Content uploaded by Daan van der Gouwe
Author content
All content in this area was uploaded by Daan van der Gouwe on Dec 08, 2017
Content may be subject to copyright.
Purity, adulteration and price of drugs bought on-line
versus off-line in the Netherlands
Daan van der Gouwe, Tibor M. Brunt, Margriet van Laar & Peggy van der Pol
Trimbos Institute, Netherlands Institute of Mental Health and Addiction, Utrecht, the Netherlands
ABSTRACT
Background and aims On-line drug markets flourish and consumers have high expectations of on-line quality and drug
value. The aim of this study was to (i) describe on-line drug purchases and (ii) compare on-line with off-line purchased
drugs regarding purity, adulteration and price. Design Comparison of laboratory analyses of 32663 drug consumer
samples (stimulants and hallucinogens) purchased between January 2013 and January 2016, 928 of which were bought
on-line. Setting The Netherlands. Measurements Primary outcome measures were (i) the percentage of samples pur-
chased on-line and (ii) the chemical purity of powders (or dosage per tablet); adulteration; and the price per gram, blotter
or tablet of drugs bought on-line compared with drugs bought off-line. Findings The proportion of drug samples pur-
chased on-line increased from 1.4% in 2013 to 4.1% in 2015. The frequency varied widely, from a maximum of 6% for
controlled, traditional substances [ecstasy tablets, 3,4-methylenedioxy-methamphetamine (MDMA) powder, amphet-
amine powder, cocaine powder, 4-bromo-2,5-dimethoxyphenethylamine (2C-B) and lysergic acid diethylamide (LSD)] to
more than a third for new psychoactive substances (NPS) [4-fluoroamphetamine (4-FA), 5/6-(2-aminopropyl)benzofuran
(5/6-APB) and methoxetamine (MXE)]. There were no large differences in drug purity, yet small but statistically significant
differences were found for 4-FA (on-line 59% versus off-line 52% purity for 4-FA on average, P=0.001),MDMApowders
(45 versus 61% purity for MDMA, P= 0.02), 2C-B tablets (21 versus 10 mg 2C-B/tablet dosage, P=0.49)andecstasy
tablets (131 versus 121 mg MDMA/tablet dosage, P= 0.05). The proportion of adulterated samples purchased on-line
and off-line did not differ, except for 4-FA powder, being less adulterated on-line (χ
2
=8.3;P<0.02). Drug prices were
mainly higher on-line, ranging for various drugs from 10 to 23% higher than that of drugs purchased off-line (six of 10
substances: P<0.05). Conclusions Dutch drug users increasingly purchase drugs on-line: new psychoactive sub-
stances in particular. Purity and adulteration do not vary considerably between drugs purchased on-line and off-line for
most substances, while on-line prices are mostly higher than off-line prices.
Keywords Adulteration, cryptomarkets, darknet, dosage, drug markets, price, purity, quality, webshops.
Correspondence to: Daanvan der Gouwe, Trimbos Institute, Netherlands Instituteof Mental Health and Addiction, Da Costakade 45, 3521 VS, Utrecht, PO Box
725, 3500 AS Utrecht, the Netherlands. E-mail: dgouwe@trimbos.nl
Submitted 26 April 2016; initial review completed 25 July 2016; final version accepted 8 December 2016
INTRODUCTION
Within the past decade, the marketing, sale and sourcing of
(illicit) drugs via the internet has grown rapidly [1–7]. The
perception of better product quality is a main trigger for
customers to purchase substances on-line [3–5,7–13],
yet there is little evidence to support this assumption. This
undue trust in on-line markets is potentially harmful,
because customers are often unaware of the presence of
unexpected or unknown substances or harmful adulter-
ants in drug products obtained on-line [14], which may
increase the risk of adverse drug events.
Furthermore, the perception of lower prices may also
attract customers to on-line markets [5,11]. However,
studies of on-line drug pricing are scarce and contradic-
ting. In fact, on-line prices were found to be higher for
certain drugs [3], to vary with drug quality [15], to be
unstable over time [6,16] and to vary geographically
[5]. Another study mentioned the relatively low prices
of on-line drugs, but proposes that this may be explained
by bulk offers intended for further trade and resale and
are thus incomparable with customer-level street prices
[17]. However, the perception of good value and of better
product quality may direct consumers to Google-indexed
© 2016 Society for the Study of Addiction Addiction
RESEARCH REPORT doi:10.1111/add.13720
webshops and cryptomarkets, or ‘darknet markets’,
which are accessible only by using encryption software
[11]. Customers reported that they purchased drugs on-
line (particularly on cryptomarkets) for anonymity, conve-
nience, customer service, the abundance of suppliers and
the larger diversity of products offered compared with
those available at local drug markets and dealers
[4,5,7,10,14,18,19]. The growth of on-line drug pur-
chases may also stem from another feature that on-line
drug markets share with regular websites: customer re-
views. Cryptomarkets in particular use mandatory feed-
back systems for ‘quality control’[16], where shoppers
comment on, for example, the reliability and security of
the supply, financial value and, importantly, perceived
chemical purity [19]. However, such reports on perceived
purity may be biased (e.g. artificial reviews from vendors
or customers) or, in fact, may refer to a completely differ-
ent substance or batch, as stocks change constantly. As
customers do not have access to information on the ac-
tual content of the substances they intend to use, which
would allow them to adapt their intake to its dosing, they
have to rely upon such subjective customer reviews.
