Place of death and factors associated with hospital death in patients who have died from liver disease in England: a national population-based study

Abstract

Liver disease is a major cause of mortality, with high numbers of hospital deaths, and disproportionately affects people younger than 65 years. This study aims to examine the place of death and factors associated with hospital death for people who died from liver disease. We did a national population-based, observational study using the National Death Registration Database from the Office for National Statistics, 2001-14. All non-accidental adult deaths (hospital and non-hospital) from liver disease in England were included. Explanatory variables were underlying cause of death, contributory causes of death (number and specific causes), age at death, sex, marital status, year of death, index of multiple deprivation, rural or urban settlement, and residential region. We applied modified Poisson regression models to assess the strength of association between hospital death and explanatory variables using an adjusted prevalence ratio (PR). 135 953 decedents were included, of whom 56 065 (41·2%) died from alcohol-related liver disease. Annual deaths from liver disease increased from 7811 in 2001, to 11 017 in 2014. Hospitals were the main place of death (66·9% [95% CI 66·6-67·1]) for patients who died from liver disease. The proportion of hospital deaths reduced from 71·5% in 2001 to 60·0% in 2014. After adjusting for sociodemographic factors, patients who died from alcohol-related liver disease had the highest chance of hospital death; people who died from liver cancer were less likely to die in hospital than people with alcohol-related liver disease (adjusted PR 0·61 [95% CI 0·60-0·61]). People with four or more contributory causes of death were more likely to die in hospital than those with no contributory causes (1·45 [1·42-1·47]). Patients with sepsis (1·24 [1·23-1·25]), hepatorenal syndrome (1·22 [1·21-1·22]), and peritonitis (1·18 [1·17-1·20]) had higher chances of hospital death than those without these respective contributory causes, and those with alcohol-related disorders (0·67 [0·66-0·69]) had lower chances of hospital death. The high risk of hospital death in patients with sepsis, hepatorenal syndrome, or peritonitis warrants further investigation, and the low chance of hospital death in patients with alcohol-related disorders also needs to be explored. Prevention strategies and end-of-life care services are urgently needed to prevent and tackle harms from liver disease. National Institute of Health Research Health Services and Delivery Research Programme, and Collaboration for Leadership in Applied Health Research and Care South London.