European Neuropsychopharmacology 15 (2005) 389 – 397

www.elsevier.com/locate/euroneuro

Problematic drug use and drug use disorders in EU countries and Norway:
An overview of the epidemiologyB
Jürgen Rehma,b,c,*, Robin Roomd, Wim van den Brinke, Ludwig Krausf
a
ISF-Addiction Research Institute, Konradstr. 32, CH 8031, Zurich, Switzerland
b
Centre for Addiction and Mental Health, Toronto, Canada
c
University of Toronto, Canada
d
Centre for Social Research on Alcohol and Drugs, Stockholm University, Stockholm, Sweden
e
Department of Psychiatry, Academic Medical Center University of Amsterdam, Amsterdam Institute of Addiction Research, The Netherlands
f
Institute for Therapy Research, Munich, Germany

Abstract

Objective: To estimate the prevalence of drug use disorders in the European Union and Norway. Method: Based on a systematic
literature search and an expert survey, publications after 1990 on prevalence of drug use disorders (DUD, defined as drug
dependence and drug abuse or harmful use) in EU countries and Norway were reviewed. The search included both direct
estimations based on general population surveys using the DSM-IIIR, DSM-IV, or ICD-10 definitions of DUD; and indirect
estimates based on other epidemiological methods, such as multiplier procedures based on treatment or legal data and capture –
recapture techniques. The indirect methods did not use diagnostic criteria, but criteria based on duration and pattern of use, labelled
as problematic drug use as a meaningful approximation. Results: The majority of DUD as estimated from direct methods using
general population surveys were cannabis use disorders, usually not included in indirect estimates. The prevalence of thus defined
disorders can be as high as 3%. For disorders other than cannabis use disorders (i.e. opioid, cocaine and amphetamine use
disorders), indirect estimates of prevalence were consistently higher than those based on direct estimates, and ranged between 0.3%
and 0.9% in European Union countries and Norway. Men have higher prevalence rates of DUD than women, but the difference was
much less pronounced in general population surveys. Younger age (18 – 25 years) is the age group with the highest estimates.
Conclusion: General population surveys typically result in a serious underestimation of the prevalence of DUD other than cannabis
use disorders, because many people with DUD are not reached by these surveys (hidden populations). Based on the more valid
indirect estimates, it is concluded that problem drug use constitutes a relatively high burden of disease and social problems in
Europe.
D 2005 Elsevier B.V and ECNP. All rights reserved.

Keywords: Illicit drugs; Drug dependence; Drug abuse; Harmful use of illicit drugs; Prevalence; General population; Hidden population; Multiplier method;
Capture – recapture

1. Introduction
B
This paper was prepared in the framework of European College of
Neuropsychopharmacology (ECNP) Task Force project on Size and Burden The recently published Comparative Risk Analysis
of Mental Disorders in Europe‘‘ (PI: Hans-Ulrich Wittchen). The paper also (CRA) led by the World Health Organization (WHO)
serves as input for the European Brain Council (EBC; www.ebc- estimated that 2.1% of the burden of disease in European
eurobrain.net) Initiative ‘‘Cost of Disorders of the Brain’’ (Steering
countries with very low mortality and 1.6% of all
committee: Jes Olesen, Bengt Jönsson, Hans-Ulrich Wittchen).
* Corresponding author. ISF-Addiction Research Institute, Konradstr. 32, European countries in the year 2000 were attributable to
CH 8031 Zurich, Switzerland. Tel.: +41 1 448 11 60; fax: +41 1 448 11 70. illegal drug use, as estimated by combining injection
E-mail address: jtrehm@aol.com (J. Rehm). opioid, cocaine, and amphetamine use (WHO, 2002;
0924-977X/$ - see front matter D 2005 Elsevier B.V and ECNP. All rights reserved.
doi:10.1016/j.euroneuro.2005.04.004
390 J. Rehm et al. / European Neuropsychopharmacology 15 (2005) 389 – 397
Ezzati et al., 2002; for a categorization of countries into
mortality strata WHO, 2000). The overwhelming majority
of countries of the old EU and Norway falls into this
category of very low mortality, with the countries joining
in 2004 being scattered in different mortality categories.
This indicates a considerable public health burden,
although it is clearly smaller than the disease burden
attributable to use of alcohol (10.2% of all disease burden
in Europe) or tobacco (12.4% of European burden; both
numbers recalculated from WHO, 2002). Before further
discussing the meaning of this burden, however, we need
to understand the nature and validity of the underlying
prevalence rates.
The WHO CRA estimates are based on the UN Drug
Control Program’s (UNDCP, 2000) Global Illicit Drug
Trends for persons over the age of 15 years and the
additional assumption, that 28% of all users in the past
year were problematic users, the latter fraction being
derived from an Australian national survey (Hall et al.,
1999). The resulting prevalence estimates were around
0.1% for problematic opioid use and cocaine use each,
and between 0.1% and 0.2% for problematic amphet-
amine use. In addition, there was considerable overlap
between problematic use of these substances (Degenhardt
et al., 2001). The definition of problematic drug use by
UNDCP depends on ‘‘the extent to which use of a
certain drug leads to treatment demand, emergency room
visits (often due to overdose), drug related morbidity
(including HIV/AIDS, hepatitis etc), mortality and other
drug-related social ills’’ (UNDCP, 2000; Degenhardt et
al., 2001).