While research on this topic is still in its infancy, the
2015 Global Drug Survey (GDS) suggested that the
perceived purity was higher for drugs purchased from
cryptomarkets than when sourced via other routes. Only
27% of an international sample of darknet purchasers
reported low purity versus 74% of those who purchased
drugs from other sources. Similarly, 11% of on-line pur-
chasers reported having received a product that did not
contain the advertised substance compared with 41% of
those who purchased drugs from other sources [9]. How-
ever, studies on the quality of on-line drug markets rely
upon consumer-perceived quality instead of objective
chemical laboratory analyses that would provide accurate
information on drug purity and the presence of adulter-
ants. It is thus unknown whether the GDS results sug-
gesting that drugs on the cryptomarkets are, in fact, of
better quality, are justified or whether this merely reflects
the reputation of the anonymous, global on-line drug
market in contrast to local and face-to-face drug markets.
One study offering chemical analyses of drug samples and
harm reduction information to cryptomarket customers
indeed suggested high substance purity, although a direct
comparison with off-line markets was not available
[12,14].
In the Netherlands, the Drugs Information and Moni-
toring System (DIMS) has monitored the composition of
consumer-derived drug samples available on the Dutch
market for more than two decades [20–22]. This study
aims to (i) describe on-line drug purchases and (ii) compare
the laboratory-analysed quality (chemical purity and pres-
ence of other psychoactive substances) and price between
on-line and off-line sourced drugs.
METHODS
Design
This study includes data on consumer drug samples col-
lected by DIMS from 1 January 2013 to 1 January 2016.
The samples were laboratory-analysed for their contents,
using gas chromatography coupled to mass spectrometry
(GC–MS) and liquid chromatography with diode array
detection (LC-DAD). For a detailed description of the testing
procedure and laboratory techniques used, see Brunt et al.
2016 [23].
DIMS monitors the Dutch drugs market at consumer
level: consumers are able to submit drug samples anony-
mously at testing offices embedded typically in regional
institutes of prevention and addiction care throughout
the country. The main reasons for testing drugs are ‘health
concern’and ‘curiosity’[22,24]. Users of these facilities
showed a relatively high education or paid employment
and were mainly of Dutch ethnicity [24,25].
Drug sample information, such as price, region of
purchase and name under which the drug was sold to
the consumer, are recorded at the testing offices. Although
the on-line purchasing source is not a mandatory
reporting category, drug-testing personnel were asked to
register this information as a string variable under ‘partic-
ularities’. Prior to 2013, the number of on-line-bought
samples submitted at DIMS was limited and not commonly
registered. On-line purchases were identified by searching
this string for internet-related words (e.g. Silk Road, inter-
net, on-line or specific web addresses or names of known
webshops; see on-line attachment). All samples identified
as having been purchased on-line were reviewed manually
to exclude misclassified samples (e.g. ‘consumer read on
the internet that…’). In addition, a random sample of
10% of all drug samples identified as having been pur-
chased off-line were reviewed manually to identify addi-
tional on-line-related words. Adding these words, the
search query and manual check was repeated twice. The
non-on-line group was labelled off-line for brevity, but it
should be noted that this category also includes samples
with no information on purchase location, and thus these
drugs could have been obtained on-line. Using the on-line-
related key words, the on-line source was specified further
as purchased in a Google-indexed webshop, a
cryptomarket or unspecified (see on-line attachment).
The category ‘no advertised drug’could, in principle,
contain other psychoactive substances as well as no
psychoactive substances. However, samples submitted at
DIMS rarely contain no psychoactive substances at all.
Therefore, the category ‘no advertised drug’generally
contains other psychoactive substances.
Mean prices are described as price per tablet (pill) or
blotter [for lysergic acid diethylamide (LSD)] or per gram
(for powders). These are calculated based on self-reported
2Daan van der Gouwe et al.
© 2016 Society for the Study of Addiction Addiction
information by DIMS customers upon submitting the sam-
ple. When prices for amounts other than per tablet, blotter
or gram were reported the prices were adjusted
accordingly.
Statistical analyses
Trends in on-line purchasing were assessed as the monthly
proportion of on-line-purchased drug samples of the total
and the proportion of on-line purchases was reported per
substance (chemical class). Further analyses were re-
stricted to substances for which a minimum of 15 on-line
samples were available. For each of those drugs, the type
of on-line source was specified. The mean chemical purity
of powders (or dosage per ecstasy tablet) were compared
between on-line and off-line purchases with unpaired
two-sided t-tests. Then, the omnibus χ
2
test was used to
compare the proportion of on-line versus off-line consumer
drug samples that were unadulterated (only containing the
advertised substance) versus adulterated (‘not containing
the advertised substance’or ‘containing the advertised
substance and other psychoactive substances’). The
other psychoactive substances that were assessed chemi-
cally included 3,4-methylenedioxy-amphetamine (MDA),
3,4-methylenedioxy-methamphetamine (MDMA), 3,4-
methylenedioxy-N-ethylamphetamine (MDEA), amphet-
amine, methamphetamine, meta-chlorophenylpiperazine
(mCPP), phenacetin, cocaine, 2,5-dimethoxy-4-
bromophenethylamine (2C-B), levamisole, gamma-
hydroxybutyric acid (GHB), gamma-butyrolacton (GBL),
LSD, 4-methylamphetamine, 4-fluoroamphetamine (4-FA/
4-FMP), lidocaine, procaine, noscapine, 5-(2-aminopropyl)
benzofuran/6-(2-aminopropyl)benzofuran (5/6-APB),
methoxetamine and other pharmacologically active com-
pounds, such as medicines or illicit drugs from the NIST
library [26]. Caffeine was not considered an adulteration.
