European Surveillance of Antimicrobial Consumption (ESAC):
outpatient penicillin use in Europe (1997–2009)
Ann Versporten
1
*†, Samuel Coenen
1,2
†, Niels Adriaenssens
1,2
, Arno Muller
1
, Girma Minalu
3
, Christel Faes
3
,
Vanessa Vankerckhoven
1
, Marc Aerts
3
, Niel Hens
3,4
, Geert Molenberghs
3,5
and Herman Goossens
1
on behalf of the ESAC Project Group
1
Laboratory of Medical Microbiology, Vaccine & Infectious Disease Institute (VAXINFECTIO), University of Antwerp, Antwerp, Belgium;
2
Centre for General Practice, Vaccine & Infectious Disease Institute (VAXINFECTIO), University of Antwerp, Antwerp, Belgium;
3
Interuniversity Institute for Biostatistics and Statistical Bioinformatics (I-BIOSTAT), University of Hasselt, Hasselt, Belgium;
4
Centre for
Health Economics Research and Modelling Infectious Diseases (CHERMID), Vaccine & Infectious Disease Institute (VAXINFECTIO),
University of Antwerp, Antwerp, Belgium;
5
Interuniversity Institute for Biostatistics and Statistical Bioinformatics (I-BIOSTAT),
Catholic University of Leuven, Leuven, Belgium
*Corresponding author. Tel: +32-3-265-2418; Fax: +32-3-265-2752; E-mail: ann.versporten@ua.ac.be
†These authors contributed equally to this work.
Background: Data on 13 years (1997–2009) of outpatient penicillin use were collected from 33 European
countries within the European Surveillance of Antimicrobial Consumption (ESAC) project and analysed
in detail.
Methods: For the period 1997–2009, data on outpatient use of systemic penicillins aggregated at the level of
the active substance were collected using the Anatomical Therapeutic Chemical (ATC)/defined daily dose (DDD)
method (WHO, version 2011) and expressed in DDD per 1000 inhabitants per day (DID). For detailed analysis of
trends over time, seasonal variation and composition of outpatient penicillin use in 33 European countries, we
distinguished between narrow-spectrum penicillins (NSP), broad-spectrum penicillins (BSP), penicillinase-
resistant penicillins (PRP) and combinations with b-lactamase inhibitors (COP).
Results: Total outpatient penicillin (ATC group J01C) use in 2009 varied by a factor of 3.8 between the countries
with the highest (16.08 DID in France) and lowest (4.23 DID in the Russian Federation) use. COP represented
45.8%, BSP 40.7%, NSP 10.8% and PRP 2.6% of total European outpatient penicillin use. Total outpatient
penicillin use significantly increased over time by 1.53 (SD 0.71) DID between 1997 and 2009. COP (mainly
co-amoxiclav) increased by 2.17 (SD 0.40) DID, which was the result of its absolute increase as well as the
observed shift from NSP and BSP towards COP. This increase exceeded 10% in 20 countries, where it coincided
with a similar decrease in either BSP (15 countries) or NSP (5 countries).
Conclusions: Penicillins represented the most widely used antibiotic subgroup in all 33 participating countries,
albeit with considerable variation in their use patterns. For Europe, a continuous increase in overall penicillin use
and of COP use was observed during the period 1997–2009.
Keywords: antibiotic use, co-amoxiclav, pharmacoepidemiology, ambulatory care
Introduction
This paper presents data from the European Surveillance of
Antimicrobial Consumption (ESAC) project and reports on
the use of penicillins in 2009, classified into five chemical
subgroups based on the Anatomical Therapeutic Chemical
(ATC) classification. It also reviews temporal trends, seasonal
variation and composition of outpatient penicillin use in data
collected for the 1997–2009 period from 33 European
countries.
Methods
In 2009, 35 countries were included in the ESAC project, of which 33
countries provided valid data. The methods for collecting use data for
systemic antibiotics were described in the introductory paper of this
series.