While this definition contains elements of defining drug
use disorders (DUD) according to ICD-10 (WHO, 1993; see
Rehm et al., 2005) or DSM-IV (American Psychiatric
Association, 2000), and while the operationalization of
Degenhardt et al. (2001) described above makes use of this
overlap, problematic drug use is less concretely operationa-
lized than the terms ‘‘drug dependence’’ and ‘‘drug abuse’’
or ‘‘harmful use’’ with respect to severity and duration of
symptom load. In fact, the UNDCP data have been often
criticized for using rather vague operationalizations, and one
of the answers to this critique has been that on a global
scale, the use of stringent criteria such as ICD-10 criteria is
impossible because of the lack of epidemiologic data. This,
however, may change globally with the introduction of
standardized surveys in many countries (e.g. the World
Mental Health Survey, cf. World Mental Health Survey
Consortium, 2004), and in fact it has changed already for the
European Union (EU) countries and Norway where several
adequate general population surveys are available. We,
therefore, can use prevalence estimates of ICD-10 and
DSM-IV DUDs from population surveys as part of our
review.
For this review, we restricted ourselves to studies where
the fieldwork was conducted after 1990 for two reasons.
First, DUD or problematic drug use prevalence rates tend
to fluctuate rather quickly; so quickly that sometimes even
terms like ‘‘epidemic’’ are used for describing the rapid
increase and later decrease of rates (e.g., Agar and
Reisinger, 2002; see also Hartman and Golub, 1999).
Second, the concept of DUD, the diagnostic criteria and
the assessment instruments have changed substantially
over the past 25 years (e.g. Room, 1998). In the past,
many different non-converging definitions were used in
different classification systems. In contrast, the newest
definitions of drug dependence (DD) in ICD-10 (WHO,
1993) and DSM-IV (1994; see American Psychiatric
Association, 2000) converge and have been shown to be
relatively stable across standard assessments and cultures,
with the exception of cannabis dependence (e.g. Üstün et
al., 1997; Compton et al., 1996; Rounsaville et al., 1993).
Unfortunately, the definitions (see Rehm et al., 2005) for
harmful use according to ICD-10, and for drug abuse (DA)
according to DSM-IV are less stable and comparable
across cultures and instruments (e.g. Üstün et al., 1997;
Rounsaville et al., 1993). In the following review, we will
thus distinguish between dependence and harmful use/
abuse.
DSM-IIIR, DSM-IV and ICD-10 diagnoses of DUD will
generally lead to similar prevalence rates (e.g. Rounsaville
et al., 1993; Pull et al., 1997; see also Table 2 below). We
therefore accepted papers from DSM-IIIR onwards. In terms
of diagnostic instruments, the most common was the
Composite International Diagnostic Interview (CIDI). The
CIDI has been shown to generate data very similar to those
obtained by the Schedules for Clinical Assessment in
Neuropsychiatry (SCAN) in the WHO/NIH cross-national
reliability and validity study (e.g. Üstün et al., 1997; Pull et
al., 1997). We decided to include only studies using these
instruments.
Assessment instruments like the CIDI are typically used
in household surveys, e.g. in the current World Mental
Health Survey (World Mental Health Survey Consortium,
2004). Based on the nature of illegal behaviour, household
surveys will not reach many illegal drug users for various
reasons. Most importantly, many users are not living in
households but are institutionalized (e.g., in inpatient
treatment or in prison) or homeless. To give one example
from a recent study in Sweden: of the problematic heavy
drug users in Stockholm in contact with the legal or
treatment system, only 46% had a fixed living place
(Olsson et al., 2001). In addition, for people living in their
own homes and not being in contact with either the
treatment or legal system, it may be difficult to admit to a
problematic use of illegal drugs for reasons of social
pressure.
To reach this so-called hidden populations, indirect
estimation methods are used based on treatment records,
policy and legal data, back calculation from mortality and/
or HIV/AIDS data, or the combination of multiple data
sources in the capture/recapture method (see Kraus et al.,
2003, for short descriptions; see also EMCDDA, 2003).
 J. Rehm et al. / European Neuropsychopharmacology 15 (2005) 389 – 397
Basically, the multiplier methods use knowledge from past
studies about the proportion of problematic drug users,
who in a given time period were in treatment, or in contact
with the legal system, or who had died, to back estimate
the number of all problematic drug users. Capture –
recapture methods use the overlap of people in different
systems (e.g. in the treatment and legal system) to estimate
the part of problematic drug users not captured by the
different data sources (hidden population). We will show
the results of such estimates, as undertaken within the
guidelines of European Monitoring Centre for Drugs and
Drug Addiction (EMCDDA; cf. Kraus et al., 1999;
EMCDDA, 2003) for sensitivity analyses and as basis
for discussion. As indirect methods use social statistics,
which with the exception of treatment statistics are not
based on medical categories, the case definitions in such
studies differ from the medical definitions such as ICD-10
or DSM-IV. For instance, problematic drug use as defined
by the EMCDDA refers to ‘‘injecting drug use (IDU) or
long duration/regular use of opioids, cocaine and/or
amphetamines, excluding ecstasy and cannabis use’’.
However, the definition of problematic drug use depends
on the choice of the data system and the estimate may
represent injectors, problematic opioid users, or problem-
atic drug users as the broadest category. The police, for
example, records individuals caught in possession of
illegal drugs, but not all offenders might show symptoms
of harmful use or dependence, and the estimate derived
from these data may include different drugs and to some
extent even non-addicted dealers.
In sum, the present report will use different estimates to
answer the following questions: What is the prevalence of
DUD in the EU and Norway based on a review of direct
estimates using general population surveys? What is the
prevalence of problematic drug use based on indirect
estimates using multiplier and capture –recapture techni-
ques? What is the meaning of the burden of disease data
after evaluation these questions?