Then, the psychoactive substances detected most frequently
were described for the adulterated samples. Finally, as for
purity, on-line and off-line prices were compared with un-
paired two-sided t-tests using the natural logarithm, as price
data were not distributed normally. The chemical analysis
data and prices were analysed using SPSS version 22.
Analysis of tablets
The tablet dosages are expressed in mg/tablet rather than
percentages of pure substance, as is the case for powders.
Therefore, tablet quality and price were reported separately
and excluded from the figures and tables for all substances
except ecstasy tablets, which comprised the majority of
tablet samples. It should be noted that ecstasy tablets are
not always laboratory-tested. If they were recognized ac-
cording to DIMS protocol, the mean value of at least two
matching tablets analysed chemically in the past 3 months
was used for recognized tablets (at face value, based on
logo, shape, colour, diameter, thickness, physical profiles,
grooves and Marquis reagent test results) [21].
RESULTS
On-line purchasing
A total of 32663 drug samples were submitted to DIMS
between January 2013 and January 2016, 928 of which
were identified as having been sourced on-line. The propor-
tion of samples registered as having been sourced on-line
increased from 1.4% in 2013 to 4.1% in 2015 (Fig. 1)
and varied largely among substances. Fewer than 6% of
controlled drugs (ecstasy 1%, MDMA powder 2%, amphet-
amine 1%, cocaine 1%, LSD 5%, and 2C-B 6%) to more
than half of non-controlled drugs [4-fluoroamphetamine
(4-FA) 32%; 5/6-APB 48%) or recently controlled drugs
(methoxetamine (MXE) 54%] were purchased on-line.
The following substances were excluded from further anal-
yses because fewer than 15 on-line samples were available:
heroin (n= 88 off-line/0 on-line), GHB/GBL (n=176
off-line/n= 1 on-line), ketamine (n= 842 off-line/n=11
on-line), mephedrone (n=71off-line/n= 5 on-line) and
‘other substances’, including NBOMes, 3-MMC, methylone
and 2C-E/I/P (in total n= 543 off-line/n= 99 on-line).
In total, 15% (n= 136) of the on-line samples were pur-
chased from cryptomarkets, 26% (n=245)fromGoogle-
indexed webshops and the majority (59%; n=547)from
unknown on-line sources (e.g. registration mentioned only
as ‘on-line’or ‘internet’). Despite this large proportion of
drug samples from unspecified on-line sources, Fig. 2
shows clearly that controlled traditional substances were
purchased rarely from Google-indexed webshops, whereas
non-controlled or recently controlled NPS were purchased
rarely from cryptomarkets.
Purity, adulteration and price
There were no large differences in average chemical purity
between drugs purchased on-line and off-line, yet small but
statistically significant differences existed. The mean purity
of 4-FA powder and 2C-B tablets were higher in on-line
samples compared with off-line samples (see Table 1)
whereas, in contrast, MDMA powders were of lower purity
on-line than off-line. For MDMA powders, the interquartile
range (IQR) varied largely, with 50% of the samples be-
tween 0 and 79 mg MDMA. Comparing the proportions
of adulterated samples between the on-line and off-line
samples revealed only a small but statistically significant
difference for 4-FA powder, whichcontained a higher purity
in the samples purchased on-line (χ
2
=8.3;P<0.02) (see
Fig. 3). Figure 3 shows that cocaine and 2C-B powders had
large proportions of adulterated samples compared with
amphetamine, 4-FA, 5/6-APB, MXE, ecstasy tablets and
On-line and off-line drug quality and price 3
© 2016 Society for the Study of Addiction Addiction
MDMA powders. While approximately half of samples sold
as LSD were unadulterated, those that were adulterated
rarely contained LSD but usually contained other psycho-
active substances. MDMA powders and ecstasy tablets
frequently contained amphetamine or, to a lesser extent,
para-methoxymethamphetamine (PMMA). The most im-
portant cocaine adulterant was levamisole.
For most other substances with fewer on-line samples,
small differences in the same direction were observed, but
these were not statistically significant. For substances for
which at least 20 on-line purchased samples were available
(i.e. ecstasy tablets, cocaine and 4-FA), similar adultera-
tions were found compared with samples bought off-line.
Although the overall purities of samples purchased
on-line and off-line were not very different, average on-
line prices [as reported by DIMS customers and prices
measured per tablet or blotter (for LSD), or per gram of
powder] were higher for six out of 10 substances
Figure 1 Proportion of samples submitted at
Drugs Information and Monitoring System
(DIMS) bought on-line between January 2013
and January 2016
Figure 2 Sourcing location of on-line-purchased samples submitted at Drugs Information and Monitoring System (DIMS) between January 2013
and January 2016. Percentage of samples per substance categorized by its sourcing location.
a
Google-indexed webshop: on-line shop indexed by
Google and other searching engines, where drugs are being marketed, usually as research chemicals.
b
Cryptomarket or darknet market: on-line plat-
form where drugs are being sold and that is not indexed by Google.
c
Unspecified: % of samples per substance bought on-line, but without information
about its specific on-line purchasing location
4Daan van der Gouwe et al.
