Years of potential life lost amongst heroin users in the Australian Treatment Outcome Study cohort, 2001-2015

June 2016
Citation: 
Darke, S., Marel, C., Mills, K. L., Ross, J., Slade, T., & Teesson, M. (2016). Years of potential life lost amongst heroin users in the Australian Treatment Outcome Study cohort, 2001-2015. Drug and Alcohol Dependence, 162, 206-210

Heroin use carries the highest burden of disease of any drug of dependence, with rates of mortality more typical of the elderly Indeed, it has been estimated that heroin users die at a rate 15 times that of the general population. Reflecting the high levels of mortality risk factors seen amongst heroin users, rates of death are elevated for all causes, including overdose, suicide, disease, trauma and homicide.

While mortality rates provide important epidemiological data, they do not measure the extent of premature mortality.

The authors examined death in the Australian Treatment Outcome Study (ATOS), a cohort 615 heroin users, recruited in 2001.  They looked at deaths over 2001-2015, causes of death, and the years of potential life lost (YPLL) associated with these deaths. At 2015, 12% of the cohort had died, a rate of around 1% per annum -- 10 times the rate expected amongst the general population. Men were 7 times more likely to die than men in the general population, and women 17 times more likely than women in the general population. Each death was associated with 28 years of life lost, taking death before age 65 as the cut-off (a conservative measure often used in disease epidemiology). Using the alternative measure years of life lost in the context of average life expectancy, there was an average of 44 years of potential life lost associated with these deaths. Accidental overdose and suicide accounted for three quarters of years of life lost. Given the prominent role of overdose and suicide, the majority of these fatalities appear preventable.

What does it mean and what can be done?

It is clear from these data that the chronicity of heroin use seen across waves is associated with a continued high risk of premature mortality. What, then, can be done? Long-term, stable treatment, such as opioid substitution or residential rehabilitation, is well demonstrated in this and other cohorts, to reduce drug use and associated harm. While cycles of relapse and abstinence appear the norm, engaging and retaining heroin users in treatment appears the most effective means to reduce risk, particularly from opioid overdose. Given the prominent role of overdose, screening for overdose risk within a treatment setting would also appear appropriate. Suicide prevention also appears to be of high importance.