Estimating inequality in alcohol-related liver disease burden in the UK, 2009 to 2020: a population-based study using routinely collected data.

Background: There is a need to understand the preventable burden of alcohol-related liver disease (ARLD) and to improve the identification of individuals at high risk. We aimed to establish reliable and stratified epidemiological data to understand the burden of ARLD and the inequalities in this burden related to ethnicity, socioeconomic factors, and region in the UK.

Methods: Data were extracted from Clinical Practice Research Datalink Aurum, a primary care database that includes 20% of UK general practices. The study period was Jan 1, 2009, to Dec 31, 2020; all patients aged 18 years and older registered at a participating practice were eligible for inclusion. Hospital admission data were extracted from linked Hospital Episode Statistics (HES) and ARLD-specific mortality data were obtained from Office for National Statistics Death Registration Data. Several analytical approaches were used, as follows: yearly cross-sectional and cohort analyses to calculate the annual prevalence and incidence of ARLD, respectively; a retrospective, matched, open cohort study to assess all-cause mortality rates (in which patients without liver disease were matched with patients with ARLD on the basis of age, sex, ethnicity, and geographical region); and a retrospective, open cohort analysis to evaluate all-cause hospitalisation rates. Hospitalisation rates were calculated in those with ARLD only. We explored different definitions of ARLD, and our primary definition was definite ARLD (ie, a coded clinical record specifying ARLD). Incidence and prevalence were stratified by age, sex, ethnicity, deprivation (Index of Multiple Deprivation [IMD] quintile) and geographical region.

Findings: During the study period, 19 534 887 patients from 1491 practices were eligible for inclusion in our study. For definite ARLD exposure, 257 544 patients were included in the all-cause mortality outcome analysis, of whom 51 510 were diagnosed with definite ARLD; while among the 50 409 patients with definite ARLD for whom HES-linked data were available, 37 142 had one or more hospital admissions. Prevalence of definite ARLD rose from 154 to 243 per 100 000 population from 2009 to 2020. Incidence increased from 18·6 to 30·3 per 100 000 person-years between 2009 and 2019, and then decreased to 24·7 per 100 000 person-years in 2020. Prevalence and incidence of ARLD by age, sex, ethnicity, geographical region, and IMD quintile increased between 2009 and 2020. The overall adjusted all-cause mortality hazard ratio (HR) for those with definite ARLD compared with no liver disease was 4·30 (95% CI 4·20-4·41). The effect of ARLD on mortality was more pronounced in younger than older age groups (eg, adjusted HR of 21·86 [95% CI 18·23-26·20]) in those aged 30-39 years vs 2·19 [2·09-2·29] in those ≥70 years) and in females than in males (5·61 [5·35-5·88] vs 3·93 [3·83-4·04]). The overall incidence rate for hospitalisations in patients with definite ARLD was 1·17 per person-year. Hospitalisation rates were higher in females (adjusted incidence rate ratio 1·03 [95% CI 1·01-1·06]) and in patients in more deprived groups (1·16 [1·10-1·21] in the most deprived IMD quintile vs the least deprived quintile).

Interpretation: Our findings indicate an increasing burden of ARLD in the UK. Raising awareness of disparities in health outcomes in affected groups could facilitate earlier and more targeted interventions.

Funding: National Institute for Health and Care Research Clinical Research Network West Midlands.