Nearly 20 years ago, public attention was drawn to the previously hidden tragedy of deaths in prison and police custody. Initially, attention was focused on deaths of Indigenous people. However, it quickly became apparent that the death rate for all people held in custody was much higher than that for the general population. The result was a detailed investigation into the causes of the problem, in the form of the Royal Commission into Aboriginal Deaths in Custody.1 The Royal Commission made over 300 recommendations on penal policy, cell design, custody management regimens, treatment programs, services to Indigenous offenders and a host of other topics. In addition, the Deaths in Custody Monitoring Program was established at the Australian Institute of Criminology (AIC) to scrutinise and report on deaths in prison or police custody.2

In parallel with the acknowledged problem of deaths in custody, there is an equally significant tragedy in the form of high death rates among released prisoners3-5 and offenders in the community.6,7 In the decade after the Royal Commission (1990 to 1999), the AIC monitoring program recorded 628 deaths in police or prison custody. Over the same period, in Victoria alone, 820 men and women who had been released from prison died unnatural deaths.4 The study by Coffey and colleagues in this issue of the Journal (page 473) shows that the problem is not confined to adult offenders but affects juveniles as well.8

The high rate of unnatural deaths among offenders living in the community is a major public health issue, but what can we do to reduce these rates? One of the groups at greatest risk is injecting drug users — over half of the unnatural deaths examined in the earlier Victorian study were heroin related,4 and drug-related offences were an indicator of high mortality risk in the study by Coffey et al.8 Drug treatment and maintenance (methadone) programs lower the risk of death by overdose.9 However, many of those at greatest risk are profoundly alienated from society, and we need to find ways to engage them. For example, heroin-dependent Indo-Chinese offenders frequently face rejection by their families and community, and are isolated from the mainstream community as well. Female offenders often come from backgrounds of extensive sexual and physical abuse, their heroin dependency is often supported by prostitution,5 and any interventions need to take account of their responsibilities as parents.10 Treatment and maintenance programs need to be delivered in ways that meet the material, social and cultural needs of those at risk.

Overdose risk can be dramatically lowered by behaviour changes, like not using drugs by yourself, being aware of variations in the purity of heroin, and not taking heroin in conjunction with alcohol or benzodiazepines. Again, the problem is partly that those at greatest risk are also the most difficult to communicate with, and tend to be unrealistic in judging risks to themselves. Peer-based education and information dissemination programs have shown they can transmit the key messages about risk reduction to this group (eg, how to avoid overdose and recognise its signs),9 but, again, a range of approaches tailored to the needs of specific groups at risk is required. Such approaches could include maximising the effectiveness of needle and syringe program workers by having them provide standardised, evidence-based messages and materials on safe injecting, overdose prevention and support options.9

The importance of heroin as a cause of unnatural death should not obscure the other dangers that offenders face. Individuals who have the combination of mental disorder and drug and alcohol misuse experience much higher risks of both overdose and suicide and can find it difficult to obtain the kind of treatment and support services that might alleviate their problems. Older offenders are at increased risk of a variety of general health problems, such as diabetes, cancer and liver disease, and there need to be programs that link at-risk offenders with healthcare and support services.11

People who are released from custody are at greatly increased risk, in part because their tolerance for heroin is reduced, and also because return to the community can be a time of great emotional stress. Prisoner release support programs like the Victorian Bridging the Gap program have shown that the period before release can be an important “window of opportunity” when offenders are motivated to plan for their release.12 A key feature of this program is intensive, outreach-based support, with the support agency helping the releasee to identify his or her specific needs and brokering access to material support, healthcare and social services. Releasees who participated in Bridging the Gap had improved outcomes as measured by accommodation stability and engagement in drug-treatment programs, and these in turn translated into lower rates of reoffending.12

Finally, we need to attend to an important lesson from the Royal Commission. Despite real improvements in custodial management, the number of deaths of Indigenous people in custody has continued to increase because there are now more Indigenous people in custody.2 The high rate of unnatural deaths among offenders is a public health problem that requires changes in the way that healthcare services are delivered to this vulnerable population. However, we also need to recognise that the justice system has a key role to play in ensuring that its goals of punishing offenders and preventing crime are properly balanced by a consideration of the health and support needs of the people who are the subjects of its interventions.