Appendix C Extent and Patterns of Drug Use

C.1 INTRODUCTION

Despite an enormous amount of research on the extent and patterns of drug use in the last few years, a number of considerable difficulties still remain in providing information on this subject. Some of these difficulties are conceptual, some arise from the types of populations studied by researchers, and some have to do with temporal changes in the phenomenon of drug use itself. Various populations are not equally amenable to social scientific investigation, and we realize that drug use patterns in certain groups (for example, among most institutionalized persons) remain invisible. In this section we set out these and other methodological difficulties and indicate the rationale of our approach to the extent and patterns of non-medical drug use. Tentative estimates of the number of current non-medical drug users in Canada, and their social characteristics, are provided in the second and third parts of this appendix, C.2 Extent of Use and C.3 Characteristics of Users. The fourth section, C.4 Patterns of Use is devoted to a description of Canadian drug use patterns.

Information on any subject to do with human behaviour always derives from a delimited group of people. Such a group is called a 'population' by social scientists. Everyone 18 years of age and over and living in Toronto in the spring of 1973 constitutes a population, and all university students in Canada in the spring of 1973 constitute another population, one that happens to overlap slightly with the first. By whatever method information is collected, from whatever type or size of 'sample', the information can, with any certainty, represent only the situation in the population from which it was obtained. Uncertainty necessarily enters when the information is projected to other populations. Research in the drug field has concentrated disproportionately on particular populations, reflecting temporal changes in public concern. In the post-war United States, when heroin use was the principal concern, lower-class people of specific ethnic minority groups were the main populations studied. Later, when cannabis and hallucinogen use began to spread, first university students and then high school students became the populations of most interest to researchers, partly because they were seen as especially 'at risk' to drug use, and partly because they were easily accessible. This tendency to study drug use in special populations creates problems of comparability from study to study, and of generalizability to larger populations. It has undoubtedly made the drug use patterns of these special populations seem more exotic than they would appear in the context of the drug use patterns of our society as a whole.

Our primary concern is non-medical drug use. But what do we mean by 'drug use' and what constitutes 'non-medical' use? Many different answers to these questions are possible, and many can be found in the studies that we depend on for data on extent and patterns of use. How drug use is defined depends on the particular problem under study, and this leads to various definitions and, consequently, a certain degree of confusion. For our purposes, however, we refer the reader to the Interim Report in which the non-medical use of drugs is defined as "all drug use which is not indicated on generally accepted medical grounds. . . ."

Many studies attempt to say something about the 'drug culture', and hence focus on the use of illegal drugs: cannabis, 'speed', and the hallucinogens, for example. If the study is of a high school population, then for most of its subjects alcohol will also be an illegal drug. However, alcohol has frequently been omitted from past consideration since researchers have not ordinarily associated it with the drug culture. Fortunately this situation is changing, as indicated by the inclusion of alcohol-related questions in many recent Canadian surveys.

Some studies are directed at adult use of pharmaceutical substances in order to evaluate the degree to which this use accords with medical norms. Here alcohol, tobacco, and illegal drugs are typically ignored. Studies of tobacco use generally define that use in terms relevant to a concern with the increased risk of disease, and studies of alcohol use tend to reflect a concern with the development of alcoholism, and thus to remove each from a broader consideration of the patterns and context of drug use. Many researchers have failed to examine the larger context within which the use of certain substances occurs and have displayed stereotyped thinking regarding both particular drugs and the populations to which they are presumed to be relevant.

In studies of the use of illegal drugs, use is typically regarded as being ipso facto non-medical use, and is therefore artificially isolated from all other current patterns of drug use. This assumption may disguise illuminating parallels between the use of these illegal drugs and the use of legal psycho-tropic substances employed for medical purposes. The opinion of the user himself as to the function of his drug use (be it medical or non-medical) is subjective, but it may be a relevant definition for certain purposes such as understanding those factors which motivate some persons to use drugs on an initial or continuing basis.

In discussing measurements of extent of use, we have made no judgment, in the first instance, as to the medical or non-medical purpose of the use. We thus have a maximum figure for the extent of use of the substance in question. We then qualify the maximum figure with estimates of non-medical use. In some cases these estimates are quantitative; in others they can only be in qualitative terms. Wherever possible we point out subcategories of users of a drug in terms of recreational, functional or medical use of the substance (see C.4 Patterns of Use, page 707), in terms of the level-of-use of the substance, and in terms of the short-term or chronic character of use. We know that these subcategories of users are smaller than the totality of users, but in most cases there are no data available that enable us to determine their size exactly.

