Import and spread of extended-spectrum beta-
lactamase-producing Enterobacteriaceae by
international travellers (COMBAT study): a
prospective, multicentre cohort study
Citation for published version (APA):
Arcilla, M. S., van Hattem, J. M., Haverkate, M. R., Bootsma, M. C. J., van Genderen, P. J. J., Goorhuis,
A., Grobusch, M. P., Oude Lashof, A., Molhoek, N., Schultsz, C., Stobberingh, E. E., Verbrugh, H. A., de
Jong, M. D., Melles, D. C., & Penders, J. (2017). Import and spread of extended-spectrum beta-
lactamase-producing Enterobacteriaceae by international travellers (COMBAT study): a prospective,
multicentre cohort study. Lancet Infectious Diseases, 17(1), 78-85. https://doi.org/10.1016/S1473-
3099(16)30319-X
Document status and date:
Published: 01/01/2017
DOI:
10.1016/S1473-3099(16)30319-X
Document Version:
Publisher's PDF, also known as Version of record
Document license:
Taverne
Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can
be important differences between the submitted version and the official published version of record.
People interested in the research are advised to contact the author for the final version of the publication,
or visit the DOI to the publisher's website.
• The final author version and the galley proof are versions of the publication after peer review.
• The final published version features the final layout of the paper including the volume, issue and page
numbers.
Link to publication
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright
owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these
rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above,
please follow below link for the End User Agreement:
www.umlib.nl/taverne-license
Take down policy
If you believe that this document breaches copyright please contact us at:
repository@maastrichtuniversity.nl
providing details and we will investigate your claim.
Download date: 10 Aug. 2022
78
www.thelancet.com/infection Vol 17 January 2017
Articles
Lancet Infect Dis 2017;
17: 78–85
Published Online
October 14, 2016
http://dx.doi.org/10.1016/
S1473-3099(16)30319-
X
See Comment
page 8
*Contributed equally
Department of
Medical
Microbiology and Infectious
Diseases, Erasmus University
Medical Centre,
Rotterdam,
Netherlands (M S Arcilla MD,
Prof H A Verbrugh PhD,
D C Melles PhD); Department of
Medical Microbiology,
Academic Medical Centre,
Amsterdam, Netherlands
(J M van Hattem MD,
C Schultsz PhD,
Prof M D de Jong PhD); Julius
Centre for Health Sciences and
Primary Care, University
Medical Centre Utrecht,
Utrecht, Netherlands
(M R Haverkate PhD,
M C J Bootsma PhD);
Department of Mathematics,
Faculty of Science, Utrecht
University, Utrecht,
Netherlands (M C J Bootsma);
Institute for Tropical Diseases,
Havenziekenhuis, Rotterdam,
Netherlands
(P J J van Genderen PhD,
N Molhoek MSc); Centre of
Tropical Medicine and Travel
Medicine, Academic Medical
Centre, University of
Amsterdam, Amsterdam,
Netherlands (A Goorhuis PhD,
Prof M P Grobusch PhD); and
School for Public Health and
Primary Care (Caphri),
Department of Medical
Microbiology
(A M Oude Lashof PhD,
E E Stobberingh PhD,
J Penders PhD) and School for
Nutrition and Translational
Research in Metabolism
(NUTRIM) (J Penders),
Maastricht University Medical
Centre, Maastricht,
Netherlands
Import and spread of extended-spectrum β-lactamase-
producing Enterobacteriaceae by international travellers
(COMBAT study): a prospective, multicentre cohort study
Maris S Arcilla*, Jarne M van Hattem*, Manon R Haverkate, Martin C J Bootsma, Perry J J van Genderen, Abraham Goorhuis, Martin P Grobusch,
Astrid M Oude Lashof, Nicky Molhoek, Constance Schultsz, Ellen E Stobberingh, Henri A Verbrugh, Menno D de Jong, Damian C Melles, John Penders
Summary
Background International travel contributes to the dissemination of antimicrobial resistance. We investigated the
acquisition of extended-spectrum β-lactamase-producing Enterobacteriaceae (ESBL-E) during international travel,
with a focus on predictive factors for acquisition, duration of colonisation, and probability of onward transmission.