© 2016 Society for the Study of Addiction Addiction
(P=<0.05) (Table 2). However, price differences varied
widely among the different types of substances: ecstasy
tablets (+10% on-line), amphetamine powder (+23%
on-line), cocaine powder (+22% on-line), 4-FA powder
(+17% on-line) and 5/6-ABP powder (+23% on-line).
2C-B powder showed a similar but non-statistically signif-
icant trend (+16% on-line).
Tablets
Ecstasy tablets were relatively unadulterated and more
expensive on-line (Fig. 3 and Table 2). As with 4-FA,
but in contrast to MDMA powder, the doses of ecstasy
tablets were higher on-line than off-line. The on-line sup-
plement shows that there were too few amphetamine
Table 1 Chemical purity of consumer drug samples bought on-line versus off-line
a
.
nPurityin%
b
On-line Off-line
a
On-line
mean (SD)
Off-line
a
mean (SD) P t
Mean difference on-line–
off-line (95% CI)
Ecstasy tablets (‘purity’=doseinmg)
b+c
197 18 363 131 mg (61)
b
121 mg (66)
b
0.05 2.0 9.3 (0.04; 18.5)
b
4-FA tablets (‘purity’=doseinmg)
b
10 273 69 mg 30 mg 0.074 2.0 40.0 (4.8; 84.8)
2C-B tablets (‘purity’= dose in mg)
b
20 409 21 mg 10 mg 0.049 2.0 10.7 (3.1; 18.3)
MDMA powders 43 1907 45% (38) 61% (44) 0.02 –2.4 16.6 (30.0; –3.1)
Amphetamine powders 43 3475 39% (26) 41% (25) 0.7 –0.4 –1.5 (9.2; 6.1)
Cocaine powders 28 4158 56% (29) 53% (27) 0.6 0.5 2.3 (7.5; 12.3)
LSD (‘purity’= dose in μg)
b
34 693 35 μg(36)
b
33 μg(46)
b
0.9 0.2 1.4 (14.4; 17.2)
b
2C-B powders 13 83 44% (37) 42% (35) 0.9 0.1 1.4 (19.6; 22.5)
4-FA powders 344 480 59% (32) 52% (35) 0.001 3.3 7.8 (3.2; 12.6)
5/6-APB powders 52 53 NA NA
Methoxetamine powders 23 20 NA NA
a
Off-line = samples not on-line or from unspecified source.
b
Purity represents the meanpercentage of pure substance. The maximum chemical purityof pow-
ders is never 100% (analysed to the base component of the substance), but variesbetween substances: 3,4-methylenedioxy-methamphetamine (MDMA) 84%,
amphetamine 73%, cocaine89%, 2,5-dimethoxy-4-bromophenethylamine(2C-B) 88%, 4-fluoroamphetamine(4-FA)81%. For tablets purityis not defined as
the percentage of pure substance, but as the mean dosage in mg per tablet. Similarly, lysergic acid diethylamide (LSD) ‘purity’is the dosage in μg per blotter.
Type of adulterants in on-line and off-line samples were found to be similar.
c
Dose of ecstasy tablets from cryptomarkets (n= 39) was analysed separat ely and
compared with the off-line samples: mean dose = 149 mg (P= 0.01). CI = confidence interval; SD = standard deviation; 5/6-APB = 5-(2-aminopropyl)
benzofuran/6-(2-aminopropyl)benzofuran; NA = not avalaible. Degrees of freedom t-test was n-2.
a
Figure 3 Proportion unadulterated and adulterated consumer drug samples bought on-line versus off-line. Off-line = samples from unspecified
source and those sourced not on-line.
a
The category ‘ecstasy’contains tablets that were sold as ecstasy or 3,4-methylenedioxy-methamphetamine
(MDMA). Amphetamine and 5/6-APB tablets were excluded. ‘Only advertised drug’contains the specified drug alone, ‘Advertised drug + other’
contains the specified drug and another active component, ‘No advertised drug’does not contain the specified drug and may or may not contain
another active component. χ
2
=omnibusχ
2
test with 2 degrees of freedom for every drug category. LSD = lysergic acid diethylamide;
2C-B = 2,5-dimethoxy-4-bromophenethylamine; 4-FA = 4-fluoroamphetamine; MXE = methoxetamine
On-line and off-line drug quality and price 5
© 2016 Society for the Study of Addiction Addiction
and 5/6-APB tablets for further analysis, whereas most
2C-B samples were tablets (n= 429 of 525). The on-line
purchased 2C-B tablets were more often unadulterated
(60%) than off-line purchased tablets (28%) and powders
(on-line and off-line about 20%). Strikingly often, 2C-B
tablets and 4-FA tablets sourced off-line did not contain
the specified substance (62 and 76%, respectively). This
was also reflected by the average tablet dosages being
lower off-line than on-line, yet prices did not differ statis-
tically significantly.