1
For the period 1997–2009, data on outpatient antibiotic use,
aggregated at the level of the active substance, were collected in accord-
ance with the ATC classification and defined daily dose (DDD) measure-
ment unit (WHO, version 2011).
2
Outpatient antibiotic use data for
2009, expressed in DDD per 1000 inhabitants per day (DID), was available
for 32 European countries, including Cyprus and Lithuania, that provided
# The Author 2011. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: journals.permissions@oup.com
J Antimicrob Chemother 2011; 66 Suppl 6: vi13–vi23
doi:10.1093/jac/dkr454
vi13
Downloaded from https://academic.oup.com/jac/article/66/suppl_6/vi13/679709 by guest on 21 August 2022
total use data. In addition, 2004 data for Switzerland are presented.
Beside the DID outcome measurement unit, the number of packages/
1000 inhabitants/day (PID) was utilized (n¼ 17 countries). Package
data became available from 2006. The calculated DID/PID ratio
allowed assessment of the number of DDD available per package.
The WHO Collaborating Centre for Drug Statistics Methodology has
assigned 44 unique ATC codes for penicillins; no new substances were
added compared with earlier descriptions of outpatient penicillin use in
Europe (Table 1).
3
The ATC group J01C (b-lactam antibacterials, penicil-
lins) comprises penicillins, b-lactamase inhibitors and their combinations
categorized into five chemical subgroups. Since b-lactamase inhibitors
are given in conjunction with broad-spectrum penicillins, outpatient
use of four main subgroups of penicillins was analysed in this paper:
broad-spectrum penicillins (BSP; ATC subgroup J01CA); narrow-spectrum
penicillins (NSP; J01CE); penicillinase-resistant penicillins (PRP; J01CF);
and combinations of penicillins with b-lactamase inhibitors (COP;
J01CR). Importantly, the DDD for the parenteral use of amoxicillin and
enzyme inhibitor (J01CR02) changed from 1 g to 3 g in 2005. To enable
comparability for subsequent years, ESAC has updated data for the
previous years.
The methods used for analysing temporal trends, seasonal variation
and changes in composition of outpatient penicillin use are described
in two tutorials in this series.
4,5
Results
Outpatient penicillin use in 2009
Five penicillins represented 96.1% of the total outpatient penicillin
use in 2009 in Europe: amoxicillin and enzyme inhibitor
(co-amoxiclav) (44.9%); amoxicillin (39.0%); phenoxymethylpeni-
cillin (8.9%); flucloxacillin (1.7%); and pivmecillinam (1.6%). Other
substances represented ,1% each, and no use at all was
recorded for 11 substances (Table 1).
Penicillins represented 47% of total outpatient use in 2009.
1
Figure 1 shows total outpatient penicillin use subdivided into
the four main subgroups in 2009 for 33 European countries,
including two countries (Cyprus and Lithuania) with total use
data and Swiss data for 2004, expressed in DID. Total penicillin
use varied by a factor of 3.8 between the countries with the
highest (16.1 DID in France) and lowest (4.2 DID in the Russian
Federation) use.
NSP represented 10.8% of total European outpatient penicillin
use. Large variations in outpatient NSP use were found, ranging
from 5.1 DID in Denmark to 0.003 DID in Italy. Phenoxymethyl-
penicillin, commonly known as penicillin V, was the most widely
used NSP in most countries. Its use represented .50% of total
outpatient penicillin use in Denmark (5.1 DID), Norway (3.6
DID) and Sweden (3.9 DID). In contrast, this substance repre-
sented ,1% of total penicillin use in Belgium, Cyprus, France,
Israel, Luxembourg, Malta, the Netherlands, the Russian Feder-
ation and Spain (representing ,0.07 DID). No use was reported
for Croatia, Italy and Portugal. Instead, a wide variety of
other NSP substances were reported. For example, benzathine
phenoxymethylpenicillin was mainly used in Austria, Croatia
and the Czech Republic, and penamecillin in Hungary and
Slovakia. Pheneticillin was exclusively used in the Netherlands.
The long-acting parenteral benzylpenicillin was mainly used in
the Czech Republic, Greece and Lithuania.
