The quantification of morbidity and mortality caused by substance abuse. Prepared for the Second International Symposium on the Social and Economic Costs of Substance Abuse, 2-5 October 1995

In this paper we identify the major issues in attributing mortality and morbidity to substance abuse, review studies of the quantification of morbidity and mortality due to substance abuse and discuss guidelines for choosing the most appropriate methods for quantifying substance-related morbidity and mortality. Attention is restricted to alcohol consumption, cigarette smoking and illicit drug use (cannabis, opiates, stimulants including cocaine and amphetamines, and hallucinogens). While it is sometimes difficult to make a distinction between licit abuse and illicit use of the latter class of substances, we have focussed our attention on those for which the majority of associated morbidity is likely to be related to illicit use. Thus, we have not considered morbidity associated with substances such as minor tranquillisers (eg benzodiazepines) and with other substances that may be misused, such as volatile solvents. Throughout the paper, the terms substance and drug will be used interchangeably to avoid undue repetition. The simplest way to calculate the number of drug-caused deaths, illnesses or injuries is to count the number which are known to be drug-caused. However, this method is only possible when one can classify individual deaths, illnesses or injuries into drug-caused or otherwise. Where multiple causal pathways lead to the condition, and the clinical manifestations of the condition do not vary according to the aetiologic mechanism, as is the case, for example, with ischaemic heart disease in relation to cigarette smoking, and where empirical induction periods are long, classification of individual cases is not possible. In these circumstances, one must apply a probability measure to individual deaths, where the measure is the probability that the death, or other event, was drug-caused. In practice, we estimate the proportion of all deaths, illnesses or injuries occurring in a population that is attributable to substance abuse and multiply by the total numbers of deaths, etc, to derive indirectly the number in a population caused by substance abuse. These indirect methods are used for the majority of diseases and injuries caused by substance abuse and are discussed in detail in the next section.