2. Methods
The following steps were undertaken in this review for
the detection of general population surveys (see also
Wittchen and Jacobi, 2005):
& Systematic computer-assisted search in Medline for
‘‘drug dependence’’, ‘‘drug problems’’, ‘‘drug abuse’’,
‘‘drug use disorders’’, and the name of the relevant
countries or Europe. Criteria for inclusion were:
indication of a sex-specific prevalence rate for DUD;
publication in English, French, German, Swedish, or
Spanish; field work in 1990 and later; a representative
general population or primary care visitors sample of an
EU country (Austria, Belgium, Cyprus, Czech Repub-
lic, Denmark, Estonia, Finland, France, Germany,
391
Greece, Hungary, Ireland, Italy, Latvia, Lithuania,
Luxembourg, Malta, Netherlands, Poland, Portugal,
Slovakia, Slovenia, Spain, Sweden, United Kingdom)
or Norway; and an assessment of DUD with either
SCAN or CIDI.
& Check with experts about the completeness of the
literature and inclusion of studies published in other
languages. The following experts were included: H.
Katschnig, Austria; J. Mendlewicz, Belgium; E. Drago-
mirecka, Czech Republic; P. Munk-Jørgensen, Denmark;
J. Lönnqvist, Finland; J. Lèpine, France; M. Kopp,
Hungary; Z. Rihmer, Hungary; C. Faravelli, Italy; C.
Pull, Luxembourg; R. de Graaf, Netherlands; I. Sandan-
ger, Norway; T. Sorensen, Norway; M. Xavier, Portugal;
J. Alonso, Spain; C. Allgulander, Sweden; J. Storbjörk,
Sweden; T. Brugha, United Kingdom.
& Extraction of information about each article according
to the following categories: country and region of the
study; type of prevalence (i.e., lifetime prevalence, 12-
month prevalence, 1-month prevalence); diagnostic
category (drug dependence, drug abuse/misuse, both);
diagnostic system (ICD10; DSM-IV, DSM-IIIR); as-
sessment instrument (CIDI, SCAN); sample informa-
tion (N males, N females, age range); year of
fieldwork; full citation and influencing factors on
prevalence rate.
& Analysis of the extracted information for similarity
between countries and diagnostic systems, influence of
socio-demographic variables and sampling.
The results from the indirect methods were taken from
the systematic effort of EMCDDA to collect such estimates
(EMCDDA, 2003; Kraus et al., 2003).
3. Results
Table 1 gives the prevalence estimates from the general
population surveys. Overall, the rates vary between less than
one (0.3%) and three (2.9%) percent for 12-month drug
dependence, with similar and sometimes higher numbers for
drug abuse/harmful use. The ratio between males and
females shows a higher prevalence for males in most
surveys; however, the differences are not as pronounced as
we would expect from treatment or untreated street user
samples.
The overwhelming majority of the cases in general
population surveys tend to be people with cannabis use
disorders (CUD). In fact, the prevalence rates in studies for
CUD alone do not differ substantially from the rates based
on all DUD; but this statement is based on few studies (see
Table 1, e.g. Perkonigg et al., 1999 for a study on CUD).
Even in comparisons combining all available data for a
country, higher CUD prevalence rates were estimated than
other DUD prevalence rates. For instance: in Germany for
the period 2000/2001 180,000 people with drug depen-
392 J. Rehm et al. / European Neuropsychopharmacology 15 (2005) 389 – 397

Table 1
Prevalence of DUD in EU countries and Norway 1990 – 2004 as measured with CIDI or SCAN
Country Prevalence in %a Diagnosis and Sample and Field work Reference
(age group with highest assessment population
prevalence and rate)
Belgium, France, males: 0.4 SD substance 12-month-prevalence, 9953 males, 2001 – 2003 ESEMeD/MHEDEA, 2000
Germany, Italy, dependence (including DSM-IV, CIDI 11 472 females, investigators, 2004
Netherlands, alcohol), 1.3 SA substance general population,
and Spainb abuse (including alcohol); 18 years and older
females: 0.1 SD, 0.2 SA
(highest prevalence:
18 – 24 years old: 2.2% c)
males: 1.8 SD, 7.4 SA; lifetime prevalence,
females: 0.4 SD, 1.0 SA DSM-IV, CIDI
Finland Helsinki and males: 2.6 DA (cannabis); 1-month-prevalence, 233 males, 1995 Aalto-Setälä et al., 2001
Jyväskylä regions females: 2.9 DA (cannabis) DSM-IV, SCAN 414 females,
general population,
20 – 24 years old
Germanyd males: 0.6 DD and DA; 1-month-prevalence, 1913 males, 1997 – 1999 Jacobi et al., 2002, 2004;
females: 0.4 DD and DA DSM-IV, CIDI 2268 females, Wittchen et al., 2000
males: 0.5 DD; females: 12-month-prevalence, general population,
0.2 DD (DD males 18 – 34 DSM-IV, CIDI 18 – 65 years old
years old, 1.2%; DD
females: 35 – 49 years old,
0.4%); males: 1.0 DD
and DA; females:
0.5 DD and DA
males: 2.3 DD and DA; lifetime prevalence,
