Lyons, Suzi
(2009)
Drug use: an overview of general population surveys in Europe.
Drugnet Ireland,
Issue 31, Autumn 2009,
p. 29.
The EMCDDA recently published an overview of general population surveys on the prevalence of drug use in Europe.1 The authors conducted a meta-analysis of the national surveys in use in 18 countries since 2001 and compared the different survey methods used.
The participating countries provided information about their national surveys under four headings: the intrinsic focus (design of the questionnaire and type of questions asked), the methodology used, financial issues, and the utility and accessibility of the results. The overview identified intrinsic and methodological issues relating to the comparability of the national surveys and the standardisation of definitions used, some of which are outlined below.
The surveys in many countries comply, wholly or in part, with the European Model Questionnaire (EMQ), which allows cross-country comparisons. However, not all countries conduct surveys specifically on drug use. Some include drug-related questions in general health or crime surveys, which limits the number of drug-related questions that can be asked. This means that the surveys conducted in different countries vary in the number of items and the number of questions included.
While most countries want to include in their surveys questions about all the psychoactive drugs that are known to be used, it is recognised that estimates obtained in this way actually underestimate the use of some drugs in the general population.
Most national surveys include three prevalence indicators for each drug included: lifetime use, last year use and last month use. However, some countries feel that these prevalence rates must be interpreted with caution because of the low number of users among the general population. Measuring the prevalence of licit drugs tends to be limited.
The authors report that countries are divided on the inclusion of an item on intravenous drug use in their surveys. The argument against including such an item is, again, linked to the low prevalence in the general population and the belief that such users, if reached, would not be willing to reply to the questionnaire.
The authors found no consensus about the measurement of the frequency of use of either licit or illicit drugs. Several countries, including Ireland, measure frequency of use in the last month only as they feel that is the most useful recall period; other countries are of the opinion that the sample size for last-month frequency tends to be too small, and therefore also measure lifetime and last-year frequency.
A high response rate is generally seen as a mark of quality in a survey. The rates reported in this overview ranged from 21% in Austria to 91% in Hungary. Information on how non-response rates are calculated is often not reported. The authors suggest that standardised definitions and formulas for calculating response rates be provided so that better cross-survey comparisons can be made.
Securing funding to repeat surveys is a major problem for some countries and leads to restrictions on the length of the questionnaire and the sample size. The authors noted that most national survey reports do not contain any information about the costs of the survey.
In conclusion, the authors suggest that some of the intrinsic and methodological issues should be standardised to improve comparability of general population surveys across Europe.
1. Decorte T, Mortelmans D, Tieberghien J and De Moor S (2009) Drug use: an overview of general population surveys in Europe. EMCDDA: Office for Official Publications of the European Communities.
