EMCDDA logo Annual report on the state of the drugs problem in the European Union 2001

Home of the Annual Report
Introduction to the annual report
Chapter 1: Drug demand and drug supply
Drug use
Health consequences
Law-enforcement indicators
Drug-market indicators
Chapter 2: Responses to drug use
Chapter 3: Selected issues
Chapter 4: The drugs problem in central and eastern European countries
Figures and tables
Download complete report in 12 languages  in Adobe PDF  format
News Releases and press information

Search Site Map Contact


Drug use

Drug use: Drug use in the general population | Problem drug use

Problem drug use

In general, prevalence rates of problem drug use seem to be highest in Italy, Luxembourg, Portugal and the United Kingdom, with between five and eight problem drug users per 1 000 inhabitants aged 15 to 64 (this is taking the midpoints of the range within a country). Rates seem lowest in Belgium (but this data refers only to IDUs and thus is an underestimate), Germany and the Netherlands, with two to three problem drug users per 1 000 inhabitants aged 15 to 64. Intermediate rates range from between three and five problem drug users per 1 000 inhabitants aged 15 to 64 in Austria, Denmark, Finland, France, Ireland, Spain and Norway. Some countries report slightly different figures than in 2000 due to improved data and estimates. Denmark and especially Finland, however, have much higher figures owing to having suppressed less reliable (and in both cases lower) estimates.

Besides estimates of prevalence, data on patterns of use are important. However, there are little data on patterns of drug use among problem drug users who are not in treatment, although this is important information for guiding policy decisions on future treatment demand. In northern countries, notably Finland and Sweden, primary amphetamine users form the majority of problem drug users (an estimated 70 to 80 % in Finland in 1997). This is contrary to other countries where problem drug users are mostly primary opiate users (but often polydrug users).

Definition and methods used to estimate problem drug use
Figure 3


n.a. = data not available.
Differences between countries have to be interpreted with caution due to different methods and target groups. For more detail, see online Tables 1OL and 2OL. Data for Austria and Belgium are for 1995, Italy for 1999 and Portugal for 1999-2000. The original Swedish estimate has been reduced by 8% to exclude cannabis addicts compatible with EMCDDA definition of problem drug use. Colour for a country indicates the midpoint of the range in estimates, except for Belgium (point estimates 3.0 and 95% confidence interval). The Belgian estimate refers to IDUs and thus underestimates total problem drug use.

Reitox national reports 2000. For Sweden: B. Olsson, C.A. Wahren, S. Byqvist, Det tunga narkotikamissbrukets, omfattning I Sverige 1998, CAN, Stockholm 2001.

'Problem drug use' is defined here as 'injecting drug use or long-duration/regular use of opiates, cocaine and/or amphetamines'. This definition excludes ecstasy and cannabis users and those who never - or irregularly - use opiates, cocaine or amphetamines. Opiates include prescribed opiates such as methadone.

The national estimates of problem drug use reported here for the EU and Norway (which also participated in the EMCDDA project on prevalence estimation) are for 1996 to 1998. Austria (1995) and Belgium (1995) could not provide estimates for this time period, due to lack of more recent data on which to base the calculations (see Figure 3), while Luxembourg, Portugal and Italy provided more recent ones (1999-2000).

The methods used to produce these estimates are mainly based on statistical models incorporating drug-related indicators and include:

  • the multivariate indicator method;

  • capture-recapture;

  • three multiplier methods based on police data, treatment data and mortality rates; and

  • a multiplier method using back-calculated numbers of intravenous drug users (IDUs) with HIV/AIDS, in combination with HIV/AIDS rates among IDUs.

The ranges given in Figure 3 are often derived from a multiple method approach; therefore the lowest and highest figures may have been obtained by different methods, both within and between countries. These methods do not always refer to the same target group - for example, HIV/AIDS back calculation and overdose mortality multipliers target IDUs, while multipliers from treatment data could only be used for the wider group of problem opiate users.

Several countries applied multiple estimation methods - two (Spain, Luxembourg and the Netherlands), three (Germany, France, Ireland, Finland and the United Kingdom) or even four (Italy and Portugal). Other countries (Belgium, Denmark, Austria, Sweden and Norway) could only apply one method, while Greece was still unable to provide an estimate. Using multiple independent estimates may cross-validate the single figures and lead to a more reliable overall estimate for a country. Therefore, a multiple method approach, if possible on a year-by-year basis, may ideally be applied.


Although important for policy evaluation, it is still not possible to calculate trends in the prevalence of problem drug use with the current methods, data quality and data availability. Incidence of problem drug use may be an additional and important way of measuring trends. Incidence is the number of newly occurring cases in a year, while prevalence is the total of all existing cases, new and old. Therefore trends in incidence are much more sensitive to epidemiological changes and to interventions that prevent new cases from occurring.

An EMCDDA project produced new incidence estimates of heroin use for Belgium, Lisbon, Budapest, London and Amsterdam suggesting that the timing of heroin epidemics in those regions may have been very different (see Figure 4, London and Amsterdam not shown).
Figure 4


Relative incidence is the incidence (number of new users per year) of opiate users who will eventually go to services. Lisbon: extremely problematic area of Lisbon, users presenting themselves to social services. Belgium: sample going for treatment in French community.