Methods Within the prospective, multicentre COMBAT study, 2001 Dutch travellers and 215 non-travelling
household members were enrolled. F
aecal samples and questionnaires on demographics, illnesses, and behaviour
were collected before travel and immediately and 1, 3, 6, and 12 months af
ter return. Samples were screened for the
presence of ESBL-E. In post-travel samples, ESBL genes were sequenced and PCR with specifi c primers for
plasmid-encoded β-lactamase enzymes TEM, SHV, and CTX-M group 1, 2, 8, 9, and 25 was used to confi rm the
presence of ESBL genes in follow-up samples. Multivariable regression analyses and mathematical modelling were
used to identify predictors for acquisition and sustained carriage, and to determine household transmission rates.
This study is registered with ClinicalTrials.gov, number NCT01676974.
Findings 633 (34·3%) of 1847 travellers who were ESBL negative before travel and had available samples after return
had acquired ESBL-E during international travel (95% CI 32·1–36·5), with the highest number of acquisitions being
among those who travelled to southern Asia in 136 of 181 (75·1%, 95% CI 68·4–80·9). Important predictors for
acquisition of ESBL-E were antibiotic use during travel (adjusted odds ratio 2·69, 95% CI 1·79–4·05), traveller’s
diarrhoea that persisted af
ter return (2·31, 1·42–3·76), and pre-existing chronic bowel disease (2·10, 1·13–3·90). The
median duration of colonisation after travel was 30 days (95% CI 29–33). 65 (11·3%) of 577 remained colonised at
12 months. CTX-M enzyme group 9 ESBLs were associated with a signifi cantly increased risk of sustained carriage
(median duration 75 days, 95% CI 48–102, p=0·0001). Onward trans mission was found in 13 (7·7%) of 168 household
members. The probability of transmitting ESBL-E to another household member was 12% (95% CI 5–18).
Interpretation Acquisition and spread of ESBL-E during and after international travel was substantial and worrisome.
Travellers to areas with a high risk of ESBL-E acquisition should be viewed as potential carriers of ESBL-E for up to
12 months after return.
F
unding Netherlands Organisation for Health Research and Development (ZonMw).
Introduction
Antimicrobial resistance constitutes an increasingly
important human health hazard worldwide.
1
The use of
antibiotics in human beings and food animals is a well
established driving force behind increasing resistance.
2
Given the enormous growth of international tourism,
from 25 million travellers in 1950 to 1·133 billion in 2014,
3
international travel might also contribute substantially to
the rise in resistance because resistant bacteria or bacterial
mobile genetic elements carrying resistance genes (eg,
plasmids) may be rapidly transported between regions.
4
An important part of antimicrobial resistance genes is
found on plasmids and codes for extended-spectrum β
lactamase enzymes ([ESBLs] eg, TEM, SHV, and CTX-M)
and carbapenemases that confer resistance to most
β-lactam antibiotics.
2,4
Additionally, ESBL-producing
Enterobacteriaceae (ESBL-E) and carba penemase-
producing Enterobacteriaceae (CPE) are typically resistant
to multiple other antibiotic classes, which leaves few to no
eff ective antimicrobial agents for prevention and
treatment of infections.
4,5
Previous studies have reported frequent acquisition of
ESBL-E associated with various predictors and sporadic
acquisition of CPE among international travellers.
6–10
However, data on ESBL-E colonisation after travel and
assessment of associated predictors for sustained carriage
and onward transmission within households are very
limited. Such data are needed to establish the public
health risk of the introduction and spread of antimicrobial
resistance by travellers, and the potential needs and
measures to monitor or manage these risks. Identifying
individuals at risk of ESBL-E carriage enables appropriate
measures to be taken to prevent introduction and spread
of ESBL-E or CPE and for empirical adjustment of
antibiotic treatment in individuals to optimise clinical
care. We investigated the acquisition of ESBL-E during
Articles
www.thelancet.com/infection Vol 17 January 2017
79
Correspondence to:
Dr John Penders, Department of
Medical Microbiology,
Maastricht UMC+, PO 5800,
6202 AZ, Maastricht,
Netherlands
j.penders@
maastrichtuniversity.nl
international travel, the associated predictive factors for
acquisition, duration of colonisation, and onward trans-
mission to household members.
