Statistical Bulletin 2017 — problem drug use

‘Problem drug use’ is defined by the EMCDDA as ‘injecting drug use or long duration / regular use of opioids, cocaine and/or amphetamines’. This definition specifically includes regular or long-term use of prescribed opioids such as methadone but does not include their rare or irregular use nor the use of other drugs, such as ecstasy or cannabis. Existing estimates of problem drug use are often limited to opioid and poly-drug use.

Data tables

Use the menu below to find data sets of interest.

Methods and definitions used

The European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) problem drug use (PDU) indicator focuses on the prevalence and patterns of high-risk drug use (HRDU), including opioids, injecting drug use, cannabis, stimulants and other drugs, in the European Union Member States, Norway and Turkey. The EMCDDA monitors prevalence estimates and the results are reported in the Data and statistics section (Statistical Bulletin) of the European Drug Report.

The EMCDDA encourages countries to both produce and harmonise their estimates (1). Nevertheless, differences exist in the methods of estimation and target group both within and between countries, and these will influence the interpretation and comparability of the results.

Users of the data should carefully consider the methods and definitions described below and the notes associated with the data in the SB. The underlying protocol and guidance to data providers provide further information necessary for interpreting the data. The individual country National reports are a source of contextual and methodological information and the basis of much of the information here. All of the above are available on the EMCDDA website (see ‘Key links’ below).

HRDU prevalence estimates provide one perspective on the drug situation and should be considered within the broader range of available information. Cross-referencing between data-sets provides both additional insights and a check of validity. The PDU indicator also considers studies of high-risk drug users not in treatment and the incidence of high-risk drug use. EMCDDA monitoring of information on drug-related deaths, and in particular treatment (treatment demand indicator, opioid substitution treatment), provides perspectives and information on the more intensive forms of drug use and trends.

Purpose and aims

Estimates of prevalence are designed to establish the number of individuals engaging in various types of high-risk drug use, and are therefore of direct use to policymakers in evaluating resource allocation. One of their main purposes is to establish the need for drug treatment (see WHO, 2000 and EMCDDA, 2013).

Estimates of prevalence are also combined with other drug indicators, either for comparison — as a scale, baseline or denominator — or as a component of a broader estimate. For example, data on drug-related infectious diseases, wastewater analysis, drug-related deaths, opioid maintenance treatment and demand reduction programmes such as needle and syringe programmes may be compared to the number of high-risk drug users, both as a validation and as a measure of the size of the population under intervention or experiencing specific drug-related problems. Demand-side estimates of the size of the opioids market, and in limited cases of amphetamines, multiply the number of high-risk drug users by estimates of amounts used to obtain estimates of consumption.

Definitions

Within the PDU indicator, high-risk drug use is defined as ‘recurrent drug use that is causing actual harms (negative consequences) to the person (including dependence, but also other health, psychological or social problems) or is placing the person at a high probability/risk of suffering such harms’.

The operational definition provides the parameters for the measurement of high-risk drug use: ‘High-risk drug use is measured as the use of psychoactive substances (excluding alcohol, tobacco and caffeine) by high-risk pattern (e.g. intensively or in risky combinations (2)) and/or by high-risk routes of administration in the last 12 months.’

These definitions are further operationalised by drug (see ‘Key links’ below for PDU guidelines).

Methods

General population surveys (GPS) that are used to estimate the prevalence of use of drugs such as cannabis are not suitable for estimating the number of high-risk drug users. This is because GPS are not likely to capture with reasonable precision a group with a relatively small prevalence in the population, and individuals with more intensive forms of drug use are likely to be outside the sampling frame of GPS (e.g. due to a lack of permanent accommodation, or because they are staying in an institution). The likelihood of valid responses also decreases because they are engaging in an activity that may be highly stigmatised or possibly illegal.

The methods used to produce PDU prevalence estimates are therefore based mainly on indirect statistical extrapolations from existing records and registers, using drug use related indicators. They include:

  • a simple multiplier method using police, treatment, mortality or other data;
  • capture–recapture methods;
  • extrapolation from multiple subnational estimates to parts of the country with missing estimates via multivariate indicator methods.

