E-Mail karger@karger.com
Research Report
Eur Addict Res 2017;23:163–169
DOI: 10.1159/000477802
Measuring Concurrent Polydrug Use in
General Populations: A Critical Assessment
Karoliina Karjalainen Kristiina Kuussaari Kati Kataja Christoffer Tigerstedt
Pekka Hakkarainen
National Institute for Health and Welfare, Helsinki, Finland
ment used, as does the course of the prevalence trends. The
concept of simultaneous polydrug use may capture the phenomenon
better compared to the concept of CPU.
© 2017 S. Karger AG, Basel
Introduction
Polydrug use gained a foothold and growing recognition
among researchers several decades ago [1, 2], and
currently there is an increasing body of research literature
in this regard [3]. According to the European Monitoring
Centre for Drugs and Drug Addiction (EMCDDA), polydrug
use has become a dominant pattern of drug use in
Europe [4] and it has been brought up as a challenge in
the formulation of the EU Drugs Strategy (2013–2020)
[5]. Polydrug use is noted especially among problem drug
users [6, 7], but it needs to be recognized among the general
population too due to its serious health and social
consequences [8–10]. In the present literature, it is often
claimed that polydrug use has increased during the last
decades [11–13]. However, as far as the authors know,
only a few studies have presented research evidence about
prevalence trends of polydrug use among the general
population [14–16].
Polydrug use refers to the use of 2 or more different
substances. Although it has been shown that polydrug use
Keywords
Polydrug use · Prevalence · Trends · General population ·
Finland
Abstract
Background/Aims: Polydrug use is a complicated phenomenon
that is measured in a wide variety of different ways.
Using Finland as an example, we aimed to demonstrate how
the prevalence and prevalence trends of concurrent polydrug
use (CPU) varied in the general population based on
the different measurements used. Methods: Populationbased
Drug Surveys conducted every 4 years during 1998–
2014 were used. CPU was measured with different measurements:
strict, medial and loose definition of CPU, which were
based on different combinations of alcohol, illicit drugs,
pharmaceutical drugs and cigarettes used during the last 12
months/30 days. Logistic regression was used to estimate
the p values for assessing trends. Results: Depending on the
measurements used, the prevalence of CPU in 2014 varied
between 2.0 and 18.7%. Different definitions also produced
contradictory trends of CPU: there was a modest increase in
prevalence if it was measured with a medial (p < 0.001) or
strict (p = 0.054) definition, but when measured with the
loose definition (only measure that included smoking), there
was a decrease in prevalence (p < 0.001). Conclusions: The
prevalence of CPU varies greatly depending on the measureReceived:
October 26, 2016
Accepted: May 25, 2017
Published online: June 23, 2017
European
Addiction
cRe es ar h
Karoliina Karjalainen, PhD
Senior Researcher
National Institute for Health and Welfare
PO Box 30, FI–00271 Helsinki (Finland)
E-Mail karoliina.karjalainen @ thl.fi
© 2017 S. Karger AG, Basel
www.karger.com/ear
Downloadedby:
MasarykUniversity
147.251.175.227-4/26/20239:57:25AM
Karjalainen/Kuussaari/Kataja/Tigerstedt/
Hakkarainen
Eur Addict Res 2017;23:163–169
DOI: 10.1159/000477802
164
may not always be problematic [17, 18], considerable
harm can be caused by it. For example, polydrug users are
at an elevated risk for developing comorbid psychiatric
and other health conditions [10] and polydrug use is also
significantly related to social consequences [8]. Polydrug
use can be very risky due to difficulties in dosing and unpredictable
effects that substantially increase the number
of deaths caused by overdose and other detrimental consequences
of drug use [9, 19, 20].