females: 1.9 DD and DA DSM-IV, CIDI
Germany males: 0.9 DD, 0.5 DA; 12-month-prevalence, 3026 males, 2000 Kraus and Augustin, 2001
females: 0.4 DD, 0.1 DA DSM-IV, CIDI 3606 females,
general population,
18 – 59 years old
Germany males: 1.8 DD, 2.7 DA; 12-month-prevalence, 1493 males, 1995 Wittchen et al., 1998;
Munich region females: 0.7 DD, 0.8 DA DSM-IV, CIDI 1528 females, Perkonigg et al., 1998a,b
males: 2.5 DD, 4.1 DA; lifetime prevalence, general population,
females: 1.6 DD, 1.8 DA DSM-IV, CIDI 14 – 24 years old
Germany males: 0.3 DD, 3.6 DA; 12-month-prevalence, 714 males, 1996/97 Perkonigg et al., 1998a,b,
Munich region females: 0.7 DD, 1.7 DA DSM-IV, CIDI 681 females, 1999
(cannabis use disorders only) general population,
14 – 19 years old
Netherlands males: 0.9 DD, 0.3 DA; 1-month-prevalence, 3304 males, 1996 Bijl et al., 1998a,b; 2002
females: 0.5 DD, 0.2 DA DSM-IIIR, CIDI 3773 females,
males: 1.0 DD, 0.6 DA; 12-month-prevalence, general population,
females: 0.7 DD, 0.3 DA DSM-IIIR, CIDI 18 – 64 years old
(DD males: 18 – 34 years old,
2.0%; DD females 18 – 34 years
old, 0.7%; incidence for DD
and DA highest in age group
18 – 24 years old)
males: 2.1 DD, 2.0 DA; lifetime prevalence,
females: 1.5 DD, 1.1 DA DSM-IIIR, CIDI
Norway males: 1.0 DD, 0.1 DA; 12-month-prevalence, 928 males, 1994 – 1997 Kringlen et al., 2001
females: 0.3 DD, 0.3 DA DSM-IIIR, CIDI 1138 females,
(highest prevalence for general population,
substance use disorders in 18 – 65 years old
age group 30 – 39 years old;
no rates given)
males: 2.6 DD, 1.7 DA; lifetime prevalence,
females: 1.3 DD, 1.3 DA DSM-IIIR, CIDI
(highest prevalence for
substance use disorders in
age group 30 – 39 years old;
no rates given)
 J. Rehm et al. / European Neuropsychopharmacology 15 (2005) 389 – 397 393

Table 1 (continued)
Country Prevalence in %a Diagnosis and Sample and Field work Reference
(age group with highest assessment population
prevalence and rate)
Sweden Less than 10 cases 12-month-prevalence, 1710 males and 2002 Directly calculated by the
Stockholm region ICD-10, CIDI 1846 females, authors from the raw data of
general population, the Stockholm county alcohol
18 years and older and drug treatment study
http://www.stakes.fi/nat/nat03/
2/roomeng.htm
United Kingdom males: 2.9 DD; females: 1.5 DD 12 month prevalence, 4859 males, 1993 Jenkins et al., 1997
ICD-10, CIS-R 4933 females,
general population,
16 – 64 years old
United Kingdom males and females: 0.3 DD 12-month-prevalence, 123 males, 1993 – 1994 McConnell et al., 2002
Derry region ICD-10, SCAN 184 females,
males and females: 0.3 DD 1-month-prevalence, general population,
ICD-10, SCAN 18 – 64 years old
a
DD: drug dependence; DA: drug abuse (if DSM was used as diagnostic instrument), or harmful use of illicit drugs (if ICD-10 was used).
b
In another publication of the study group, the following country-specific 12-month prevalence rates for SUD in total were given: Belgium 1.2%; France
0.7%; Germany 1.1%; Italy 0.1%; Netherlands 3.0%; Spain 0.3%; SUD include alcohol use disorders.
c
No gender-specific prevalence rates were given.
d
Any illicit substance use includes non-prescribed use of psychotropic medication.

Table 2
National prevalence estimates of problematic drug use in EU countries and Norway (rates of problematic drug use per 1000 inhabitants aged 15 – 64; cf. Kraus
et al., 2003, p. 481)
Country Total population Problem opioid/problem Problem drug use (Current)
(age 15 – 64 in 1998) opioid or cocaine/problem injecting
opoid or amphetamines
Austria 5,444,856 3.2 (CI: 2.9 – 3.4) (5)g 3.2 (2.2 – 4.2) (3)c
Belgium 6,706,370
Denmark 3,551,204 3.9 (3.6 – 4.3) (5)f 4.3 (3)c
Finland 3,433,511 3.4 (3.1 – 3.9) (5)g 3.1 (2.0 – 4.1) (3)c
3.6 (3.2 – 4.1) (5)h
France 38,039,603 4.7 (1)
3.9 (2)a
4.7 (6)
Germany 55,992,553 3.2 (3.0 – 3.5) (1) 2.6 (2.3 – 3.0) (3)c
3.1 (2.7 – 3.4) (2)b
Ireland 2,436.798 2.6 (5)f 5.6 (5)g 1.9 (3)d
3.2 (3)d,e
Italy 39,127,434 7.1 (1)
7.2 (2)b
7.7 (6)
Luxembourg 284,005 9.2 (2)a
8.3 (7.8 – 8.7) (2)b 4.8 (4.7 – 4.9) (3)c
7.3 (5.6 – 9.0) (5)f 7.5 (7.4 – 7.6) (3)c,e
8.2 (5.3 – 13.8) (5)g
The Netherlands 10.661.543 2.6 (2.4 – 2.8) (1) 2.8 (2.4 – 3.2) (2)a
Norway 2,852,913 4.3 (3.7 – 4.9) (3)d
Portugal 6,760,830 6.7 (6.2 – 7.2) (1) 4.1 (2.7 – 5.5) (3)c
7.8 (7.4 – 8.3) (2)a
Spain 26,923,202 6.6 (1) 3.1 (3)d
Sweden 5,650,799 4.6 (4.3 – 5.0) (5)i
United Kingdom 38,446,154 6.3 (1) 7.0 (6) 4.2 (3)c
(1) Treatment multiplier. (2) Police multiplier: prevalence approach; incidence approach. (3) Mortality multiplier: all-cause mortality; doverdose;
a b c
e
adding cases in treatment. (4) HIV/AIDS multiplier. (5) Capture – recapture: ftwo-sample; gthree-sample; hfour-sample; ifive-sample. (6) Multivariate
indicator.