BSP represented 40.7% of total European outpatient penicillin
use. Outpatient BSP use ranged from 9.0 DID in France to 1.1 DID
in Malta. Amoxicillin was the most commonly used BSP. The
Table 1. Classification of penicillins and their combinations (ATC classification,
2011 version)
Penicillins with extended spectrum
J01CA01 ampicillin
a
J01CA02 pivampicillin
J01CA03 carbenicillin
b
J01CA04 amoxicillin
J01CA05 carindacillin
b
J01CA06 bacampicillin
J01CA07 epicillin
b
J01CA08 pivmecillinam
J01CA09 azlocillin
J01CA10 mezlocillin
J01CA11 mecillinam
J01CA12 piperacillin
J01CA13 ticarcillin
J01CA14 metampicillin
b
J01CA15 talampicillin
b
J01CA16 sulbenicillin
b
J01CA17 temocillin
b
J01CA18 hetacillin
b
J01CA20 combinations
b
J01CA51 ampicillin, comb.
b-Lactamase-sensitive penicillins
J01CE01 benzylpenicillin
J01CE02 phenoxymethylpenicillin
J01CE03 propicillin
J01CE04 azidocillin
J01CE05 pheneticillin
J01CE06 penamecillin
J01CE07 clometocillin
J01CE08 benzathine benzylpenicillin
J01CE09 procaine penicillin
J01CE10 benzathine phenoxymethylpenicillin
J01CE30 combinations
b-Lactamase-resistant penicillins
J01CF01 dicloxacillin
a
J01CF02 cloxacillin
J01CF03 methicillin
b
J01CF04 oxacillin
J01CF05 flucloxacillin
b-Lactamase inhibitors
J01CG01 sulbactam
J01CG02 tazobactam
b
Combinations of penicillins with b-lactamase inhibitor
J01CR01 ampicillin and enzyme inhibitor
J01CR02 amoxicillin and enzyme inhibitor
J01CR03 ticarcillin and enzyme inhibitor
J01CR04 sultamicillin
J01CR05 piperacillin and enzyme inhibitor
J01CR50 combinations of penicillins
Bold type indicates that use represented .1% of total penicillin use in
2009.
a
Use represented .1% of total penicillin use in 2003.
b
No use of this penicillin was reported in 2009.
Versporten et al.
vi14
Downloaded from https://academic.oup.com/jac/article/66/suppl_6/vi13/679709 by guest on 21 August 2022
highest proportional outpatient BSP use was observed for the
Russian Federation (75.9%), followed by Estonia, Germany,
Latvia and Lithuania (.65%) (range 2.9–7.4 DID). Propor-
tional use was .50% in 11 countries. Malta had the lowest pro-
portional use (12.3%) (1.1 DID). Pivmecillinam was commonly
used in the Nordic countries Norway, Denmark, Finland and
Sweden as well as in Israel (20.3%, 15.7%, 11.2%, 7.6% and
5.2%, respectively). The use of ampicillin represented 9.1% (0.4
DID) of total outpatient penicillin use in the Russian Federation
and 7.7% (0.8 DID) of total outpatient penicillin use in Lithuania.
Bacampicillin was only used in Italy and pivampicillin only in
Denmark.
COP represented 45.8% of total European outpatient penicillin
use. In 2009, the highest use of COP was observed in Cyprus
(10.1 DID) (total care data) followed by Italy (9.8 DID) and the
lowest in Norway (0.002 DID). The most commonly used COP sub-
stance was amoxicillin and enzyme inhibitor (co-amoxiclav). Its
proportional use ranged from 86.6% in Malta (7.9 DID) to 0.03%
in Norway (0.002 DID). Co-amoxiclav use represented .50% of
total penicillin use in Austria (4.5 DID), Belgium (8.9 DID), Croatia
(5.1 DID), Cyprus (10.0 DID; hospital use data are included),
Hungary (4.6 DID), Ireland (5.5 DID), Italy (9.8 DID), Luxembourg
(8.4 DID), Portugal (9.0 DID), Romania (2.3 DID), Slovakia (5.2
DID) and Spain (7.6 DID). In contrast, a low proportional use of
co-amoxiclav was seen in Denmark (0.4 DID), Germany (0.2 DID)
and Sweden (0.2 DID) (all ,6%). It represented 10%–20% of
total penicillin use in Finland (1.2 DID), Lithuania (1.6 DID) and
the UK (1.1 DID). Sultamicillin represented .1% of total outpatient
penicillin use in Estonia, Germany, Lithuania and Slovakia.