Apart from these consequences, polydrug use is complicated
both as a phenomenon and as a term. There is no
established theory, or a unified concept of polydrug use or
an established practice to measure it. Indeed, as Schensul
et al. [17] have argued, “existing approaches to measuring
polydrug use are confusing and inconsistent.” Often it is
approached from the perspective of illicit drugs, where the
use of 2 or more illicit drugs only [21] or illicit drugs combined
with alcohol is defined as polydrug use [8]. Nonmedical
use of psychoactive pharmaceutical drugs may be
included as one of the substances constituting polydrug
use, while sometimes cigarettes can also be included in the
combination [22], especially when dealing with adolescents
[23]. However, illicit drugs are not necessarily included
at all, since the use of alcohol and prescription
drugs may form a pattern of polydrug use [24]. Moreover,
the degree of alcohol use as a part of polydrug use can vary
greatly, for example, from at least one drink during the last
year [8] to hazardous drinking [22], whereas the amounts
of illicit drugs used are rarely taken into consideration
[13].
In addition to different substances, polydrug use is
measured by the timing of ingestion. “Concurrent polydrug
use” (CPU) refers to the use of 2 or more different
substances in a given time period, such as during the last
30 days or the last 12 months. Another measure is “simultaneous
polydrug use” (SPU), which indicates the use of
different substances at the same time or in temporal proximity
[11]. Although occasional concurrent use of 2 or
more substances, for example, during a year does not
make much sense from the point of view of harms caused
by the joint effects of different substances, it is more often
used as a measure of polydrug use compared to the simultaneous
use [11]. The lack of a separate measure for SPU
in surveys may partly explain the reason for this practice.
Thus, due to different measures and definitions, the
overall impression of polydrug use among the general
population remains vague and nebulous. In order to develop
further actions, for example, for the prevention or
harm reduction of polydrug use, we should enhance our
understanding of the phenomenon [17, 25, 26] and measure
it appropriately. To confirm this, we aimed to demonstrate
how the prevalence of CPU varies among the
general population according to different measurements
by using the data of population-based, comparable Drug
Surveys conducted in Finland during 1998–2014 as an
example. To the best of our knowledge, this is the first
study to show the effect of different measures on the prevalence
of CPU by using the same data.
Methods
Data
In Finland, population-based Drug Surveys concerning drug
use and drug-related opinions and attitudes have been conducted
approximately every 4 years since 1992. There have been some
changes and modifications in the questionnaire, and therefore, in
order to ensure the comparability of the data, 1998–2014 Drug
Surveys were used in this study (Table 1). Data collection conducted
by Statistics Finland followed the same protocol each time: representative
random samples of the population were drawn from
the Finnish Population Information System, and respondents
were contacted by post. The institutionalized population, those
without permanent address, as well as the Åland Islands were excluded.
In each survey, data was collected by self-administered
anonymous postal questionnaires, which the respondents received
by mail and were asked to return in a prepaid envelope. In 2010
and 2014, responding via Internet was also possible.
The age of the target population was in the range 15–69 years.
Since 2002, younger age groups were oversampled in order to increase
the power in the age group most actively using drugs. The
size of the random samples varied between 3,250 (in 1998) and
7,000 (in 2014). The response rate for the Drug Surveys decreased
from 66% in 1998 to around 50% in 2010 and 2014. Due to a deTable
1. Basic information about the Finnish Drug Surveys conducted in 1998–2014
1998 2002 2006 2010 2014
Age of target population, years 15–69 15–69 15–69 15–69 15–69
Oversampling (age in years) No 15–34 15–34 15–39 15–39
Sample size, n 3,250 4,053 5,500 4,250 7,000
Response rate, % 66 63 55 48 50
Number of respondents, n 2,143 2,541 3,029 2,023 3,485
Downloadedby:
MasarykUniversity
147.251.175.227-4/26/20239:57:25AM
Measuring Polydrug Use Eur Addict Res 2017;23:163–169
DOI: 10.1159/000477802
165
creasing trend in response rates, a non-respondent study was conducted
in 2014. This analysis showed that the prevalence of illicit
drug use was very similar both among non-respondents and respondents
of the original survey [27].