394 J. Rehm et al. / European Neuropsychopharmacology 15 (2005) 389 – 397
dence other than cannabis are estimated, in addition to
240,000 people with cannabis dependence (Bühringer et al.,
2002). But, as indicated above, the ratio in general
population surveys is usually highly skewed towards
cannabis dependence because of the hidden population
problem. This also explains the relatively equal rates of
males and females in general population surveys, as
cannabis use is relatively equally distributed in various
countries (see Hibell et al., 2000, for an overview of school
surveys; Ministry of Public Health of Belgium, 2002, for
studies in adults).
There are few incidence studies based on general
population surveys. The Dutch NEMESIS study found
12-month incidence rates for drug dependence of 0.32%
for females, and 0.21% for males, and respective rates
for drug abuse of 0.07% for females and 0.48% for
males. Thus, the incidence rate ratio between genders
was reversed from one diagnosis to the other (Bijl et al.,
2002). In terms of care, people with DUD, together with
people with alcohol use disorders, were the least likely
among people with psychiatrically defined problems to
enlist professional care (Bijl and Ravelli, 2000; Jacobi et
al., 2004). There are, however, specialized non-psychiat-
ric treatment systems in various countries, partly com-
pletely outside of the medical treatment system
(Klingemann and Hunt, 1998). Also, the comorbidity
rates for mental diagnoses are slightly lower in these
general population samples, as compared for instance to
the comorbidity rates for mood and anxiety disorders
(Jacobi et al., 2004).
The prevalence of problematic use of drugs other than
cannabis in Europe is considerably higher when based on
other estimation methods, such as multiplier procedures
and capture –recapture strategies (Kraus et al., 2003; see
Table 2).
The prevalence of problematic opioid use in various
combinations is quite similar in different European
countries, despite different databases (i.e. treatment statis-
tics, legal statistics, morbidity and mortality statistics),
estimation methods as well as different drug policies: it
varies between 0.26 and 0.92 percent. Similarly, the
prevalence rates of injection drug use are fairly similar
between 0.19% and 0.48%, when estimated without
including treatment data (see Table 2). With the exception
of Scandinavian countries where amphetamine use is most
prevalent, opioids are the main drug class that leads to
physical, psychological, social or legal harm in all other
EU countries. Not only are the prevalence rates of
problematic drug use in EU countries and Norway much
smaller than the rates of alcohol use disorders (see Rehm
et al., 2005), but also the variance of problematic drug use
is much smaller. This small variance persists, although
drug policies in these countries differ widely, but no effect
was found for different orientations in terms of more
Fstrict_ or more Fliberal_ policies (Reuband, 1995). For
instance, countries at either end of the continuum, such as
Germany and the Netherlands appear to have similar
prevalence rates. However, more specific measures, such
as regulations concerning substitution therapy or harm-
reduction approaches have not yet systematically been
investigated. Thus, although the age structure of the
population, the rate of poverty and unemployment
(Fergusson et al., 2001), social and political reactions to
drug use (Pacula and Chaloupka, 2001), or patterns of
drug use administration (Gossop et al., 1996) have all been
found to be related to problematic drug use, there is yet no
clear evidence of consistent effects on national prevalence
rates.
4. Discussion
General population surveys seem to underestimate drug
use disorders (DUD) of so-called hard drugs, i.e. opioids,
cocaine, and amphetamines, but they can in generally be
used to estimate cannabis use disorders (CUD), as
respondents seem to be willing to admit to cannabis use,
and as such surveys reach the majority of cannabis users
in European populations. However, we have no clear idea
about the validity of such estimates, as for several reasons.
First, the measurement of CUD is fraught with measure-
ment error (e.g. Üstün et al., 1997; Compton et al., 1996;
Rounsaville et al., 1993). Second, most surveys do not
report cannabis separately, but only DUD as a general
category, from which CUD can only be estimated with
problems. Third, there is a lack of validity studies. It can
also be suspected that the validity of prevalence estimates
for CUD may depend on the legal status of cannabis in
the culture (i.e. is cannabis use tolerated by law en-
forcement) as well as on the question whether CUD is the
only or main substance of dependence/abuse. However,
with CUD, the problematic of hidden populations seems
less evident as with drugs such as opioids, cocaine, or
amphetamines.
In any case, to establish better comparability of
prevalence rates between countries, prevalence of CUD
should be reported separately from other drugs. Given the
increases in cannabis use since the beginning of the 1990s
(Hibell et al., 2000), there may be increases in CUD in
current or future surveys, given the lag time between first
use and developing a treatment need (Hartnoll, 2003).