PRP represented 2.6% of total European outpatient penicillin
use. Their proportional use in 2009 varied from 20.5% in Sweden
to no use in Hungary and Bulgaria. PRP represented .1% of
total outpatient penicillin use in 14 countries. All these countries
only reported the use of one particular substance except for
France. Flucloxacillin was used in Sweden, the UK, Ireland, the
Netherlands, Portugal, Belgium and Luxembourg (20.5%, 14.2%,
9.2%, 8.4%, 4.6%, 1.7% and 1.4%, respectively); dicloxacillin in
Denmark, Iceland and Norway (11.3%, 9.2% and 8.4%, respect-
ively); cloxacillin in Slovenia, Spain and France (1.6%, 1.6% and
1.5%, respectively); and oxacillin in Romania and France (3.2%
and 1.1%, respectively). Methicillin use was not recorded in 2009.
Figure 2 shows total outpatient penicillin use expressed in PID
for 17 countries for 2009 (Italy, 2008 data; Ireland and the Czech
Republic, 2007 data). Based on package data, Italy showed the
highest use of penicillin (3.4 PID), followed by Greece, Ireland
and the Russian Federation (1.5, 1.3 and 1.3 PID, respectively).
Lowest PID figures were seen for the Netherlands, Estonia and
Sweden (0.6–0.7 PID). Based on ranking, the Russian Federation
shifted from position 17 in DID (low-prescribing country) to pos-
ition 4 in PID (high-prescribing country), Portugal from position 4
in DID to position 9 in PID and Belgium from position 2 in DID to
position 5 in PID. The ranking was similar for the Czech Republic,
Denmark, Estonia and Italy. Consequently, the lowest mean DDD
per package was found in the Russian Federation (3.3 DDD/pack)
16
14
12
10
8
6
4
DID
2
0
Cyprus
Italy
Luxembourg
Greece
Belgium
Spain
Slovakia
Slovenia
Russian Fed.
France
Bulgaria
UK
Switzerland
Hungary
Malta
Germany
Israel
Croatia
Austria
Czech Rep.
Romania
Poland
Portugal
Lithuania
Netherlands
Finland
Latvia
Sweden
Estonia
Iceland
Ireland
Denmark
Norway
b-Lactamase-sensitive penicillins (narrow-spectrum penicillins; J01CE)
Penicillins with extended spectrum (broad-spectrum penicillins; J01CA)
Combinations of penicillins, incl. b-lactamase inhibitors (J01CR)
b-Lactamase-resistant penicillins (J01CF)
Figure 1. Outpatient use of penicillins in 33 European countries in 2009 in DID (2004 data for Switzerland). For Cyprus and Lithuania, total care data
are reported.
Outpatient penicillin use in Europe (ESAC)
vi15
JA
C
Downloaded from https://academic.oup.com/jac/article/66/suppl_6/vi13/679709 by guest on 21 August 2022
and the highest in Belgium (12.1 DDD/pack), followed by Portugal
and Sweden (10.8 and 10.6 DDD/pack, respectively). The mean
DDD for a penicillin package was higher in all countries compared
with the overall mean DDD for an antibiotic package (J01),
except for the northern countries Denmark, Finland and Sweden.