Measurements of CPU
Similar to the content of previous literature, with CPU we refer
to the use of 2 or more different substances during the last
12 months or 30 days without knowing the time interval in more
detail. As shown in a recent review article [28], the measurements
of polydrug use vary notably, indicating that there is no established
practice of how polydrug use is measured. Therefore, in
order to demonstrate how distinct prevalence estimates different
measurements provide, 4 measures (variables) representing different
degrees of strictness were formed. These measures are
based on different time frames (last 30 days/last 12 months) and
different combinations of
– Smoking (“Do you smoke cigarettes, cigars or pipe?”),
– Alcohol use (“How often do you drink alcoholic beverages?”)
or binge drinking (“How often do you drink 6 units1 or more
on one occasion [if you are a man]/4 units or more [if you are
a woman2
])?”,
– Illicit drug use (“Have you during the last 12 months/30 days
tried or used illicit drugs?”3) and
– Non-medical use of psychoactive pharmaceutical drugs (“Have
you during the last 12 months/30 days used sedatives, anxiolytics
or painkillers non-medically?,” hereafter referred to as prescription
drug misuse).
The questions used to construct these measures are all commonly
applied in population-based drug surveys, and thus they are
commonly used to measure CPU in other studies too.
The CPU measures in this study are as follows:
1. Strict CPU: this measure includes those who during the last
30 days reported the use of 2 different illicit drugs or at least 2 of
the following – binge drinking (at least 4/6 units of alcohol per one
occasion) or illicit drug use or prescription drug misuse.
2. Medial CPU (a): this is otherwise the same as Strict CPU, but
the time frame is longer. Thus, this measure includes those who
during the last 12 months reported the use of 2 different illicit
drugs or at least 2 of the following – binge drinking (at least 4/6
units of alcohol per one occasion) or illicit drug use or prescription
drug misuse.
3. Medial CPU (b): this is otherwise the same as Medial CPU
(a), but the alcohol measure is not so stringent, since it does not
take into account the quantity of alcohol used. Thus, this measure
includes those who during the last 12 months reported the
use of 2 different illicit drugs or at least 2 of the following – alcohol
use (any amount) or illicit drug use or prescription drug mis-
use.
4. Loose CPU: this is otherwise the same as Medial CPU (b),
but it also includes smoking. Thus, this measure includes those
who during the last 12 months reported the use of 2 different illicit
drugs or at least 2 of the following – alcohol use (any amount)
or illicit drug use or prescription drug misuse or current daily
smoking.
Statistical Analysis
Frequency tables were used to describe the prevalence and
trends of CPU according to the different definitions. In order to
estimate whether the trends observed were statistically significant,
a logistic regression model was used and a p value for trend as well
as ORs and their 95% CIs were calculated. When studying the
trends of CPU during 1998–2014, either the Strict CPU, Medial
CPU (a), Medial CPU (b), or Loose CPU was used as an out-
come variable. All these were binary variables, where 0 = no CPU
and 1 = CPU by each definition. The explanatory variable was the
year of the survey and it was used as a continuous variable. p value
<0.05 was considered statistically significant.
Due to the differences in response activity and the oversampling
of younger age groups, weighting coefficients were used in
all the analyses to restore the population representation. The
weighting coefficients were calculated by Statistics Finland and
they were based on age, gender, education and level of urbanization.
SPSS Statistics software version 22 was used to analyze the
data.
Research Ethics
The study protocol was approved by the Ethical Review Board
of the National Institute for Health and Welfare.
Results
As shown in Figure 1, the prevalence of CPU varied
notably among the general Finnish population. During
1998–2014 from one fourth to one fifth of Finns were defined
as polydrug users according to the loosest measure,
whereas with Strict CPU the proportion of polydrug users
was only 1–2%. Thus, the wider the time frame the higher
the prevalence. Similarly, the substances included had
an impact on the prevalence of CPU. The measure including
current daily smoking (Loose CPU) produced the
highest prevalence and the difference with the other measures
was considerable. The amount of alcohol used,
however, did not seem to have an impact on the prevalence
(Medial CPU [a] vs. Medial CPU [b]).
