"Understanding variation in health-related quality of life domains for patients with end-stage liver disease- findings from the EU COPE trial."

Introduction: Understanding the extent of patient morbidity is important for managing patient care: end-stage liver disease (ESLD) is the terminal stage of liver disease necessitating transplantation, and for which there are 11 clinical classifications with varying impacts on the patient’s symptoms and severity. Evaluating health-related quality of life (HRQoL)2 of patient groups provides insights into how specific health issues affect well-being. Given the clinical heterogeneity across the sub-group classifications, patient-reported HRQoL measures may exhibit high levels of variability. Understanding the variation is crucial to avoid misleading HRQoL reporting which may inform decision making by policy makers. The aim is to examine the extent of variation across HRQoL domains, utilising data from the EuroQol EQ-5D-5L alongside a European trial for liver transplantation. Methods: This analysis examined the responses across each of the EQ-5D-5L domains6 for HRQoL self-assessment (mobility, self-care, usual activities, pain/discomfort, anxiety/depression) using the EU COPE liver transplantation trial data for ESLD overall and by clinical sub-group. Variation in domain response at baseline and the follow-up periods, post-transplantation (30 days, 6 months) was examined. It was investigated whether other patient-level characteristics were drivers of variation usingvariation using ordinal regression as well as the probability of higher vs. lower categories being dominant. Computation was carried out using Microsoft Excel and Matlab. Results: This analysis focused on HRQoL collected alongside a small European RCT and included UK responses only. Sub-clinical groups sample sizes varied from n=2 to n=41(total sample used was n= 145). Baseline HRQoL overall exhibited the highest domain responses i.e. ‘no issues’ for Mobility (Level 1:46%), Self-Care (Level 1:66%), Usual Activities (Level 1:29%), and Depression/Anxiety (Level 1: 41%); with Pain/Discomfort (Level 2:32%) having the highest proportion of ‘mild issues’. Clinical subgroups indicated Hepatocellular Carcinoma without Cirrhosis and Non-Alcoholic Steato-Hepatitis had the lowest domain scores comparatively, while Hepatitis B and Non-Alcoholic Fatty Liver Disease had the highest. Conclusions: Clinical sub-groups report different HRQoL profiles. A “one size fits all” approach to disease management guided by inference from descriptive summaries of outcomes, neglects to understand the heterogeneity across self-reported morbidity. It is important to provide detail on the uncertainty in patient-reported HRQoL and the drivers of variation to ensure decisions made are patient-centred and reflect individual needs due to the impact of disease severity. References: [1]Philips CA, Kedarisetty CK. Palliative Care for Patients with End-Stage Liver Disease. Clinical and Experimental Hepatology. 2022 Aug [2]Yin S, Njai R. Summarizing health-related quality of life (HRQOL): development and testing of a one-factor model Population Health Metrics . 2016 Jul 11 [3]Yang MS, Lai CY. A robust EM clustering algorithm for Gaussian mixture models. Pattern Recognition. 2012 Nov [4]Herdman M, Gudex C et al. Development and preliminary testing of the new five-level version of EQ-5D. Quality of Life Research. 2011 Apr 9 Contributorship: Eugene McCarthy4, Simon Knight5, David Nasralla6, Peter Friend5, Rutger Ploeg5 on behalf of the COPE. Department of Biopharmaceutical and Medical Science, Atlantic Technological University, Galway Nuffield Department of Surgical Sciences, University of Oxford Royal Free London, NHS Foundation Trust