Longitudinal data analysis (1997–2009)
For Europe, a significant increase in total outpatient penicillin use
of 0.03 (SD 0.01) DID per quarter was found, starting from 8.09
(SD 0.62) DID in the first quarter of 1997, and there was a signifi-
cant seasonal variation with an amplitude of 1.87 (SD 0.20) DID,
which decreased non-significantly over time by 0.001 (SD 0.004)
DID per quarter (Figure 3). Furthermore, the longitudinal analysis
showed that both upward winter and downward summer peaks
of outpatient penicillin consumption shifted significantly from
one year to another and that there was a positive correlation
between volume of use and seasonal variation. This means
that, in terms of absolute amount, high and low penicillin-
consuming countries tend to have high and low seasonal
variation in penicillin use, respectively.
Table 2 provides an overview of outpatient penicillin use for all
33 participating European countries between 1997 and 2009.
Outpatient use of penicillins was stable or increasing in most
countries. The highest continuous increase since 2003 was
observed in the Russian Federation, with a rise of 100% (from
2.1 DID to 4.2 DID), followed by Austria, showing an increase
of 62% (from 4.4 DID to 7.1 DID), and Belgium, with an increase
of 61% (from 9.4 DID to 15.1 DID). In Ireland, Luxembourg and
Italy, penicillin use steadily increased by .40%. In Slovakia,
however, a large drop of 7 DID (a decrease of 76%) was observed
in overall penicillin use between its peak in 2001 and 2009.
Penicillin use decreased in Slovakia by 76% since 2001 (from
16.8 DID to 9.6 DID), in Lithuania by 49% since 2006
(from 15.1 DID to 10.1 DID) and in Croatia by 27% since 2003
(from 12.3 to 9.7). In the other countries, no remarkable
trends were observed.
The overall increase in penicillin use was mainly the result of
the increased use of COP. For Europe, a significant increase in
total COP use of 0.04 (SD 0.01) DID per quarter was found, start-
ing from 1.98 (SD 0.34) DID in the first quarter of 1997, and there
was a significant seasonal variation with an amplitude of 0.58
3.5
3
2.5
2
1.5
1
0.5
0
PID
Italy
Greece
Belgium
Slovenia
Russian Fed.
Bulgaria
Croatia
Austria
Czech Rep.
Portugal
Lithuania
Netherlands
Finland
Sweden
Estonia
Ireland
Denmark
b-Lactamase-sensitive penicillins (narrow-spectrum penicillins; J01CE)
Penicillins with extended spectrum (broad-spectrum penicillins; J01CA)
Combinations of penicillins, incl. b-lactamase inhibitors (J01CR)
b-Lactamase-resistant penicillins (J01CF)
Country BE PT SE AT CZ HR LT DK GR IE BG FI NL SI EE IT RU
Ranking PID
5 9 15 14 13 10 8 7 2 3 11 12 17 6 16 1 4
Ranking DID
2 4 12 11 13 8 6 7 3 5 10 14 15 9 16 1 17
DDD/package J01C
12.1 10.8 10.6 10.4 9.9 9.2 9.1 8.9 8.8 8.7 8.6 8.3 7.9 7.7 6.7 4.5 3.3
DDD/package J01
10.9 9.0 11.8 8.1 8.1 7.4 6.6 9.3 7.3 8.6 5.7 9.5 7.4 6.8 6.5 2.6 3.0
Figure 2. Total outpatient penicillin use in 17 European countries in 2009 in PID, the ranking in DID versus PID, and the number of DDD per outpatient
penicillin package. For Lithuania, total care data are used. For Italy, 2008 data are used. For the Czech Republic and Ireland, 2007 data are used.
Countries are shown in order (left to right) of decreasing DDD/package J01C, as shown in the table. AT, Austria; BE, Belgium; BG, Bulgaria; CZ,
Czech Republic; DK, Denmark; EE, Estonia; FI, Finland; GR, Greece; HR, Croatia; IE, Ireland; IT, Italy; LT, Lithuania; NL, Netherlands; PT, Portugal; RU,
Russian Federation; SE, Sweden; SI, Slovenia.
