A good deal of research has been undertaken into non-fatal heroin overdose in the past decade. Early investigations identified its high prevalence among injecting drug users, the protective benefit of being in treatment and heroin users' perceptions of overdose risk (Darke et al. 1996; McGregor et al. 1998; Powis et al. 1999). These earlier findings suggested the need to educate users about the likelihood of overdosing in an effort to alter their risk perceptions and hopefully their drug using behaviours.
As the clinical issues of dual diagnosis and suicide became increasingly strategic priorities in mental health services in Britain (Appleby et al. 2001; Department of Health 2002), the focus of research shifted from the circumstances of non-fatal overdose to the psychological correlates or causes of overdose. To what extent were accidental drug overdoses among heroin users actually suicide attempts or markers of poor psychological functioning in general was the question a number of studies sought to answer. Suicide and accidental drug overdose seemed less a dichotomy and more a spectrum of intention, ranging from suicide through attempted suicide, cries for help and high-risk behaviour with its attendant indifference to risk of death, to an entirely accidental and unintended overdose (Farrell et al. 1996; Rossow & Lauritzen 1999; Neale 2000). Nevertheless, the need to undertake careful mental health assessment of patients in drug treatment and to refer to appropriate psychiatric, psychology and other services emerged as an additional recommendation to reduce the incidence of drug overdose in this population (Darke et al. 2004).
Now Neale & Robertson 2004) have reported that psychosocial or ‘life’ problems were linked with likelihood of having a non-fatal overdose among a sample of heroin users before entering drug treatment in Scotland. Specifically, they found that accommodation problems, recent death of a loved or close person and relationship breakdown were associated independently with having a non-fatal overdose in the 90 days prior to entering drug treatment. Due to the cross-sectional design of their study they noted, critically, that their data did not enable them to establish any causal relationship between these life problems and overdose.
Life problems, of course, are also prime reasons why heroin users present to treatment services (Weatherburn & Lind 2001; Sell & Zador 2004), so it is not surprising to find a high proportion of subjects entering treatment in the present study reporting a variety of recent life problems. Another study, comparing prevalence of social problems among a sample of mental health patients subdivided into groups with and without problems of dual diagnosis with a sample of substance misuse patients with and without dual diagnosis, suggested that the association between substance misuse and social problems was greater than that between mental health and social problems (Todd et al. 2004).
Building on from these earlier studies, Neale & Robertson (2004) have concluded that drug users' social or life problems warrant focused targeting if the incidence of overdose is to be reduced. They go on to recommend that that burden should fall to drug treatment services. However, all agencies in contact with illicit drug users—homeless agencies, social benefits branches, social care services and others—need to be informed of the importance of addressing social risk factors as a possibly effective measure in its own right in preventing overdose. The findings of Neale & Robertson 2004) deserve to be disseminated widely.
Given the increasing attention to social risk factors among this population and now the finding of its independent link with overdose, reducing drug overdose is a remit for more than drug treatment agencies—it is an all-of-society problem requiring an all-of-society response.