Different definitions also produced different and even
contradictory trends for CPU in Finland during 1998–
2014 (Fig. 1). According to Medial CPU and Strict CPU,
CPU has somewhat increased, although the observed increase
was statistically significant only in Medial CPU (a:
from 3.7 to 6.5%, p for trend <0.001, OR 1.1, 95% CI 1.1–
1.2; b: from 4.0 to 6.7%, p for trend <0.001, OR 1.1, 95%
CI 1.1–1.2), but not in Strict CPU (from 1.2 to 2.0%, p for
1
  One unit refers to 12 g of ethanol. The respondents were instructed by several
illustrated examples how to calculate alcohol units that they had used.
2
  In 1998 and 2002, it was 6 units even for women.
3
 Due to emerging illicit drugs, the list of substances has changed over the
time of the drug survey series. In 1998, the list included hashish/marijuana,
amphetamine, heroin/morphine, cocaine/crack, LSD, ecstasy and magic
mushrooms, whereas in 2014, it covered in addition methamphetamine,
buprenorphine or methadone (other than in substitution treatment), other
opioids illegally, GHB/GBL, synthetic cathinones, synthetic cannabinoids
and mephedrone.
Downloadedby:
MasarykUniversity
147.251.175.227-4/26/20239:57:25AM
Karjalainen/Kuussaari/Kataja/Tigerstedt/
Hakkarainen
Eur Addict Res 2017;23:163–169
DOI: 10.1159/000477802
166
trend = 0.054, OR 1.1, 95% CI 1.0–1.2). However, if measured
with Loose CPU, the trend was in the opposite direction:
the prevalence seemed to be decreasing throughout
the 2000s, and the trend was statistically significant
(from 24.3 to 18.7%, p for trend <0.001, OR 0.9, 95% CI
0.88–0.93).
Discussion
Main Findings and Their Interpretation
We used a series of Finnish population-based Drug
Surveys to demonstrate the effect of different measurements
to the prevalence estimates of CPU among the general
population. It was no surprise to find out that different
measurements provided significant variation in the
prevalence figures, but the level of difference is worth noting.
More importantly, different measurements had an
impact on the course of prevalence trends.
In our study, where different measures were tested
with the same data, the prevalence of CPU was notably
spread across a wide range: according to the strictest definition
only 1–2% of Finns are polydrug users, but with
the loosest definition, almost as much as one fifth to one
fourth of the population are classified as polydrug users.
Given this, what can be concluded from the results of previous
studies where the prevalence of CPU varies between
3.3% [8] and 21% [29] among adults, between 12% [24]
and 26% [30] among students, and between 3.5% [23] and
30% [16] among adolescents? All of these describe the
prevalence of CPU among the general population, but
since different measurements have been used, the results
are barely comparable. Furthermore, it is impossible to
conclude how large or small the differences really are,
since all the results of previous studies are based on different
data. When measuring CPU, the use of 12 months’
time span, in particular, may easily overestimate the phenomenon,
especially if the amount and the frequency of
substances used are not taken into consideration, and the
overall impression about the prevalence of polydrug use
remains vague. Hence, definitions and measurements
that are currently in use make a big difference when
studying the prevalence of polydrug use and the estimates
vary according to the timing of ingestion and the substances
included.
The definitions of the measurements were especially
crucial when observing the trends. It turned out that even
the course of the trend – whether increasing or decreas-
ing – depended on the measurement. Although it is often
claimed that polydrug use has increased, there is not
much research evidence to confirm this, at least not
among the general population. An increasing trend was
found among French adolescents during 1993–1999 [14],
but otherwise the changes in the prevalence of CPU have
been very modest (this study) or the trend has remained
stable [15, 16].
