver disease (ACLD) have reduced life-expectancy and many are not candidates for curative options like transplantation. Key barriers to palliative care (PC) delivery in ACLD is the unpredictability of disease trajectory. The Bristol Prognostic Score (BPS) was developed to identify ACLD patients at risk of 12-month mortality; a score of &gt;2 was shown to predict 12-month mortality.¹ <h3>Aims</h3> The aim of this study was to evaluate the effectiveness of the BPS in predicting 12-month mortality in a cohort of patients admitted with ACLD and determine its use for timely PC referrals. <h3>Methods</h3> Retrospective data were collected for patients with ACLD at an NHS trust in South-East England between May to October, in 2021 and 2022. An ACLD admission was defined as inpatient episodes relating to decompensated liver disease. Data included demographics, traditional prognostic scores (Child-Pugh score [CPS], MELD, UKELD) and clinical outcomes. Survival analysis was conducted using Kaplan-Meier estimates to evaluate if BPS&gt;2 predicted 12-month survival and logistic regressions were performed to assess independent predictors of 12-month mortality. <h3>Results</h3> 98 patients met our inclusion criteria; and n=53 (56.4%) died within 12-months. The BPS&gt;2 predicting 12-month mortality had sensitivity=67.9%, specificity=58.5%, NPV=58.5% and PPV=67.9%. Kaplan-Meier analysis showed that individuals with BPS&gt;2 had a significantly lower median survival compared to those with BPS of 2 or less, (8 vs 25 months respectively, p=.0057). Multivariate analysis, however, found that only CPS and age were significant predictors of 12-month mortality (OR=1.61, p=1.05x10–2, CI=1.13–2.38 and OR=1.09, p=9.98x10–4, CI=1.04–1.16, respectively). Of the 53 individuals who died within 12-months, n=24 (45.3%) weren’t referred to PC. Independent predictors of referral to PC were age (OR = 1.06, p = 0.014, CI = 1.01 – 1.1) and WHO performance status (OR=3.38, p=0.012, CI=1.36–9.44.) BPS&gt;2 was not associated with increased rates of PC referrals (OR=0.79, p=0.716, CI=0.02–2.77). <h3>Conclusions</h3> While a BPS&gt;2 was associated with worse median survival; sensitivity and specificity of the BPS&gt;2 for predicting 12-month mortality was lower compared to the original study.² Accurately predicting mortality in ACLD therefore remains a challenge. Our data suggests that the BPS might have use in risk-stratifying patients. However, integration with other prognostic scores such as the CPS might be needed to better predict short-term mortality, thereby assisting clinicians in making timely PC referrals. <h3>References</h3> Bowers S, Clare K, Hagerty L, McColl K, Smith EH, Brown-Kerr A, <i>et al.</i> Predicting 1-year mortality among patients with decompensated cirrhosis: results of a multicentre evaluation of the Bristol Prognostic Score. 2022 Mar 1 [cited 2023 Jul 28];9(1):e000822–2. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8943768/ Hudson BE, Ameneshoa K, Gopfert A, Goddard R, Forbes K, Verne J, <i>et al.</i> Integration of palliative and supportive care in the management of advanced liver disease: development and evaluation of a prognostic screening tool and supportive care intervention. <i>Frontline Gastroenterology</i> [Internet]. 2016 Oct 5;<b>8</b>(1):45–52. Available from: https://doi.org/10.1136%2Fflgastro-2016-100734.