Estimates are produced at the national level and the study methodology and results are reported to the EMCDDA through an online data collection system called Fonte (see the PDU guidelines and the data collection template in the ‘Key links’ below).

Limitations

Estimates of the prevalence of high-risk drug use should be interpreted with caution. Various factors will affect the validity of the estimates and their comparability between countries and over time.

As with any statistical technique, the appropriateness of the indirect estimation methods applied will depend on the underlying assumptions being met. Each of the methods used has its own set of assumptions, and the robustness of the estimates when these assumptions are not fully met should be considered.

For example, all methods dependent on linking two or more sources require accurate and complete methods of identification. Simple capture–recapture methods will rely on assumptions of independence of the data sources and a constant probability of capture, which, it has been argued, are unlikely to fully hold in epidemiological studies (see Jones et al., 2014; Dahlberg and Anderberg, 2013; Tilling, 2001; Domingo-Salvany et al., 1998).

In order to take into account the statistical uncertainty associated with a particular estimate, confidence intervals should be considered where available. These also limit the interpretation of trends.

Comparability between countries and over time for a single country will depend on a similar estimation method, case definition and data sources being adopted. Guidelines and protocols define a common standard for reporting the prevalence estimates of high-risk drug use. However, the data sources available and their quality differ between countries, and this will have defined what methods and case definition have been used. It is clear that resulting estimates from Member States differ due to these important characteristics. Moreover, it is important to consider the patterns of use and substances used within the individual countries when comparing prevalence estimates.

The characteristics of data reporting within individual countries have been compiled to inform valid interpretation of the data. Individual estimates are commonly published as studies in the scientific literature. Cross-referencing the prevalence estimates with EMCDDA and other data collections is recommended.

Key links

For further information please refer to the following:

•    PDU guidelines (under review): National | Local
•    The high-risk drug use web page
•    National reports
•    Data collection template

Other publications

Dahlberg, M. and Anderberg, M. (2013), ‘The hidden population: some methodological issues about estimation of problematic drug use’, Nordic Studies on Alcohol and Drugs 30(3), pp. 149–66.

Domingo-Salvany, A., Hartnoll, R. L., Maguire, A., et al. (1998), ‘Analytical considerations in the use of capture–recapture to estimate prevalence: case studies of the estimation of opiate use in the metropolitan area of Barcelona, Spain’, American Journal of Epidemiology 148(8), pp. 732–40.

EMCDDA (2013), PDU (Problem drug use) revision summary, EMCDDA, Lisbon.

Frischer, M., Hickman, M., Kraus, L., Mariani, M. and Wiessing, L. (2001), ‘A comparison of different methods for estimating the prevalence of problematic drug misuse in Great Britain’, Addiction 96, pp. 1465–76.

Jones, H. E., Hickman, M., Welton, N. J., et al. (2014), ‘Recapture or precapture? Fallibility of standard capture–recapture methods in the presence of referrals between sources’, American Journal of Epidemiology 179(11), pp. 1383–93.

Kraus, L., Augustin, R., Frischer, M., et al. (2003), ‘Estimating prevalence of problem drug use at national level in countries of the European Union and Norway’, Addiction 98, pp. 471–85.

Origer, A. (2011), ‘Prevalence of problem drug use and injecting drug use in Luxembourg: a longitudinal and methodological perspective’, European Addiction Research 18(6), pp. 288–96.

Scalia Tomba, G. P., Rossi, C., Taylor, C., Klempova, D. and Wiessing, L. (2008), Guidelines for estimating the incidence of problem drug use. EMCDDA, Lisbon.

Tilling, K. (2001), ‘Capture–recapture methods – useful or misleading?’ International Journal of Epidemiology 30, pp. 12–14.

WHO, UNDCP and EMCDDA (2000), Evaluation of psychoactive substance use disorder treatment, workbook 3: needs assessment, World Health Organization, Geneva.

(1) The EMCDDA compiles guidelines and expert opinion on a range of HRDU estimation methods, holds annual meetings of both experts and data providers, provides direct assistance to data providers, and monitors and reviews the academic literature. Further information is available on the EMCDDA website.

(2) These ‘risky combinations’ may include the licit substances mentioned above, in particular alcohol.

Loading