Furthermore, as shown in our results, the substances
included into the CPU combination have a significant influence
on the prevalence trends. This was evident especially
when current daily smoking was considered one of
the substances (Loose CPU). With this measure, the trend
was decreasing, and thus in the opposite direction to the
other measures, which did not include smoking (Strict
CPU and Medial CPU). The decreasing trend was due to
the significant reduction in smoking in Finland during
the 2000s [31]. Similarly, since 2008, the use of alcohol is
slowly decreasing among the Finnish general population
[32], and therefore, the increase in CPU in our example
may mostly be due to the increasing use of illicit drugs
[33]. These alterations in the use of single substances
should be taken into consideration when measuring polydrug
use, since CPU in particular may be prone to these
kinds of changes. Moreover, even though including
smoking as a part of polydrug use is justified from the
perspective of health and long-term effects of substance
0
5
10
15
20
25
30
1998 2002 2006 2010 2014
Percentage
Loose CPU
Medial CPU (b)
Medial CPU (a)
Strict CPU
Fig. 1. The trends of concurrent polydrug use (CPU) based on different
measurements among the Finnish general population during
the period 1998–2014.
Downloadedby:
MasarykUniversity
147.251.175.227-4/26/20239:57:25AM
Measuring Polydrug Use Eur Addict Res 2017;23:163–169
DOI: 10.1159/000477802
167
use (since it provides a measure of all the different substances
a person uses), it is also somewhat problematic,
especially if those using alcohol only become classified as
polydrug users due to their smoking, as noted elsewhere,
too [25].
The alcohol measure (alcohol use [any amount]/binge
drinking) did not seem to have an effect on the prevalence
or the prevalence trends of Medial CPU. It may be that
those classified as alcohol users in this study were in fact
mostly also binge drinkers. Or it may be that those who
used alcohol but were not binge drinkers did not report
about the use of other substances, and therefore they did
not become classified as polydrug users. Indeed, at least
in Finland, especially heavy episodic drinking has shown
to be associated with other drug use [26]. We used binge
drinking as a measure for risky drinking, but the results
may have been somewhat different if we had used, for example,
hazardous or harmful alcohol use based on the
AUDIT (Alcohol Use Disorders Identification Test)
score.
Although there is some evidence that the number of
drugs used concurrently could be used as a proxy of the
severity of polydrug use [34], CPU still fails to measure
whether the substances are used at the same time intentionally,
for instance, to get more intoxicated or to enhance
the effects of other drugs [17]. This is of concern,
since the harms and risks of SPU may have more detrimental
consequences than those of CPU, for instance,
unpredictable joint effects of different substances or
greater intoxication, which increases the risk of injury
or overdose [9, 35, 36]. If a rationale for the study of
polydrug use as a separate phenomenon is that it induces
particular negative health consequences [37], occasional
concurrent use of 2 or more substances during a
year does not make much sense. The concept of CPU
may help to identify those possibly at risk, but by adhering
to joint effects of different substances, SPU would be
even more useful in preventing or reducing harms
caused by polydrug use. Although SPU would be a better
measure from this point of view and although the
relevance of SPU as a measure was emphasized al-
ready in the 1990s [38], CPU continues to be very widely
used.
As shown in this study, there are problems with CPU
as a measure, and therefore more emphasis should be put
on measuring SPU, in spite of the finding that the majority
of CPU are also using drugs simultaneously [11, 39].
Regardless of the high correlation of CPU and SPU, they
have remained 2 distinct phenomena [34, 40]. Hence, we
argue that as a measure, SPU would be more recommendable
in order to improve the comparability of the results
of different studies across time and cultures, and these
measures should be developed further in population surveys
worldwide.
EMCDDA has stated that: “it is not possible to arrive
at a single definition of polydrug use, which would
be necessary to develop standardized measures” [41],
but in our opinion it is important, and possible, to strive
towards a better and more comprehensive understanding
of the phenomenon using more uniform mea-
sures. Furthermore, it should be noted that the prevalence
alone does not reveal the harmfulness of polydrug
use; rather, additional factors such as the intensity of use
or potential harms of combining different substances
need to be taken into consideration [12, 13]. One possibility
to measure SPU in population surveys would be
to include questions concerning, for instance, differ-
ent combinations of substances used at the same time or
in temporal proximity, frequency and motivation of
such use as well as perceived harms of SPU. Such questions
might reveal detrimental combinations and their
prevalence, for example, among adolescents and
young people. However, more profound research, espe-
cially qualitative research, is needed in order to devel-
op appropriate, effective and comparable measures of
SPU.