DISCUSSION
This is one of the first studies, to our knowledge, to
compare the purity and prices of on-line and off-line drug
markets using laboratory-verified consumer samples. At
DIMS there has been an increase in on-line drug pur-
chases since 2013, yet the vast majority of collected sam-
ples were still sourced off-line. This trend is in line with
results from the Global Drug Survey 2016, which reports
ariseinlastyear’s darknet purchases from 4.5 to 6.7%
among its global participants [13]. As on-line prices were
generally higher than those from off-line markets and
the purity did not vary considerably between the on-line
and off-line groups, our findings neither confirm nor
refute the good reputation of on-line drug quality
compared with that of off-line drug markets, at least in
the Netherlands.
On-line purchasing
Almost all (on-line and off-line) consumer samples in this
study were psychostimulants; most samples were ecstasy
tablets, as ecstasy is the most frequently used illicit sub-
stance in the Netherlands after cannabis [27]. However,
unlike previous studies in other countries [5,10,16] 4-FA,
not ecstasy, was the drug purchased most frequently on-
line in our study. This is in line with findings from the
Global Drug Survey 2015; the proportion of Dutch GDS
participants who reported 12-month on-line drug pur-
chases was comparable with that in the total sample (8.8
versus 9.3%). However, within this group, the proportion
reporting on-line purchases of NPS was much higher
among Dutch respondents (72 versus 28% in the total
sample), whereas the proportion reporting on-line pur-
chases of traditional illicit drugs was much lower than
the total sample (37 versus 76% in the total sample)
(Winstock, personal communication, March/April 2016).
Typically, Dutch drug users do not search on-line for tradi-
tional illicit substances (but when theydo, this is mainlyon
cryptomarkets), whereas for non-controlled NPS they tend
to access Google-indexed webshops but not cryptomarkets
[14]. This is in correspondence with Caudevilla’s
Table 2 Prices
a
of consumer drug samples bought on-line versus off-line
b
.
n Price in €
a
Ln (price in €
a
)
b
n
On-
line
Off-
line
b
On-line
mean (SD)
Off-line
b
mean (SD)
On-line
mean (SD)
Off-line
b
mean (SD) P t
Mean difference
ln(price) on-line–
off-line (95% CI)
Ecstasy tablets
c
185 13 831 4.2 (1.6) 3.8 (1.6) 1.37 (0.36) 1.28 (0.42) 0.001 3.2 0.10 (0.04; 0.16)
MDMA powders 39 1507 21.4 (13.2) 20.3 (9.9) 2.87 (0.69) 2.83 (0.75) 0.7 0.35 0.04 (0.19;0.28)
Amphetamine
powders
41 2990 9.4 (5.1) 7.2 (4.4) 2.09 (0.57) 1.81 (0.61) 0.003 2.94 0.28 (0.09; 0.47)
Cocainepowders 26 3696 65.1 (17.5) 51.1 (11.8) 4.13 (0.33) 3.88 (0.45) 0.005 2.84 0.25 (0.08; 0.42)
LSD 32 610 6.0 (2.0) 5.5 (5.8) 1.73 (0.35) 1.53 (0.55) 0.04 2.04 0.20 (0.01; 0.40)
2C-B tablets 19 340 21.1 (29.0) 9.7 (18.0) 1.20 (0.53) 1.29 (0.43) 0.4 –0.84 0.09 (0.29;0.12)
2C-B powders 12 61 48.8 (30.2) 41.1 (29.8) 3.55 (1.03) 3.29 (1.22) 0.5 0.70 0.26 (0.49;1.01)
4-FA powders
d
320 413 14.6 (8.0) 12.1 (8.2) 2.51 (0.65) 2.23 (0.77) <0.001
e
5.30 0.28 (0.18; 0.39)
5/6-APB
powders
50 43 26.4 (7.0) 20.3 (10.1) 3.22 (0.38) 2.77 (0.88) 0.001
e
3.28 0.45 (0.18; 0.72)
Methoxetamine
powders
23 18 21.0 (6.6) 21.3 (8.9) 2.97 (0.47) 2.90 (0.71) 0.7 0.37 0.07 (0.31;0.44)
a
Prices are expressed in price per tablet [per blotter for lysergic acid diethylamide (LSD)] or per gramfor powders.
b
Off-line = samples not on-line or from un-
specified source.
c
The mean price of on-line ecstasy tablets specifically from cryptomarkets(n= 38) was also higher than off-linesamples (mean price = €4.7,
ln(price) = 1.47, P=0.004).
d
As price data were available for only seven on-line 4-fluoroamphetamine (4-FA) tablets, these data were not included in this
table (mean on-line price = €3.6, mean off-line = €4.3 n=248).CI=confidence interval. The degrees of freedom for every t-test was the total nminus 2.
The total nis lower than in Table 1 due to missing price data.
e
The non-normally distributed price data were log-transformed (after imputing €1for€0) to
allow t-test. Results for 4-FA powders and 5/6-APB powders should be interpreted with caution, as the standard deviations after transformation differed
for on-line and off-line samples. 2C-B = 2,5-dimethoxy-4-bromophenethylamine; MDMA = 3,4-methylenedioxy-methamphetamine; 5/6-APB = 5-(2-
aminopropyl)benzofuran/6-(2-aminopropyl)benzofuran.
