P329 Measuring the burden and impact of comorbidity in cirrhosis: insights from clinical database analysis

Abstract

<h3>Introduction</h3> Cirrhosis is a serious long-term condition, with high rates of morbidity and mortality. Effectively managing this long-term condition requires consideration of interlinked factors such as comorbidity, which can influence treatment decisions, the appropriateness of cancer surveillance, and decision making during acute illness. This study aims to determine the burden of comorbidity within our cirrhosis population. <h3>Methods</h3> The cirrhosis database in our centre, comprising over 800 individuals, was searched for comorbidities using the widely recognised Charlson comorbidity index (CCI). CCI was calculated from hospital discharge records using an ICD-10 coding algorithm. To mitigate the risk of missing comorbidities not documented at discharge, the CCI was corroborated through a secondary search of Primary Care (PC) records. Only those with both ICD-10 and complete PC data were included. <h3>Results</h3> 331 patients with cirrhosis were included (female 42%; n = 138) with<b> a</b> median age of 61 years (range 30 - 91). Aetiology was predominantly alcohol-related (ArLD) (42.6%; n = 141) and metabolic dysfunction associated steatotic liver disease (MASLD) (18.1%; n=60), among other causes. These characteristics were representative of the entire database. Median CCI score was significantly higher by PC record (5) than ICD-10 (2) (difference 2.8; 95% CI 2.56 - 2.97; p &lt; .05), indicating that ICD-10 under-represents the degree of comorbidity. Consequently, PC records were used for further analysis. The most prevalent comorbidity was diabetes mellitus (n = 119; 36%), followed by COPD (n = 44; 13.3%). MASLD and diabetes were closely linked, with substantial overlap; 39.8% of individuals with diabetes had MASLD, and 78.3% of those with MASLD had diabetes. Almost two thirds of the cohort had five or more comorbidities (CCI ≥5; 57.7%), and those with MASLD had the greatest proportion of severe comorbidity (79.3%). 37 patients in the cohort died during follow up. Whilst there was no significant association between liver aetiology and death (p = 0.124), CCI grade was significantly associated with mortality (p = 0.048). <h3>Conclusions</h3> We found high levels of comorbidity, particularly diabetes, in our cirrhosis population. The greatest burden was observed in those with MASLD. CCI grade was associated with increased mortality, and recording this will inform both individual management plans and development of care pathways. Relying solely on ICD-10 codes underestimates the true burden of comorbidity, and linked databases are recommended for future analysis to provide a more comprehensive understanding of the association between cirrhosis and comorbid disease.