Separate listing is also suggested with respect to non-
prescribed use of psychotropic medication, for similar
reasons. And of course, prevalence rates for disorders of
all substances used combined such as reported in the
European Study of the Epidemiology of Mental Disorders
(ESEMeD/MHEDEA, 2000 investigators, 2004) are very
hard to interpret and thus make use of in health care
planning.
Finally, multiplicator methods do not make that much
sense in estimating CUD. With respect to treatment, there
seem to be no standards, and treatment rates depend to a
 J. Rehm et al. / European Neuropsychopharmacology 15 (2005) 389 – 397
large degree on cultural norms (e.g. when do parents
consider use grave enough to send their child to therapy?)
including the way school systems handle cannabis use. With
respect to involvement with the legal system, toleration and
enforcement practices of cannabis use vary widely between
and even within countries. And with respect to back
calculating from morbidity and mortality, the overmortality
of people with cannabis is not well established and there are
also no clear disease markers to back calculate. Thus, we are
left with general population surveys as the only current
practical methodology to estimate CUD. This being the
case, with respect to CUD, more efforts should be
undertaken to achieve a valid assessment. Although most
studies in the early or mid-90s found problematic validity
and reliability, and although CUD are the most prevalent of
all illicit drug disorders in most societies, no other studies
have been conducted to derive better instruments to assess
CUD.
With respect to estimating prevalence rates of opioid,
cocaine and/or amphetamine use disorders, indirect methods
seem to yield a more realistic picture, as they include hidden
populations. Also, the estimates under the umbrella of the
European Monitoring Centre for Drugs and Drug Addiction
are surprisingly similar for different EU countries and
Norway, independently if they were based on treatment
statistics, legal statistics, or were derived from morbidity
and mortality data. However, estimates derived from
indirect methods are limited with respect to the target
population and validity: The defined target groups do not
directly correspond to the DSM-IV or ICD-10 definitions of
harmful use/abuse or dependence, and the available data are
by no means without measurement error. Most secondary
data systems are biased by systematic measurement errors,
such as imprecise definitions, incompleteness or flaws in
data entry. As a consequence, confidence intervals seem to
be meaningless and cross-national comparisons are limited.
Clearly the estimated target population depend to a large
extent on the definitions of questions such as ‘‘Who gets
into treatment?’’, ‘‘Who is picked up by police?’’, or ‘‘Who
dies as a consequence of his or her drug use?’’. Overall,
while indirect methods give a first approximation to the
extent of the drug problem in Europe, they have to be
refined and further developed to serve as a valid basis for
interventions.
With regard to estimating drug-related consequences,
such as burden of disease, three points should be made.
Firstly, with illicit drugs, social and legal consequences
by far outweigh health consequences as illustrated by
several cost studies (Rehm et al., 2003, 2004). Thus, the
public health importance of illicit drugs would be
underestimated if judged only by the burden of disease.
Secondly, the current global estimates within the CRA
(WHO, 2002) seem to underestimate the burden for
Europe because it assumes prevalence rates that are too
low for problematic opioid, cocaine and amphetamine
use. And thirdly, drug-related harmful consequences to a
395
large degree do not stem from the chemical properties of
opioids and other substances used illegally, but are
related to conditions and contexts of their use. Typically,
the compromised health status of users, contaminated
drugs and drug equipment, dependence on black markets
to obtain drugs, and a lack of access to adequate
prevention and care have been shown to primarily
contribute to the harms and costs associated with illicit
drug use. In turn, adequate interventions, like access to
clean syringes and injecting equipment, and state-of-the-
art treatment have been shown to positively influence the
degree of harm associated with drug use (Ling et al.,
2003; Vlahov et al., 2001; Ward et al., 1999). In future,
such interventions can and should be informed by
epidemiology to a larger extent (Musto and Sloboda,
2003). The methods described above can be considered
as an important step in this direction.
References
Aalto-Setälä, T., Marttunen, M., Tuulio-Henriksson, Poikolainen,
K., Lonnqvist, J., 2001. One-month prevalence of depression
and other DSM-IV disorders among young adults. Psychol.
Med. 31, 791 – 801.
Agar, M., Reisinger, H.S., 2002. A heroin epidemic at the intersection of
histories: the 1960s epidemic among African Americans in Baltimore.
Med. Anthropol. 21 (2), 115 – 156.
American Psychiatric Association, 2000. Diagnostic and Statistical Manual
of Mental Disorders, Text Revision [DSM-IV-TR] (Vol. 4th edition).
American Psychiatric Association, Washington, DC.
Bijl, R.V., Ravelli, A., 2000. Psychiatric morbidity, service use, and
need for care in the general population: results of The Netherlands
mental health survey and incidence study. Am. J. Publ. Health 90 (4),
602 – 607.
Bijl, R., van Zessen, G., Ravelli, A., 1998a. The Netherlands Mental Health
Survey and Incidence Study (NEMESIS): objectives and design. Soc.
Psychiatry Psychiatr. Epidemiol. 33, 581 – 586.
Bijl, R., Ravelli, A., van Zessen, G., 1998b. Prevalence of psychiatric
disorder in the general population: results of The Netherlands Mental
Health Survey and Incidence Study (NEMESIS). Soc. Psychiatry
Psychiatr. Epidemiol. 33, 587 – 595.
Bijl, R.V., de Graaf, R., Ravelli, A., Smit, F., Vollebergh, W.A.M., 2002.
Gender and age-specific first incidence of DSM-III-R psychiatric
disorders in the general population. Soc. Psychiatry Psychiatr.
Epidemiol. 37, 372 – 379.