6Daan van der Gouwe et al.
© 2016 Society for the Study of Addiction Addiction
suggestion [12] and with the results of the Global Drug
Survey 2016, that reports that 50.5 % of respondents
source NPS on-line [13].
Purity, adulteration and price
A main objective of this study was to compare on-line and
off-line drug quality in the Netherlands, assessed as the
proportion of unadulterated samples and their overall
chemical purity. The purity did not vary considerably,
hereby contrasting with Caudevilla, who finds that
cryptomarket-sourced cocaine samples were more pure
and less adulterated compared to those bought on the
Spanish drug market [12]. However, on-line sourced 4-FA
was slightly but statistically significantly less adulterated
(7 percentage points) and of higher purity (also 7 percent-
age points) than samples bought off-line. Our finding, that
for most other substances small but statistically insignifi-
cant differences in the same direction were observed, may
represent a lack of power to identify similar small differ-
ences. The only exception was that MDMA powder was
more pure off-line than on-line. Nevertheless, as variations
among samples were much larger than the on-line/off-line
differences, these small differences were deemed clinically
irrelevant. For example, the difference was only 10 mg
MDMA for ecstasy tablets; the probability of experiencing
desirable effects peaks at 81–100 mg MDMA and adverse
effects tend to exceed desirable effects at doses above
160 mg [22]. Also, for the other types of substances, simi-
lar small differences in purity between off-line and on-line
purchases are not expected to result in major health
consequences.
The proportion of unadulterated samples was higher
for non-controlled substances than for controlled sub-
stances, with the exception of ecstasy tablets and LSD, both
controlled substances that were relatively unadulterated,
which are not powders and therefore presumably not adul-
terated easily after production. Tentatively, one may argue
that this illustrates the delicacy of the decision to control
substances emphasizing their health-related risks to poten-
tial users, and the potential side effect of creating a more
harmful adulterated market. The risk of contributing to a
more adulterated market, however, depends upon the tox-
icity of the adulterants used. This paper paints a general
picture of frequently detected adulterants per substance,
especially PMMA in ecstasy/MDMA and levamisole in co-
caine, which can cause severe health risks over and above
those of unadulterated MDMA and cocaine [28–30]. In the
only comparable study using chemical analyses of on-line
drug samples (International Drug Testing Service) almost
half the cocaine samples were adulterated, as 42 of 103
samples contained levamisole [12,14]. The average purity
of cocaine samples reported in this study was also much
higher (72%) than we found for on-line (56%) and off-line
(53%) cocaine powders. On-line/off-line sourcing does not
provide an explanation for the difference in findings, and
although it is early to draw conclusions, local differences
between drug markets and different time-frames may play
a more important role in this discrepancy. The comparison
of specific adulterants (particularly those that are more
toxic than the drug itself) between the on-line and off-line
markets remains an issue to be addressed to inform drug
consumerscomprehensibly and form a drug policystrategy.
In contrast to the small purity differences found, on-line
prices for various drugs were 10–23% higher, despite the
prevailing perception of better values on-line [5,11]. Tenta-
tively, such higher prices may be interpreted by the con-
sumer as higher quality, which might result in the intake
of lower amounts. Apparently, on-line drug shopping has
a number of advantages for which a minority of Dutch
customers are willing to pay a little extra.
Notwithstanding several plausible reasons for price dif-
ferences observed between on-line and off-line markets, the
findings in our study should be interpreted within the con-
text of the Dutch setting. Prices on the Dutch drug market
are relatively low compared with those in other European
countriesorAustralia[7,31–33]. Given the global nature
of on-line drug markets, this may in itself explain the
higher on-line prices compared with the local Dutch drug
market. Furthermore, the effort it takes to set up a newen-
terprise in a new market as well as advertising and ‘trans-
action costs’, such as the risk of arrest and of seizure of the
product, of low product quality and lack of choice, as well
as the risks of violence, may vary locally [34]. Finally, costs
related to shipments (includingloss of packages) may influ-
ence the (global) on-line markets to a lesser extent. More-
over, the Netherlands are a main producer of ecstasy and
amphetamine [7,32,35–37], which may augment local
off-line availability and low prices. Particularly because
the average doses of MDMA in ecstasy are historically high
combined with minimal risks of being caught and relatively
low sentencing, theremay be little urgency for Dutch users
to buy traditional drugs on-line.
This exemplifies difficulties in the extrapolation of our
results, as the Dutch market may not be representative of
foreign markets. Moreover, even though DIMS runs the
most advanced monitoring system to date allowing for val-
idation of consumer drug samples by laboratory analysis
(and provision of targeted prevention messages and
warnings), this monitoring system could be optimized by
improving the level of information retrieved from its
customers, especially regarding internet-sourced samples.
Although our data showed an increase in on-line pur-
chases in line with previous studies [11,14], it should be
noted that this study had to rely upon reporting of on-line
sources as registered in an open text field, which may have
led to under-reporting. Moreover, such data could lead to
bias with an increased awareness of testing staff over time
On-line and off-line drug quality and price 7
© 2016 Society for the Study of Addiction Addiction
or differential awareness (for reporting of the on-line
source) for different substances. Misclassification may also
arise when drug samples were purchased from a dealer,
or were given by a friend who, in turn, had purchased it
on-line. While overcoming these issues would increase
our estimate of the proportion of on-line purchased sam-
ples, this would dilute our on-line–off-line comparisons.
