Introduction The mortality from alcohol related liver disease in the UK has more than doubled in the last decade. Services to assist people with alcohol related health problems are available but identification of those at risk of avoidable mortality remains poor.

Aim To establish the demographic details of people dying of alcohol related liver disease in a health community, determine trends and demographic associations with ALD mortality and ascertain secondary healthcare systems attendances (and reasons for them) prior to death. These are settings where alcohol interventions may be offered.

Method Identification of ALD related deaths was established via Public Health Mortality Data (ONS) and Hospital Episode Statistics for in-patient and Emergency Department attendances. Deprivation (Index of Multiple Deprivation, IMD) and population mosaic data are available for the local communities studied.

Results 755_people died with a primary or underlying diagnosis of ALD from 2007 to 2010. Mortality rates overall are stable over this time period. Age at death was 20–34 in 2.4%, 35–39 in 24%, 50–64 in 39%, 65–79 in 25% and 80+ in 10%. Standardised mortality rates (DSR) varied by geographical area from 18 (CI 13 to 23) to 46 (CI 40 to 52) and there was_ a strong correlation with social deprivation (R²=0.83) but people dying of ALD had higher deprivation scores that their local population across all geographical areas (Mean IMD score general population 21.8 vs ALD 25.9, p=0.001). Mosaic groups at highest risk of ALD death varied by geographical area, in the highest mortality area, young, well-educated city dwellers and families in low-rise social housing with high levels of benefit need have the highest number of deaths but in these groups the mortality from ALD was relatively under-represented in comparison to the proportion of population in those groups (44% of population in those groups, 32% of ALD deaths) whereas the ALD deaths were over-represented in other mosaic groups (Elderly people reliant on state support, Young people renting flats in high density social housing and Residents with sufficient incomes in right-to-buy social houses, 40.76% of ALD deaths but 19.7% of population). 72% of deaths from ALD were in hospital, 24% at home with the remainder being in another community setting. Hospice care was rare (0.13%). 95% of people who died had a previous hospital admission prior to death but only 42% had a previous hospital admission with a diagnosis of ALD. 80% of ALD deaths had a prior Emergency department attendance. In those people with a prior hospital admission 32% had a discharge diagnosis of ALD suggesting that already advanced liver disease is not diagnosed or coded in 68% of hospital attendees in other specialty areas. The admissions were in a wide variety of specialty groups the most common being cardiac disease (7.9%), neoplasia (6.4%), injury/poisoning (6.3%) and mental and behavioural disorders (5.6%). There was a mean of 7.06 hospital admissions per death from ALD in the 5 years preceding death.

Conclusion Death from ALD is preceded by a substantial number of medical attendances in secondary care with the opportunity for intervention. Liver disease is unsuspected in a significant number of people in contact with secondary care. Mortality rates over a 4-year period remain high and cluster in particular demographic groups which has importance for public health interventions.