Bühringer, G., Kraus, L., Augustin, R., 2002. Prävalenz des Konsums
psychoaktiver Substanzen und Zahl der jährlichen Behandlungsfälle in
Deutschland. http://www.ift.de/IFT_deut/Daten/praevalenz.htm (last
assessed 22.01.04).
Compton, W.M., Cottler, L.B., Dorsey, K.B., Spitznagel, L., Mager, D.E.,
1996. Comparing assessments of DSM-IV substance dependence
disorders using CIDI-SAM and SCAN. Drug Alcohol Depend. 41,
179 – 187.
Degenhardt, L., Hall, W., Warner-Smith, M., Lynskey, M., 2001. Compar-
ative Risk Assessment: Illicit Drug Use. Report to WHO from the
National Drug and Alcohol Research Centre. University of New South
Wales, Australia.
EMCDDA, 2003. National prevalence estimates of problem drug use in the
European Union, 1995 – 2000. Final Report CT.00.RTX.23. Lisbon:
European Monitoring Centre for Drugs and Drug Addiction.
ESEMeD/MHEDEA, 2000 investigators, 2004. 12-Month comorbidity
patterns and associated factors in Europe: results from the European
396 J. Rehm et al. / European Neuropsychopharmacology 15 (2005) 389 – 397
Study of the Epidemiology of Mental Disorders (ESEMeD) project.
Acta Psychiatr. Scand. 109, 28 – 37.
Ezzati, M., Lopez, A.D., Rodgers, A., Vander Horn, S., Murray, C.J.L.,
2002. Comparative risk assessment collaborating group. Selected major
risk factors and global and regional burden of disease. Lancet 360,
1347 – 1360.
Fergusson, D.M., Horwood, L.J., Woodward, L.J., 2001. Unemployment
and psychosocial adjustment in young adults: causation or selection?
Soc. Sci. Med. 53, 305 – 320.
Gossop, M., Griffiths, P., Powis, B., Williamson, S., Strang, J.,
1996. Frequency of non-fatal heroin overdose: survey of heroin
users recruited in non-clinical settings. Br. Med. J. 313, 402.
Hall, W., Teesson, M., Lynskey, M., Degenhardt, L., 1999. The 12-month
prevalence of substance use and ICD-10 substance use disorders in
Australian adults: findings from the national survey of mental health
and well-being. Addiction 94, 1541 – 1550.
Hartman, D.M., Golub, A., 1999. The social construction of the crack
epidemic in the print media. J. Psychoact. Drugs 31 (4), 423 – 433.
Hartnoll, R.L., 2003. Drug epidemiology in the European institutions:
historical background and key indicators. UNODC Bull. Narc. 55,
53 – 71.
Hibell, B., Andersson, B., Ahlström, S., Balakireva, O., Thoroddur, B.,
Kokkevi, A., Morgan, M., 2000. The 1999 ESPAD Report (The
European School Survey Project on Alcohol and Other Drugs). Alcohol
and Other Drug Use Among Students in 30 European Countries.
Stockholm: The Swedish Council for Information on Alcohol and Other
Drugs (CAN).
Jacobi, F., Wittchen, H., Holting, C., Sommer, S., Lieb, R., et al., 2002.
Estimating the prevalence of mental and somatic disorders in the
community: aims and methods of the German National Health
Interview and Examination Survey. Int. J. Methods Psychiatr. Res. 11,
1 – 18.
Jacobi, F., Wittchen, H., Holting, C., Hofler, M., Pfister, H., et al.,
2004. Prevalence, co-morbidity and correlates of mental disorders
in the general population: results from the German Health
Interview and Examination Survey (GHS). Psychol. Med. 34, 1 – 15.
Jenkins, R., Lewis, G., Bebbington, P., Brugha, T., Farrell, M., Gill, B.,
et al., 1997. The National Psychiatric Morbidity Surveys of Great
Britain: initial findings from the Household Survey. Psychol. Med.
27, 775 – 789.
Klingemann, H., Hunt, G., 1998. Drug Treatment Systems in an
International Perspective: Drugs, Demons, and Delinquents. Sage,
London.
Kraus, L., Augustin, R., 2001. Repräsentativerhebung zum Gebrauch
psychoaktiver Substanzen bei Erwachsenen in Deutschland 2000. Sucht
47, S3 – S86.
Kraus, L., Kümmler, P., Augustin, R., Hartnoll, R, Wiessing, L.,
1999. Methodological guidelines to estimate the prevalence of
problem drug use on the national level. Lisbon: European
Monitoring Centre for Drugs and Drug Addictions. http://www.
emcdda.org/multimedia/project _ reports/situation/guidelines _ national _
pdu.pdf.
Kraus, L., Augustin, R., Frischer, M., Kümmler, P., Uhl, A., Wiessing, L.,
2003. Estimating prevalence of problem drug use at national level in
countries of the European Union and Norway. Addiction 98, 461 – 485.
Kringlen, E., Torgersen, S., Cramer, V., 2001. A Norwegian psychiatric
epidemiological study. Am. J. Psychiatry 158, 1091 – 1098.
Ling, W., Rawson, R.A., Compton, M., 2003. Clinical treatment of
opioid addiction and dependence. Methods Mol. Med. 84, 285 – 295.
McConnell, P., Bebbington, P., McClelland, R., Gillespie, K., Houghton, S.,
2002. Prevalence of psychiatric disorder and the need for psychiatric
care in Northern Ireland: population study in the district of Derry. Br. J.
Psychiatry 181, 214 – 219.