The current presented estimates are likely to be conserva-
tive. Currently, the data collection is improved by structur-
ally reporting the on-line source as a category, overcoming
such problems as missing on-line samples (50% missing
data regarding the specific on-line source in the current
study) and reducing the possibility of reporting bias.
Although drug samples purchased on-line still com-
prise a minority of the total number of samples, the internet
as a source for marketing or obtaining drugs is well and
truly a phenomenon of the contemporary world that is
likely to expand further. Despite our findings that on-line
customers receive on average equal quality for a higher
price than off-line, we hypothesize that consumers are will-
ing to pay more for the convenience of purchasing drugs
on-line. Despite the current modest role for on-line pur-
chasing controlled substances via cryptomarkets, Google-
indexed webshops already source up to half of the non-
controlled substances. Therefore, it seems advisable that
prevention professionals and harm reduction organizations
expand their territories to on-line markets [38]. The provi-
sion of harm reduction advice to cryptomarket shoppers, as
pioneered by Doctor X (e.g. Silk Road) might be monitored,
possibly evaluated and intensified.
It is clear that monitoring information of (on-line) drug
testing services, such as DIMS and the Energy Control
International Drug Testing Service [12], provide essential
insights into the substances to which increasing numbers
of (young) people expose themselves, and these data are
unavailable in other regions. These monitoring systems
improve our understanding of drug markets and allow us
to follow and respond to trends in on-line markets; for ex-
ample, a possible transition from centralized market-places,
such as Silk Road, to decentralized marketplaces such as
Open Bazaar. Avenues of further research include weighing
law-enforcement interventions against their potential in-
terference with the above-mentioned monitoring and
harm-reduction efforts, and the (potential) health gains of
controlling new psychoactive substances against the poten-
tial side effect of creating a more adulterated market.
Declaration of interests
None.
References
1. European Monitoring Centre for Drugs and Drug Addiction.
Online Sales of New Psychoactive Substances/‘Legal Highs’:
Summary of Results from the 2011 Multilingual Snapshots
[Briefing Paper]. Lisbon: EMCDDA; 2011, pp. 3–7.
2. Barratt M. J. Silk Road: eBay for drugs. Addiction 2012; 107:
683.
3. Van Hout M. C., Bingham T. Surfing the Silk Road: a study of
users’experiences. Int J Drug Policy 2013; 24:524–9.
4. Van Hout M. C., Bingham T. Silk Road, the virtual drug
marketplace: a single case study of u ser experiences. Int J Drug
Pol i c y 2013; 24:385–91.
5. Van Buskirk J., Roxburgh A., Bruno R., Naicker S., Lenton S.,
Sutherland R. et al. Characterising dark net marketplace
purchasers in a sample of regular psychostimulant users. Int
J Drug Policy 2016; 35:32–7.
6. Christin N. Traveling the Silk Road: a measurement analysis of
a large anonymous online marketplace. Proceedings of the
22nd international conference on World Wide Web. Geneva,
Switzerland: International World Wide Web Conferences
Steering Committee; 2013.
7. Kruithof K., Aldridge J., Decary-Hetu D., Sim M., Dujso E.,
Hoorens S. Internet-facilitated Drugs Trade. Santa Monica,
CA/Cambridge, UK: RAND Corporation; 2016, pp. 21–32.
8. Buxton J., Bingham T. The Rise and Challenge of Dark Net Drug
Markets. Policy Brief 7 (2015). Swansea: University of
Swansea; 2015, pp. 1–5.
9. Winstock A. R. Global Drug Survey. 2016-10-03. Avail-
able at: http://www.globaldrugsurvey.com (accessed 4
December 2016) (Archived at http://www.webcitation.
org/6kyfAscMz).
10. Barratt M. J., Ferris J. A., Winstock A. R. Use of Silk Road, the
online drug market-place, in the United Kingdom, Australia
and the United States. Addiction 2014; 109:774–83.
11. Martin J. Lost on the Silk Road: online drug distribution and
the ‘cryptomarket’.Criminol Crim Just 2014; 14:351–67.
12. Caudevilla F., Ventura M., Fornis I., Barratt M. J., Vidal C.,
Lladanosa C. G. et al. Results of an international drug testing
service for cryptomarket users. Int J Drug Policy 2016; 35:
38–41.
13. Winstock A. R. Global Drug Survey. 2016-10-03. Available
at: http://www.globaldrugsurvey.com (accessed 8 May
2016) (Archived at http://www.webcitation.org/6kyfAscMz).
14. Mountenay J., Oteo A., Griffiths P. The Internet and Drug
Markets (Insights 21). European Monitoring Centre for
Drugs and Drug Addiction (EMCDDA): Luxembourg; 2016,
pp. 13–17; 69–75.
15. Martin J. Drugs on the Dark Net: How Cryptomarkets are
Transforming the Global Trade in Illicit Drugs. Basingstoke:
Palgrave Macmillan; 2014, pp. 36–41.
16. Soska K., Christin N. Measuring the longitudinal evolution of
the online anonymous marketplace ecosystem. Washington,
DC: 24th USENIX Security Symposium (USENIX Security
15); 12–14 August 2015.