In preventing polydrug use, health education should
not separate different substances but rather treat them in
conjunction, and attention should be paid, in particular,
to their unexpected joint effects. In addition, an integrated
approach and a shift from substance-specific policies
towards a policy of psychoactive substances are needed.
However, it is not possible to formulate adequate preventive
actions until the phenomenon is accurately mea-
sured.
Strengths and Limitations
The data used covered a long period of time, and since
the survey questions used have remained very similar
throughout the time series, the data was highly comparable.
New illicit substances that emerged during the survey
period have been subsequently included in the questionnaires
and, as a result, the list of illicit substances was
notably longer in 2014 compared to that in 1998. Given
the sampling protocol, where the institutionalized population
and those without permanent address were excluded,
the results may be somewhat underestimated,
since this kind of restriction excludes, for instance, problem
drug users from the sample. In addition, the decreasing
response rate and a rather small numbers of observa-
Downloadedby:
MasarykUniversity
147.251.175.227-4/26/20239:57:25AM
Karjalainen/Kuussaari/Kataja/Tigerstedt/
Hakkarainen
Eur Addict Res 2017;23:163–169
DOI: 10.1159/000477802
168
tions in certain variables can lead to unstable prevalence
estimates. The data used was based on self-report and
people may not want to reveal their illicit drug use patterns.
On the other hand, the survey was confidential and
anonymous, and this could probably reduce the response
bias.
Conclusion
The prevalence of CPU varies greatly depending on
the measurement used, as does the course of the prevalence
trends. It is difficult to develop adequate preventive
actions until the phenomenon is accurately measured.
The concept of SPU may capture the essence of the phenomenon
better than the concept of CPU.
Acknowledgement
The authors wish to thank the Academy of Finland for their
financial support (grant No. 274415).
Disclosure Statement
The authors declare that they have no conflicts of interest.
References
1 Douglass FM, Khavari KA: The drug use index:
a measure of the extent of polydrug usage.
Int J Addict 1978;13:981–993.
2 Kaufman E: Polydrug abuse or multidrug
misuse: it’s here to stay. Br J Addict Alcohol
Other Drugs 1977;72:339–347.
3 Hakkarainen P: Polydrug use – a challenge to
research and policy. Conference paper presented
in 10th Annual ISDDP (The International
Society for the Study of Drug Policy)
Conference. Sidney, Australia, 2016.
4 EMCDDA: Annual Report 2011: The State of
Drug Problem in Europe. Luxembourg, Publications
Office of the European Union, 2011.
5 Council of the European Union: EU Drugs
Strategy (2013–2020). C 402/1, 2012.
6 Babor T, Caulkins J, Edwards G, Fischer B,
Foxcroft D, Humphreys K, Obot I, Rehm J,
Reuter P, Room R, Rossow I, Strang J: Drug
Policy and the Public Good. Oxford, Oxford
University Press, 2010.
7 Byqvist S: Patterns of drug use among drug
misusers in Sweden. Gender differences.
Subst Use Misuse 2006;41:1817–1835.
8 Midanik LT, Tam TW, Weisner C: Concurrent
and simultaneous drug and alcohol use:
results of the 2000 National Alcohol Survey.
Drug Alcohol Depend 2007;90:72–80.
9 Coffin PO, Galea S, Ahern J, Leon AC, Vlahov
D, Tardiff K: Opiates, cocaine and alcohol
combinations in accidental drug overdose
deaths in New York City, 1990–98. Addiction
2003;98:739–747.