By 'extent of use' of a drug we mean how widespread its use is in a population. Since the studies on which this appendix is based do not use a single precise definition of extent of use, we are in no position to do so either. The term becomes more precise when level-of-use' is defined, and a period of time is specified. We may then be able to say what proportion of Canadians have used, for example, barbiturates at least once a month in a period of a year. Or we may define the phenomenon much more broadly, and ask what proportion of Canadians have used barbiturates in their life time. The second approach to determining extent of use will undoubtedly yield a different phenomenon in terms of the social characteristics of users and of the patterns of barbiturate use, and of drug use more generally, than the first definition.

The want of sophistication that still plagues studies of non-medical drug use shows up most obviously in the treatment of levels-of-use. By levels-ofuse' we mean the frequency and regularity (and, in some cases, dose levels) with which individual users consume a particular drug, or the total number of times that they have used it. A minimum level-of-use must be specified in order to define a user. We may ask how many people have ever used a particular substance, even once. This is the simplest measure of drug use in a population, and for most purposes the least useful since the levels-of-use, and whether the use is current or took place in the more or less distant past, are not known. In some studies the measure is sharpened to include only those who have used the substance at least once in a fixed period of time (for example, in the last six months), but this still leaves room for a wide variation in levels-of-use. Despite these problems, 'ever used' and 'any use in the last six months' remain the most commonly employed operational definitions of drug use. The matter becomes more complicated when we consider that it has become typical of studies of non-medical drug use to survey the use of a range of substances, and to define that use in terms of the same levels-of-use definitions for all drugs. However, the relevance of a level-of-use definition depends on the particular substance: cannabis used once a month would probably not be considered heavy use, but LSD used once a month might well be.

The distribution of levels-of-drug-use in a population at a given point in time has received increasing attention in recent years. The majority of the relevant studies have focussed on per capita alcohol consumption, although there is also more limited evidence that similar distributions of use occur with a variety of other psychotropic drugs as well. It would appear from available data that levels-of-use of a drug in the user population may be distributed in a way which can be described by a continuous smooth curve, in some circumstances approximately a log-normal distribution (see Figure C.1). Although the exact mathematical specifications of the distribution are not essential here, its general form can provide significant information.", 130,

Within this distribution, the majority of those who use a drug use relatively little, and as the amount used is increased, the number of individuals involved at these levels decreases, at first rapidly and then more slowly, but without any break that would differentiate various levels-of-use. Available data suggest that extreme use is on a continuum with light and moderate consumption, and that discrete levels-of-use categories do not exist as such. The unimodal shape (i.e., having only one peak or 'mode') of this distribution may be of considerable importance. If, for example, users of a particular substance became 'fixed' at a given level of consumption, we would expect some clustering at such a level which would be indicated by a bimodal distribution curve (i.e., having two peaks or 'modes') rather than a unimodal one. No evidence of such a subpopulation, defined only by levels-of-use, has been indicated by the available data.

For practical purposes of analysis, however, discrete and necessarily arbitrary divisions on the levels-of-use continuum are made in much of the following discussion.

In this appendix level-of-use is employed as an indicator of an individual's stage in a social process, the process of becoming a user of a drug, his use governed by the norms of the using culture. The examination of this social process or 'career' of drug use can be viewed as the study of how an individual changes his position in the per capita consumption distribution over time (which is a within-subject analysis over time, as opposed to a between-subject analysis at a given point in time).51 This social process by which an individual becomes a user of a particular substance suggests that the first conceptual distinctions in levels-of-use are between non-users and 'experimental' users, and between experimental users and all more experienced users. These more experienced users can then be subdivided into `occasional' and 'regular' level-of-use categories. Experimental use, occasional use and regular use are defined more concretely in C.4 Patterns of Use, where regular users are further examined in terms of the moderate or heavy character of their use.

Differences in levels-of-use of a drug may be an indicator of different drug-using subcultures. When that is the case, the differences in level-of-use are usually accompanied by a number of other differences in style of use. If the definition of use of a substance does not in some way take into account these differences in style of use, then disparate phenomena may be analysed together as if they were the same. 'Speed freaks', for example, use amphetamines, but are not typical of amphetamine users as regards level-of-use, dosage patterns, mode of consumption, or subcultural values. If a study fails to recognize these distinctions and still goes on to examine the social correlates of drug use, the results will be misleading. When the extent and patterns of use of a drug are examined, information on the style of use as well as the level and duration of use of that drug must be available. It is only with such information that there is any hope of elucidating the social context of use, and of evaluating the physiological and psychological consequences of varying consumption patterns.