Ministry of Public Health of Belgium, 2002. Cannabis 2002 report. A joint
international effort at the initiative of the Ministers of Public Health of
Belgium, France, Germany, the Netherlands, Switzerland. Brussels:
Ministry of Public Health.
Musto, D.F., Sloboda, Z., 2003. The influence of epidemiology on drug
control policy. UNODC Bull. Narc. 55, 9 – 22.
Olsson, B., Adamsson Wahren, C., Byqvist, S., 2001. Det tunga
narkotikamissbrukets omfattning i Sverige 1998. Rapport, vol. 61.
CAN, Stockholm.
Pacula, R.L., Chaloupka, F.J., 2001. The effects of macro-level interven-
tions on addictive behavior. Subst. Use Misuse 36, 1901 – 1922.
Perkonigg, A., Lieb, R., Wittchen, H.-U., 1998a. Substance use, abuse and
dependence in Germany: a review of selected epidemiological data. Eur.
Addict. Res. 4 (1 – 2), 8 – 17.
Perkonigg, A., Lieb, R., Wittchen, H.-U., 1998b. Prevalence of use,
abuse and dependence of illicit drugs among adolescents and young
adults in a community sample. Eur. Addict. Res. 4 (1 – 2), 58 – 66.
Perkonigg, A., Lieb, R., Höfler, M., Schuster, P., Sonntag, H., Wittchen,
H.U., 1999. Patterns of cannabis use, abuse and dependence over time:
incidence, progression and stability in a sample of 1228 adolescents.
Addiction 94, 1663 – 1678.
Pull, C.B., Saunders, J.B., Mavreas, V., Cottler, L.B., Grant, B.F., et al.,
1997. Concordance between ICD-10 alcohol and drug use disorder
criteria and diagnoses as measured by the AUDADIS-ADR, CIDI and
SCAN: results of a cross-national study. Drug Alcohol Depend. 47,
207 – 216.
Rehm, J., Fischer, B., Frick, U., 2003. ‘‘Public Health’’ und ökonomische
Studien im Bereich Sucht. Suchttherapie 4, 1.
Rehm, J., Frick, U., Fischer, B., 2004. Zur Ökonomie des
Suchtmittelgebrauchs. In: Tretter, F., Erbas, B., Sonntag, G. (Eds.),
Oekonomie der Sucht und Suchttherapie. Pabst, Lengerich, Berlin, pp.
233 – 244.
Rehm, J., Room, R., van den Brink, W., Jacobi, F., 2005. Alcohol use
disorders in EU countries and Norway: an overview of the epidemiology.
Eur. Neuropsychopharmacol. 15, 377 – 388.
Reuband, K.-H., 1995. Drug use and drug policy in Western Europe.
Epidemiological findings in a comparative perspective. Eur. Addict.
Res. 1, 32 – 41.
Room, R., 1998. Alcohol and drug disorders in the international
classification of diseases: a shifting kaleidoscope. Drug Alcohol Rev.
17, 305 – 317.
Rounsaville, B.J., Kendall, B., Babor, T., Kranzler, H., Kadden, R., 1993.
Cross system agreement for substance use disorders: DSM-III-R, DSM-
IV and ICD-10. Addiction 88, 337 – 348.
UNDCP, 2000. Global illicit drug trends 2000. United Nations Drug
Control Programme, Vienna.
Üstün, B.T., Compton, W., Mager, D., Babor, T., Baiyewu, O., et al., 1997.
WHO study on the reliability and validity of the alcohol and drug use
disorder instruments. Overview of methods and results. Drug Alcohol
Depend. 47, 161 – 169.
Vlahov, D., Des Jarlais, D., Goosby, E., Hollinger, P., Lurie, P., Shriver, M.,
Strathdee, S., 2001. Needle exchange programs for the prevention of
human immunodeficiency virus infection: epidemiology and policy.
Am. J. Epidemiol. 154, S70 – S77.
Ward, J., Hall, W., Mattick, R., 1999. Role of methadone maintenance in
opioid dependence. Lancet 353, 221 – 226.
WHO, 1993. The ICD-10 Classification of Mental and Behavioural
Disorders: Diagnostic Criteria for Research. World Health Organization,
Geneva.
WHO, 2000. Health Systems: Improving Performance. World Health
Organization, Geneva.
WHO, 2002. World Health Report 2002: Reducing Risks, Promoting
Healthy Life. World Health Organization, Geneva.
Wittchen, H.-U., Jacobi, F., 2005. Size and Burden of mental disorders in
Europe: A critical review and appraisal of 27 studies. Eur. Neuropsy-
chopharmacol. 15, 357 – 376.
Wittchen, H., Nelson, C., Lachner, G., 1998. Prevalence of mental disorders
and psychosocial impairments in adolescents and young adults.
Psychol. Med. 28, 109 – 126.
Wittchen, H.-U., Pfister, H., Schmidtkunz, B., Winter, S., Müller, N.,
2000. German National Health Interview and Examination Survey—
 J. Rehm et al. / European Neuropsychopharmacology 15 (2005) 389 – 397 397

Mental Health Supplement (GHS—MHS), Part 2: Tables and
instructions for public use file (supplement to final report BMBFBW
01 EH 9701/8). Munich: Max-Planck-Institute for Psychiatry, Clinical
Psychology and Epidemiology. [The public use file can be ordered at:
jacobi@psychologie.tu-dresden.de].
World Mental Health Survey Consortium, 2004. Prevalence, severity
and unmet need for treatment of mental disorders in the world
health organization mental health surveys. J. Am. Med. Assoc. 291,
2581 – 2590.