10 Connor JP, Gullo MJ, White A, Kelly AB:
Polysubstance use: diagnostic challenges, patterns
of use and health. Curr Opin Psychiatry
2014;27:269–275.
11 Martin CS: Timing of alcohol and other drug
use. Alcohol Res Health 2008;31:96–99.
12 MammoneA,FabiF,ColasanteE,SicilianoV,
Molinaro S, Kraus L, Rossi C: New indicators
to compare and evaluate harmful drug use
among adolescents in 38 European countries.
Nordic Stud Alcohol Drug 2014;31:343–358.
13 Höhne B, Pabst A, Hannemann T, Kraus L:
Patterns of concurrent alcohol, tobacco, and
cannabisuseinGermany:prevalenceandcorrelates.
Drugs (Abingdon Engl) 2014;21:102–
109.
14 Choquet M, Morin D, Hassler C, Ledoux S: Is
alcohol, tobacco, and cannabis use as well as
polydrug use increasing in France? Addict Behav
2004;29:607–614.
15 Kelly AB, Chan GC, White A, Saunders JB,
Baker PJ, Connor JP: Is there any evidence of
changes in patterns of concurrent drug use
among young Australians 18–29 years between
2007 and 2010? Addict Behav 2014;39:
1249–1252.
16 Kokkevi A: Polydrug use by European adolescents;
in Hibell B, Guttormsson U, Ahlström
S, Balakireva O, Bjarnason T, Kokkevi A,
Kraus L (eds): The 2011 ESPAD Report – Substance
Use Among Students in 35 European
Countries. Stockholm, The Swedish Council
for Information on Alcohol and Other Drugs
(CAN), 2012, pp 166–170.
17 Schensul JJ, Convey M, Burkholder G: Challenges
in measuring concurrency, agency and
intentionality in polydrug research. Addict
Behav 2005;30:571–574.
18 Quintero G: Controlled release: a cultural
analysis of collegiate polydrug use. J Psychoactive
Drugs 2009;41:39–47.
19 Stinson FS, Grant BF, Dawson DA, Ruan WJ,
Huang B, Saha T: Comorbidity between
DSM-IV alcohol and specific drug use disorders
in the United States: results from the National
Epidemiologic Survey on Alcohol and
Related Conditions. Drug Alcohol Depend
2005;80:105–116.
20 Salasuo M, Vuori E, Piispa M, Hakkarainen P:
Suomalainen Huumekuolema 2007. Poikkitieteellinen
Tutkimus Oikeuslääketieteellisistä
Kuolinsyyasiakirjoista [Drug-Related Deaths
in Finland in 2007. A Multidisciplinary Study
on Forensic Causes-of-Death Data]. Helsinki,
Yliopistopaino, Nuorisotutkimusseura, 2009.
21 Smith GW, Farrell M, Bunting BP, Houston
JE, Shevlin M: Patterns of polydrug use in
Great Britain: findings from a national household
population survey. Drug Alcohol Depend
2011;113:222–228.
22 Carter JL, Strang J, Frissa S, Hayes RD; SELCoH
Study Team, Hatch SL, Hotopf M: Comparisons
of polydrug use at national and inner
city levels in England: associations with demographic
and socioeconomic factors. Ann
Epidemiol 2013;23:636–645.
23 Olszewski D, Matias J, Monshouwer K, Kokkevi
A: Polydrug use among 15- to 16-
year  olds: similarities and differences in
Europe. Drugs (Abingdon Engl) 2010;17:
287–302.
24 McCabe SE, Cranford JA, Morales M, Young
A: Simultaneous and concurrent polydrug
use of alcohol and prescription drugs: prevalence,
correlates, and consequences. J Stud Alcohol
2006;67:529–537.
25 Ives R, Ghelani P: Polydrug use (the use of
drugs in combination): a brief review. Drugs
(Abingdon Engl) 2006;13:225–232.
26 Hakkarainen P, Metso L: Joint use of drugs
and alcohol. Eur Addict Res 2009;15:113–
120.