There are various approaches to, and sources of, information on the extent and patterns of non-medical drug use, differing in reliability, validity and generalizability. The first and most common approach to drug use data is surveys and, under certain circumstances, surveys based on random samples provide the most valid extent of use information for a population. However, if the number of users in the total population is very small, or if they are geographically "clustered" (i.e., live close together in specific areas) or frequently institutionalized (hospitalized or imprisoned), then the size of the sample that would yield valid estimates of numbers of users may be impractically large. Thus, most surveys may yield valid data on alcohol use, but not on cocaine or heroin use because of the relatively small number of users of such drugs in the general population. Additionally, the legal status of a drug or the relative stigma associated with its use may affect the likelihood of accurate responses and, thus, the overall validity of drug survey findings.

The quality of social research in the Canadian drug field has risen considerably in the past few years. In the spring of 1971 the Addiction Research Foundation of Ontario conducted the first random sample survey, apart from samples drawn for the Commission, of a general population of adults, and a broad range of drug use patterns was examined.82. 239 It is to be hoped that this represents the beginning of a trend to less specialized samples, a broader definition of the phenomenon under scrutiny, and a more profound analysis of the relationships that are found. The unsophisticated polling of correlates of drug use that characterized many studies in the past yielded prevalence of use rates which have some value but which now need to be carefully interpreted. A much more sophisticated approach is presently required. Specifically researchers must be more analytical and precise in the questions they ask when investigating the phenomenon of non-medical drug use, and they must be prepared to conduct open-ended and extensive interviews and, if necessary, apply ethnographic techniques to ensure sociologically meaningful measures of the phenomenon. The time for exploratory surveys is past.

If the use of a substance is known to be concentrated within a particular subculture or relatively small geographical area, then anthropological techniques, of a qualitative sort, may be the most effective means of determining extent and patterns of use. There are two difficulties in using this second approach to drug use data to describe the situation in the general population. First of all, the assumption cannot usually be made that use is concentrated in a subculture. Subcultural trappings may make certain kinds of drug use more visible and lead the observer to the conclusion that that sort of drug use is always found in association with those trappings: for example, assuming that all hallucinogen users affect long hair, beads and dishevelled dress. The use of a substance may, however, be mediated by more general social norms, and in that case the use is liable to be less visible, but still definitely present. A second difficulty is that anthropological studies provide no basis for predicting the consequences of an increase in the extent of use of a drug. If the extent of use of a drug increases because the particular beliefs and practices that define the subculture are spreading, then no problem arises. However, it is possible that the extent of use is increasing because the use of the drug has expanded beyond the boundaries of the subculture. In this case, many of the concomitants of use will drop away, having been the consequence not of the substance itself, but of the subculture. This appears to be the case with cannabis, which has now escaped the boundaries of the `hippie' subculture and has consequently shed these subcultural trappings. As a result, the social characteristics of cannabis users have changed, relegating the early studies of communities of cannabis users to a position of a largely historical interest.

Drug use may also be measured indirectly through such data as illicit drug analyses, licit drug sales, psychiatric or medical epidemiology (for example, liver cirrhosis and adverse reaction statistics), arrest statistics, or police seizures. This third source of drug use information was the usual mode of determining extent of use before drug-related surveys became popular in the mid-1960s. Some of these measures are highly sophisticated (in particular those used to estimate numbers of alcoholics and heroin dependents), but they are often a source of interpretive disagreement. The ultimate origin of such information is commonly the tabulations of the consequences of the activities of such groups as police and hospital officials. These data are therefore susceptible to changes resulting from policy alterations within the agencies that generate them. Furthermore, any such measures depend on an explicit or implicit function or relationship linking it to the extent of drug use that it is to measure. If the true extent of drug use changes, the function may also change, and this is a second reason why such measures may yield false conclusions.

The Commission has made use of all three of these sources of information: surveys, more qualitative anthropological approaches, and indirect measures.

The York University Survey Research Centre, in collaboration with Le Centre de Sondage de l'Universite de Montreal, conducted national surveys for the Commission on drug use of the population of Canada 12 years of age and over in the spring of 1970. These surveys (the only methodologically rigorous general national surveys to date) were intended to measure the extent of drug use, including individual drug use histories, attitudes toward and knowledge of drugs, and attitudes toward and knowledge of the law, together with a number of social and social-psychological variables.142, 143, 144

There were three national Commission surveys: one of high school students, one of college and university students (including both undergraduates and graduates), and one of a group that is most conveniently called 'adults' and that was defined as everyone over the age of 12 who was not attending a primary or secondary school.

