The increasing mortality burden of liver disease among opioid-dependent people: cohort study

Hepatitis C (HCV) infection is highly prevalent among injection drug users (IDUs) and likely to cause significant mortality over time, but little research attention has focused upon the magnitude of this risk, particularly among ageing users. This study examined trends over time in mortality attributed to liver disease, and in particular contrasting this with other more commonly studied causes of death [acquired immune deficiency syndrome (AIDS), suicide and overdose] among an ageing cohort of heroin-dependent people in Australia.

Data linkage study of methadone treatment entrants with the National Deaths Index.

A cohort entering methadone treatment for heroin dependence in New South Wales, Australia, 1980-85.

A total of 2489 people entering methadone treatment for heroin dependence and 54,847 person-years (PY) of follow-up.

Linkage of data on all methadone entrants between 1980 and 1985 with data from the Australian National Deaths Index, linked using probabilistic record linkage software.

There were 8.2 deaths per 1000 PY [95% confidence interval (CI) 7.5-9.0], with standardized mortality ratios (SMRs) of 4.6 (95% CI 4.2-5.0). Almost one in five (17%) of deaths were from underlying liver-related causes, most commonly viral hepatitis. The overall mortality rate for any liver cause was 1.4 deaths per 1000 PY (95% CI 1.1-1.7), 17 times higher than to the general population (95% CI 13.4-21.3), with relative elevations more marked for females (SMR 27.9; 95% CI 17.7-41.9) than males (SMR 14.5; 95% CI 10.8-19.0). Liver mortality increased over time, becoming the most common cause of death by the end of follow-up.

Liver disease has become the most common cause of mortality among ageing opioid-dependent people in an ageing Australian cohort. There is an imperative to reduce the long-term risks of HCV and other risks to the liver, including alcohol consumption, which are typically not the major clinical focus for this group.