27 Karjalainen K, Savonen J, Hakkarainen P:
Suomalaisten Huumeiden Käyttö ja
Huumeasenteet – Huumeaiheiset Väestökyselyt
Suomessa 1992–2014 [Drug Use and
Drug Attitudes among Finns – Drug-Related
Population Surveys in Finland 1992–2014].
Helsinki, National Institute for Health and
Welfare, 2016.
28 Tomczyk S, Isensee B, Hanewinkel R: Latent
classes of polysubstance use among adolescents
– a systematic review. Drug Alcohol Depend
2016;160:12–29.
29 Reyes J, Perez C, Colon H, Dowell M, Cumsille
F: Prevalence and patterns of polydrug
use in Latin America: analysis of populationbased
surveys in six countries. Rev Eur Stud
2013;5:10–18.
Downloadedby:
MasarykUniversity
147.251.175.227-4/26/20239:57:25AM
Measuring Polydrug Use Eur Addict Res 2017;23:163–169
DOI: 10.1159/000477802
169
30 Oliveira LG, Alberghini DG, Santos Bd, Andrade
AG: Polydrug use among college students
in Brazil: a nationwide survey. Rev Bras
Psiquiatr 2013;35:221–230.
31 Varis T, Virtanen S: Tupakkatilasto 2014 [Tobacco
Statistics 2014]. Statistical Report
18/2015. Helsinki, National Institute for
Health and Welfare, 2015.
32 Jääskeläinen M, Virtanen S and Räik-
könen  O: Päihdetilastollinen Vuosikirja
2016: Alkoholi ja Huumeet [Yearbook of Alcohol
and Drug Statistics 2016]. Helsinki, Juvenes
Print – Suomen Yliopistopaino Oy,
National Institute for Health and Welfare,
2017.
33 Hakkarainen P, Karjalainen K, Ojajärvi A,
Salasuo M: Huumausaineiden ja kuntodopingin
käyttö ja niitä koskevat mielipiteet
Suomessa vuonna 2014 [Drug use, doping
and public opinion in Finland: results from
the 2014 Drug Survey]. Yhteiskuntapolitiikka
2015;80:319–333.
34 Baggio S, Studer J, Mohler-Kuo M, Daeppen
JB, Gmel G: Concurrent and simultaneous
polydrug use among young Swiss males: use
patterns and associations of number of substances
used with health issues. Int J Adolesc
Med Health 2014;26:217–224.
35 Karjalainen K, Lintonen T, Impinen A,
Mäkelä P, Rahkonen O, Lillsunde P, Ostamo
A: Mortality and causes of death among
drugged drivers. J Epidemiol Community
Health 2010;64:506–512.
36 Subbaraman MS, Kerr WC: Simultaneous
versus concurrent use of alcohol and cannabis
in the national alcohol survey. Alcohol Clin
Exp Res 2015;39:872–879.
37 Leri F, Bruneau J, Stewart J: Understanding
polydrug use: review of heroin and cocaine
co-use. Addiction 2003;98:7–22.
38 Martin CS, Clifford PR, Clapper RL: Patterns
and predictors of simultaneous and concurrent
use of alcohol, tobacco, marijuana, and
hallucinogens in first-year college students. J
Subst Abuse 1992;4:319–326.
39 Quek LH, Chan GC, White A, Connor JP,
Baker PJ, Saunders JB, Kelly AB: Concurrent
and simultaneous polydrug use: latent class
analysis of an Australian nationally representative
sample of young adults. Front Public
Health 2013;1:61.
40 Earleywine M, Newcomb MD: Concurrent
versus simultaneous polydrug use: prevalence,
correlates, discriminant validity, and
prospective effects on health outcomes. Exp
Clin Psychopharmacol 1997;5:353–364.
41 EMCDDA: Polydrug Use: Patterns and Responses.
EMCDDA Selected Issue 2009. Luxembourg,
Office for Official Publications of
the European Communities, 2009.
Downloadedby:
MasarykUniversity
147.251.175.227-4/26/20239:57:25AM