The adult and high school surveys were based on multi-stage samples: the first stage was geographical areas, with far northern areas being excluded. Households were listed within sampled areas, and randomly sampled in turn. Since households formed the sampling frame, or units of analysis, at this stage, people in prisons, hospitals, old people's homes, on Indian reserves and in institutional dormitories were excluded. The high school sampling frame included all students in grades 7 to 12 or 13 who were over the age of 11 and living in the sampled households. The adult frame included everyone else over the age of 11. The sample of university students was randomly selected from registrars' lists from universities and colleges selected in a purposive sample to represent all Canadian colleges and universities. Each survey used a different method of data collection: the adult survey used an interview, the high school survey used a questionnaire distributed by the interviewer to all eligible children in the household and completed by them in private, and the college and university survey used a mailed questionnaire. The adult survey yielded 2,749 usable interview schedules, while the high school and college and university surveys each yielded 1,213 respondents. The response rate for the adult survey was 79 per cent, and for the college and university survey, 73 per cent. Since the high school survey was dependent on the national adult sample, it is not possible to present a meaningful response rate for this population.

The Commission also conducted a special survey of adult drug users in five Canadian cities,'" and a smaller study of regular cannabis users to quantitatively determine their drug consumption patterns. 1°5 In addition, the Commission made use of many published and unpublished survey studies of various populations in Canada and other countries, as well as extensively reviewing the social scientific literature related to extent and patterns of drug use. The Commission also conducted participant-observation studies in several `street-level' drug-using communities in 1970,103 and regularly monitored drug use patterns in several of these communities for over two years. Recently, Commission observers returned to a variety of cities across Canada to further question knowledgeable persons on changes in patterns of drug use since
1970.104
The Commission attempted to evaluate the extent of certain drug-related deaths in Canada between 1969 and 1972.173 Besides these special analyses of national data on drug deaths supplied by Statistics Canada, the Commission surveyed coroners in each province requesting coroners' reports and related documents pertaining to the drugs of primary interest. The Commission also surveyed all psychiatric hospitals in Canada in April 1971 to determine the extent to which certain drugs, including alcohol, were mentioned in the diagnoses of hospitalized patients.110 In addition, special analyses were done of national mental health data provided by Statistics Canada. The results of these studies are discussed elsewhere in this report (see Appendix A The
Drugs and Their Effects).

The Commission has had access to the annual tabulations of numbers of `known habitual illicit narcotic drug users' compiled by the Bureau of Dangerous Drugs of the Health Protection Branch, Department of National Health and Welfare (some of these tabulations for 1972 are presented in Annex 2 of this appendix), as well as many studies measuring numbers of addicts and alcoholics in several jurisdictions using various indirect indicators. In addition, R.C.M. Police estimates of Canada's addict population were available to the Commission.

In this appendix we primarily make use of a particular model to explain human behaviour, that which interprets human conduct as social conduct governed by definitions and evaluations learned from other persons. The individual and his drug use are linked within a social context by definitions, practices and values, in other words, by the elements of a culture. If there are different patterns of drug use within a society, these are mediated by different cultures. By culture, we mean merely a body of tools, definitions, norms and values bearing upon some ongoing human activity. Thus there can be cultures within cultures. Those cultural elements which are common to a whole society make up the core of any individual's culture. However, if cultural elements which cover sufficient of the tasks of social life as to mark an individual off from others are shared among a small proportion of the population, they constitute a subculture. For certain purposes society can be seen as a mosaic of subcultures. We are concerned with subcultures only insofar as they display distinct subcultural patterns of drug use.

If the use of a drug spreads from one subculture to another, it will not necessarily take its cultural baggage with it. Therefore, the social concomitants of use may change. The more widespread use becomes, the more the social characteristics of users will approach those of the general population, if they were not the same to begin with. This process, as has been previously noted, is strikingly illustrated by the case of cannabis.

We must make use of past information to arrive at a description of the present, and therefore must project the past into the present in some way. There are so few observations on drug use in the same population over time in Canada (or, in fact, in any other country) that formal statistical techniques of extrapolation or projection are not useful. However, certain assumptions about the social context within which the diffusion of drug use occurs, and about social processes more generally, allow us to make descriptive statements that reflect all of the available data and the consensus of opinion on the subject in the social sciences. These assumptions are more specifically delineated in the following section.

No society ever stands still. However, we can imagine a society in which all social forces are in equilibrium, in which no social change is taking place. Each drug would then presumably be consumed by a constant proportion of the population, and the various patterns of use of each drug would be relatively fixed. All drug use would then be at a plateau. If there is a change, and the social forces move to a new equilibrium, then the extent and patterns of use of each drug would probably change and move to a new plateau, there to remain until another shift in social forces takes place. This is not the only way of conceiving of social processes, but it is a useful way of thinking about significant changes in the extent and patterns of drug use. Estimating changes in extent and patterns of drug use in the society as a whole thus becomes a process of estimating the plateaux of use in the various subcultures of a society, and determining the culturally conditioned style of use for each subculture. The results of this method of analysis follow.