NDARC Monograph No. 20 (1994)


It is well documented that heroin users are at substantially greater risk of premature mortality than their non-heroin using peers. Longitudinal studies generally suggest that approximately 2-3% of heroin users die each year. The excess mortality rates among heroin users in these studies have been variously estimated to be between 6 and 20 times those expected among peers of the same age and gender. In Australia, and other countries, the major contributor to the excess mortality among this population has been narcotic overdose. In 1990, 457 opioid-related deaths occurred in Australia.

Unlike alcohol and tobacco related deaths, which have declined, the incidence of opioid related deaths in Australia appears to be rising significantly, with death rates due to opioids increasing by 170% between 1981-1990.

The magnitude of the problem of opioid overdoses in Australia is best illustrated by comparison to deaths due to HIV/AIDS among injecting drug users. To date, 170 injecting drug users have died of AIDS in Australia. As noted above, 457 people died of opioid overdose in 1990 alone. Overdose is currently a far greater public health problem among heroin users than is HIV.

Despite the mortality caused by overdose among heroin users, surprisingly little research has been conducted on this topic. Such studies that have been conducted have overwhelmingly been retrospective ones of either coronial records, or of hospital emergency admission records. These studies have identified several factors that may be related to opioid overdose deaths. One major factor is the use of alcohol. Ruttenber and Luke, for example, reported that 74% of heroin-related deaths between 1976-1979 in Washington, D.C. had alcohol present at autopsy, with a mean blood ethanol level of 0.09mg/100mls. Walsh found that 48% of opioid deaths in Newcastle, Australia had alcohol present, with a mean blood ethanol level of 0.14mg/100mls. Benzodiazepine use has also been implicated in several studies. Other factors include being female, not currently being in drug treatment, being single, longer heroin using careers and week-end use.

While such retrospective record studies are useful, they cannot provide information on the prevalence of non-fatal overdose among heroin users, or the actions taken by those present at another's overdose. To the best knowledge of the authors, only three published studies have involved interviews with heroin users about their experiences with overdose. Bammer and Sengoz interviewed heroin users in the Australian Capital Territory. The results indicated that overdose is a common experience among heroin users, with a third of the sample reporting an overdose in the preceding twelve months. Apart from the Bammer and Sengoz study, the only other published study on the prevalence of non-fatal overdose is that of Toet & Ven, cited in Grund. This study was based upon reports of non-fatal overdoses among Rotterdam methadone maintenance clients, registered in the Rotterdam Drugs Information System. The authors reported that, in 1988, 27% of clients had a lifetime prevalence of non-fatal overdose. Again, overdose would appear to be a common experience among heroin users.

Manning et al conducted psychological autopsies following 37 heroin-related deaths and near-deaths among U.S. military personnel based in Germany. The study involved interviews with friends of the victims, those present at the overdose, and the victims themselves in the 13 near-death overdoses. Questions were asked concerning the circumstances leading to overdose and the actions that were taken upon overdose. This was not, however, an epidemiological study concerning the prevalence of such experiences.

The current study was designed to investigate the prevalence of overdose and the circumstances surrounding overdose among a large sample of Sydney heroin users. It also aimed to investigate actions taken at others' overdoses; obstacles to help-seeking; the beliefs of heroin users about overdose and its causes; and their attitudes towards the opioid antagonist naloxone. The motive for the study was to develop interventions to reduce overdose and overdose deaths among injecting drug users (IDU) who continue to use heroin.



Shane Darke, Joanne Ross, Julia Cohen, Wayne Hall
Date Commenced
31 Oct 1994
Resource Type