PTH-104 Hospital Admissions and Associated Costs of Alcoholic Liver Disease in Scotland Between 1991 and 2011

Abstract

Introduction Alcoholic liver disease and its associated hospitalisations due to acute and chronic liver disease is a significant worldwide problem with Scotland having worse statistics than most European and Western countries with regards to health burden including mortality and healthcare costs. This SC ottish A lcoholic L iver disease E valuation (SCALE) study was set up to identify a cohort of people hospitalised for the first time with Alcoholic Liver Disease (ALD) in Scotland from 1991 to 2011 and identify trends in these incident admissions, in-hospital mortality, readmissions and mortality after discharge as well as life time costs compared to controls. Methods Hospital records from 1981 to 2011 were used to identify cases of ALD presenting 1991–2011 with no prior hospitalisation due to ALD. These incident cases were classified by sex, age group, socio-economic deprivation and study year as well as decompensated or not decompensated ALD. In-hospital mortality, time to all cause death, readmissions and related hospital costs were calculated by length of stay and hospital-specific per diem cost and compared to matched controls. Parametric survival analysis was used to estimate remaining life expectancy and remaining lifetime hospitalisation related costs were calculated. Results We identified 35,208 incident ALD admissions over the period 1991 to 2011. Incident ALD hospitalisation rates decreased between 2002 and 2011(Men: 63 to 54; Women: 27 to 24 per 100,000). Inpatient mortality was 17.4% during the index admission. 5 (10) year mortality post-discharge was 59% (74%) for patients with decompensated ALD and 49% (65%) without. The mean (median) number of readmissions per year was 3.1 (1.3). The estimated annual cost of hospitalisation for the remaining life of a man, aged 50, living in the most deprived SIMD fifth with an incident ALD admission in 2011 was £ 118,000 higher compared to matched controls. Conclusion After a peak in 2007 / 2008 incident rates of ALD fell in Scotland but remain high in comparison to the epidemiological literature. Prognosis for incident cases of ALD is poor and costs associated with incident ALD cases are very high compared to matched controls.