17. Aldridge J., Décary-Hety D. Not an ‘Ebay for Drugs: The
Cryptomarket Silk Road as a Paradigm Shifting Criminal Innova-
tion. Manchester: University of Manchester; 2014, pp. 2–20.
18. Orsolini L., Francesconi G., Papanti D., Giorgetti A., Schifano
F. Profiling online recreational/prescription drugs’customers
and overview of drug vending virtual marketplaces. Hum
Psychopharmacol 2015; 30:302–18.
19. Bancroft A., Scott R. P. Concepts of illicit drug quality among
darknet market users: Purity, embodied experience, craft and
chemical knowledge. Int J Drug Policy 2016; 35:42–9.
20. Vogels N., Brunt T. M., Rigter S., van Dijk P., Vervaeke H.,
Niesink R. J. Content of ecstasy in the Netherlands:
1993–2008. Addiction 2009; 104:2057–66.
8Daan van der Gouwe et al.
© 2016 Society for the Study of Addiction Addiction
21. Brunt T. M., Niesink R. J. The Drug Information and Monitor-
ing System (DIMS) in the Netherlands: implementation,
results, and international comparison. Drug Test Anal 2011;
3:621–34.
22. Brunt T. M., Koeter M. W., Niesink R. J., van den Brink W.
Linking the pharmacological content of ecstasy tablets to
the subjective experiences of drug users. Psychopharmacology
(Berl) 2012; 220:751–62.
23. Brunt T. M., Nagy C., Bucheli A., Martins D., Ugarte M.,
Beduwe C. et al. Drug testing in Europe: monitoring results
of the Trans European Drug Information (TEDI) project. Drug
Test Anal 2016; DOI: 10.1002/dta.1954.
24. Benschop A., Rabes M., Korf D. J., Eggerth H. Pill Testing,
Ecstasy and Prevention: A Scientific Evaluation in Three Euro-
pean Cities. Amsterdam: Rozenberg Publishers; 2002, pp.
29–38.
25. Korf D. J., Benschop A., Brunt T. M., Dallas M. Pillen testen in
Nederland: een onderzoek naar versterking van de monitor
uitgaansdrugs [Pills Tests in the Netherlands: a Study to
Strengthen the Monitoring of Party Drugs].Amsterdam:
Rozenberg Publishers; 2003, pp. 39–45.
26. Brunt T. M., Rigter S., Hoek J., Vogels N., van Dijk P., Niesink
R. J. An analysis of cocaine powder in the Netherlands: con-
tent and health hazards due to adulterants. Addiction 2009;
104:798–805.
27. van Laar M. W., van Ooyen-Houben M. M. J., Cruts A. A. N.,
Meijer R. F., Croes E. A., Ketelaars A. P. M. National Drugs
Monitor. Annual Report 2015. Trimbos-instituut: Utrecht;
2015, p. 36.
28. BuchananJ.A.,VogelJ.A.,EberhardtA.M.Levamisole-in-
duced occlusive necrotizing vasculitis of the ears after use of
cocaine contaminated with levamisole. J Med Toxicol 2011;
7:83–4.
29. Gaertner E. M., Switlyk S. A. Dermatologic complications
from levamisole-contaminated cocaine: a case report and
review of the literature. Cutis 2014; 93:102–6.
30. Vevelstad M., Oiestad E. L., Middelkoop G., Hasvold I., Lilleng
P., Delaveris G. J. et al. The PMMA epidemic in Norway:
comparison of fatal and non-fatal intoxications. Forensic Sci
Int 2012; 219:151–7.
31. Van der Gouwe D. Jaarbericht 2014. Drugs Informatie en Mon-
itoring Systeem (DIMS). Utrecht: Trimbos-instituut; 2015.
32. European Monitoring Centre for Drugs and Drug Addiction
(EMCDDA) European Drug Report 2015. Trends and Develop-
ments. Luxembourg: EMCDDA, 2016, pp. 27–30.
33. Van der Gouwe D. Annual Report 2015. Drugs Information and
Monitoring System (DIMS). Utrecht: Trimbos-instituut; 2016.
34. Belackova V., Maalste N., Zabransky T., Grund J. P. Should I
buy or should I grow? How drug policy institutions and drug
market transaction costs shape the decision to self-supply
with cannabis in the Netherlands and the Czech Republic.
Int J Drug Policy 2015; 26:296–310.
35. European Monitoring Centre for Drugs and Drug Addiction
(EMVDDA). Europol European Drug Markets Report. In-depth
Analysis. Lisbon/The Hague: EMCDDA; 2016, p. 116.
36. Van Buskirk J., Naicker S., Roxburgh A., Bruno R., Burns L.
Who sells what? Country specific differences in substance
availability on the Agora cryptomarket. Int J Drug Policy
2016; 35:16–23.
37. Aldridge J., Decary-Hetu D. Hidden wholesale: the drug diffus-
ing capacity of online drug cryptomarkets. Int J Drug Policy
2016; 35:7–15.
38. Barratt M. J., Lenton S., Maddox A., Allen M. What if you live
on top of a bakery and you like cakes? Drug use and harm
trajectories before, during and after the emergence of Silk
Road . Int J Drug Policy 2016; 35:50–7.
On-line and off-line drug quality and price 9
© 2016 Society for the Study of Addiction Addiction
