Polydrug use and overdose: overthrowing old myths

It is an extraordinary state of affairs, but it is only in the past decade that overdose has received consistent and serious attention. This despite the fact that overdose is consistently the single biggest killer of heroin users around the world, even in high HIV prevalence countries. However, the study of overdose has been hindered over the years by a number of cherished myths. These include: the ‘fact’ that overdose victims were young, inexperienced users; that death occurred almost instantaneously; and that overdose deaths were due to impurities. The foremost, and most insidious, of these myths relates to the term ‘overdose’ itself. Overdose has been conceptualized as the consumption of a quantity or purity of a drug that is in excess of the person's tolerance. Thus, for example, a heroin user who overdoses simply had too much heroin, or it was of particularly high purity. This has great intuitive appeal. It just does not coincide with reality, as studies from the 1970s onward have shown [1-5]. Brescher [1], writing 30 years ago, noted that many overdoses appeared to be ‘underdoses’. While this may sound a somewhat arcane and academic debate, it has major implications for how we go about reducing the morbidity and mortality that accrue from overdose. It is only by understanding what truly causes overdose that we can hope to reduce overdose rates. It is in this context that the study Coffin et al. [6] of overdose fatalities in New York City, presented in this issue of Addiction, represents a major advance. In their large-scale study, the authors convincingly relate variations in the fatal overdose rate to those in multiple drug deaths. Variance in overall drug overdose deaths was not due to differences in the single-substance drug deaths but, rather, to the number of multiple-substance deaths. My colleagues and I have examined the issue of heroin overdose in detail [7]. We have consistently argued that polydrug toxicity is a major factor in heroin ‘overdoses’. Coffin and colleagues have confirmed this relationship and, furthermore, extended it to cocaine ‘overdoses’. This is of crucial importance, particularly in countries where cocaine use is widespread. As evident from the title of their article, the role of alcohol in what are termed ‘heroin’ or ‘cocaine’ deaths cannot possibly be overstated. An injection of heroin or cocaine that may be well tolerated when sober may well kill when drunk. No-one, of course, is arguing that true pharmacological overdoses do not occur. However, the fact remains that most ‘overdose’ deaths involve multiple drugs. Coffin and colleagues are to be congratulated. This important work provides serious toxicological/epidemiological evidence of the extent to which polydrug use influences overdose rates, and highlights how overdose-related mortality and morbidity can be reduced. I strongly recommend all who are concerned about overdose as a clinical or research issue to read this important work.