Versporten et al.
vi16
Downloaded from https://academic.oup.com/jac/article/66/suppl_6/vi13/679709 by guest on 21 August 2022
(SD 0.12) DID, which increased significantly over time by 0.004
(SD 0.002) DID per quarter (Figure 4). This trend was observed
for most countries, mainly due to an increase in co-amoxiclav
use. Countries with an initial low use of co-amoxiclav increased
their use more than 3-fold; these were Austria, Bulgaria,
Estonia, Finland, Germany, Norway and Poland. Almost all
other countries doubled their co-amoxiclav use over the years.
Figures S1 and S2 (available as Supplementary data at JAC
Online) show the seasonal variation of outpatient penicillin
use in 27 European countries able to deliver quarterly data
(no quarterly data were available for Bulgaria, France, Malta
and Norway). Seasonal variation was less pronounced in low
penicillin-consuming countries (Denmark, Iceland, the Nether-
lands, Sweden and the UK) compared with higher penicillin-
consuming countries (Austria, Belgium, Portugal and Spain).
Iceland and Sweden in Northern Europe and the Czech Republic
decreased both their total penicillin use and seasonal variation
over time. Austria, Belgium, Italy, Lithuania, Luxembourg,
Poland and the Russian Federation increased both their overall
and seasonal variation over time. For the other countries differ-
ent yearly and seasonal variations were observed.
Seasonal variations of penicillin use correlated to a high degree
with the total antibiotic use described in the introductory paper in
all but two countries; the UK showed substantially higher seasonal
variation of penicillins compared with all antibiotics, whereas in
the Netherlands the seasonality of penicillin use was lower than
for antibiotics in general, mainly due to the increased use of tetra-
cyclines (country-specific results not shown).
Compositional data analysis (1997–2009)
For Europe, higher outpatient penicillin use resulted in a signifi-
cant increase in the proportional use of COP relative to NSP
and BSP. Proportional use of NSP was significantly smaller relative
to the other three main subgroups when total penicillin use was
higher (Table 3). The relative use of COP and BSP significantly
increased over time with respect to that of NSP and PRP, and
the relative use of COP significantly increased over time with
respect to BSP (Table 4).
Trends of proportional outpatient penicillin use in individual
countries are shown in Figure S3 (available as Supplementary
data at JAC Online). The proportion of COP increased continuous-
ly in most European countries between 1997 and 2009. This
increase exceeded 10% in 20 countries, where it coincided
with a similar decrease in either BSP in 15 countries or NSP in
5 countries. Its proportional use was stable in Spain, France,
Luxembourg and Slovenia since 2005–06. Denmark, Germany
and Sweden had the lowest proportional COP use since 1997
(,10%). COP use in Norway was negligible.
The Northern European countries Denmark, Norway and
Sweden had the highest proportional use of NSP, with a con-
stant 60% of total penicillin use over time, but their proportion
decreased over time. In contrast, proportional NSP use was
low since 1997 in Belgium, France, Italy, Latvia,
Luxembourg, Malta, Portugal and Spain (,2%). NSP use in
the Russian Federation and Poland decreased to ,2% over
the last 3 years.
0
5
10
15
20
25
30
Time (1997–2009)
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Observed mean
Predicted mean
Trend
8.0902 (0.6158)* 0.0295 (0.0136)* 1.8696 (0.1976)* −0.0014 (0.0045) 0.3825 (0.0234)*
DID
Parameters
b
0
b
1
b
0
S
b
1
S
d
Figure 3. Estimated linear trend and seasonal variation of outpatient penicillin use based on available quarterly data for 1997–2009.
b
0
(intercept),
predicted average outpatient use in the first quarter of 1997;
b
1
(slope), predicted average increase (if positive)/decrease (if negative) in use per
quarter;
b
0
S
(seasonal variation), predicted average amplitude of the upward winter and downward summer peak in use;
b
1
S
(damping effect),
predicted average increase (if positive)/decrease (if negative) of the amplitude of the upward winter and downward summer peak in use per
quarter;
d
(phase shift), shift in timing of the upward winter and downward summer peak from one year to another. *Significant (P, 0.05).
Outpatient penicillin use in Europe (ESAC)
vi17
JA
C
Downloaded from https://academic.oup.com/jac/article/66/suppl_6/vi13/679709 by guest on 21 August 2022