EMCDDA project CT99.EP.05, 'Study on incidence of problem drug use and latency time to treatment in the European Union', Lisbon 2001 (coordinated by University Rome ' Tor Vergata').

Figure 5 shows how, even within one country, in this case Italy, heroin epidemics may occur in different regions at different moments. The results also suggest that there might be an initial (epidemic) period of rapid spread followed by saturation effects (that is most susceptible persons have become heroin users) and lower but constant further spread (endemic situation) — a mechanism which is similar to the spread of infectious diseases. This may have important implications for decision-makers as regards the optimal balance between policy measures such as primary prevention, which needs to be timely to prevent high prevalence, and secondary prevention (drugs treatment and harm-reduction measures), which becomes more important once high prevalence has been established.

Figure 5

Figure 5


Relative incidence is the incidence (number of new users per year) of opiate users who will eventually go to services.

EMCDDA project CT99.EP.05, 'Study on incidence of problem drug use and latency time to treatment in the European Union', Lisbon 2001 (coordinated by University Rome ' Tor Vergata').

In Figure 5, it is notable that the heroin epidemic in Puglia seems to have started earlier than in Sicilia and Campania. Puglia, the 'heel' of Italy, is situated on an important heroin trade route from the Balkan region into Europe, which may have led to earlier spread of heroin use. This confirms earlier results based on geographical analysis of treatment data (2000 annual report, EMCDDA).

Differences in the prevalence of drug use are influenced by a variety of factors in each country. As countries with more liberal drug policies (such as the Netherlands) and Those with a more restrictive approach (such as Sweden) have not very different prevalence rates, the impact of national drug policies (more liberal versus more restrictive approaches) on the prevalence of drug use and especially problem drug use remains unclear. However, comprehensive national drug policies are of high importance in reducing adverse consequences of problem drug use such as HIV infections, hepatitis B and C and overdose deaths. Other factors that may affect prevalence of problem drug use are the availability and price of drugs, unemployment and poverty or other societal problems (war), the age structure of a country and the proportion of urban and rural areas. Furthermore, drug use seems to behave in an epidemic manner (see for example Figures 4 and 5). It may thus follow long-term epidemic cycles that for a large part depend on the demographic replenishment of new generations of 'susceptible' young people, who have never experienced the problems associated with heavier patterns of use. At present, therefore, a final causal interpretation of varying prevalence rates within EU Member States cannot be made.

Additional qualitative or local information on trends in problem drug use is available for some countries or cities.

In Austria, problem drug use prevalence seems currently stable in most regions, although in Vienna street workers have registered a decline in the number of young newcomers on the scene. In France, heroin consumption is in general declining although it is not clear what this means in terms of numbers of heroin users and there are some indications of new, local increases. In Finland, prevalence estimates for the greater Helsinki area for 1995-97 suggest that the use of hard drugs has increased in this area by a minimum of 40 % in two years. This increase was most pronounced in males, those over 26 years old and in amphetamine users.

In Germany, heroin use is primarily found in metropolitan areas, prevalence rates and seizures in rural areas being much lower. In the new LŠnder, heroin use is still scarce. Surveys suggest that heroin use is only slightly increasing or stagnating since 1992; however, cocaine shows stable and uniform growth. In Greece, data from indirect indicators (treatment, deaths, low-threshold services) suggest that problem drug use is increasing. In Ireland, drug users are young reflecting the general demographic situation. Until about 1996, the trend in heroin use was towards increased smoking; however, more recently, injecting is again increasing. In Italy, drug use levels seem higher in northern regions but problem drug use may be more evenly distributed between northern, central and southern regions even if the level of 'at risk' population is different between geographical areas. (The total population of 15—54-year olds is around 32 million - 44 % live in the northern regions, 19 % in the central regions and 37 % in the southern regions.) In addition, northern regions may have relatively more use of 'recreational drugs'.

In Luxembourg, the national registration system showed a marked increase of 42 % of new registrations between 1998 and 1999, however this is mainly due to drug law offenders and may reflect policing activity. There is a large proportion of non-natives from Luxembourg (48 %) among problem users. The estimated prevalence shows an upward trend; while injecting drug use has decreased from over 90 % to about 66 %. In Amsterdam, the Netherlands, the number of heroin users has moderately decreased over the years, while the proportion of injectors among them has strongly decreased. In Portugal, drug use problems are more significant in the districts of Lisbon, Porto, Setubal and Faro. In Spain, heroin use, especially injected, is becoming less relevant and cocaine problems are becoming more important among drug problems. In Sweden, there has been an increase in the number of severe drug users during the 1990s and heroin use is becoming more common in younger groups of problem users. In the United Kingdom, a report published in 1998 suggested an increase of new heroin outbreaks among young people in most regions of England since about 1996, while a recent study in London suggested a two-fold increase in incidence of non-injected heroin use between 1991 and 1997.

Drug use: Drug use in the general population | Problem drug use

Drug use | Health consequences | Law-enforcement indicators | Drug-market indicators
© EMCDDA 2001 | home | webmaster | disclaimer