Methods
Study design and participants
The study design and methods have been described in
detail elsewhere.
11
Briefl y, we did a multicentre,
longitudinal, prospective cohort study involving travellers
who were followed up from 1–3 weeks before travel
departure until 12 months after return. To study
household transmission, we also assessed non-travelling
household members in the same period.
Eligible participants were adults (age ≥18 years)
planning to travel for at least 1 week and up to 3 months.
They were recruited at three outpatient travel clinics
across the Netherlands from November, 2012, to
November, 2013. The study was approved by the Medical
Research Ethics Committee, Maastricht University
Medical Centre (METC 12-4-093). All participants
provided written informed consent.
Procedures
Participants were provided with faeces collection kits and
instructed to self-collect faecal swabs (appendix) before
and immediately and 1 month after travel. If any of these
samples contained ESBL-E, the traveller and his or her
household members were asked to provide further
samples at 3, 6, and 12 months after travel. If no samples
were positive for ESBL-E, no additional samples were
collected. Questionnaires were also collected at all
timepoints to obtain information on potential risk factors
for ESBL-E acquisition, including demographics,
illnesses, and behaviour before, during, and after travel.
Samples were processed immediately after receipt. They
were inoculated in tryptic soy broth supplemented with
vancomycin (50 mg/L) to select for Entero bacteriaceae.
The broth was then subcultured on chromID ESBL
(bioMérieux, Marcy l’Etoile, France). All morphologically
distinct colonies were characterised to the species level
with matrix-assisted laser desorption/ionisation time-of-
fl ight mass spectrometry (Bruker Microfl ex LT, Bruker,
London, UK). Antibiotic minimum inhibitory
concentrations were measured with the automated
susceptibility testing system Vitek 2 (bioMérieux) for all
Enterobacteriaceae. ESBL production was phenotypically
confi rmed by the combination disc diff usion test,
according to current national Dutch guidelines.
12
All phenotypically confi rmed ESBL-E isolates acquired
during travel were screened for the presence of ESBL
Research in context
Evidence before this study
We searched PubMed on
Aug 17, 2015, with the search terms
“Gram negative bacteria”, “Enterobacteriaceae”, “Escherichia”,
”Klebsiella”, ”Salmonella”, “Shigella”, “Yersinia”, “travel”, “tourist”,
“tourism”, “turista”, “aviation”, “air transport”, “airport”,
“resistance”, “colonisation”, “antibiotic”, “susceptibility”,
“carriage”, and “carrier”. We did a systematic review and
identifi ed 11 eligible studies. We updated this search on
April 14, 2016, and found no new prospective studies. The
results of the 11 prospective cohort studies showed high
acquisition rates of extended-spectrum β-lactamase-producing
Enterobacteriaceae (ESBL-E) among travellers who had returned
from southern Asia and northern Africa. Four travellers who
visited India acquired carbapenemase-producing
Enterobacteriaceae (CPE). However, whether antibiotic use and
traveller’s diarrhoea are predictors for ESBL-E acquisition was
unclear. Moreover, these studies did not suffi ciently address
duration of ESBL-E carriage among travellers or onward
transmission within households. One study asked travellers to
provide stool samples up to 12 months after return, but duration
of carriage was defi ned by ESBL phenotype. One other study
looked at household transmission, but because only
11 household contacts were included, no reliable conclusion
could be inferred about the risk of household transmission.
Added value of this study
In this large-scale, longitudinal cohort study, we followed up
travellers and their non-travelling household members for up
to 12 months after travel. The large sample size meant that
we could investigate ESBL-E acquisition among travellers who
had returned from a large number of countries across the
world, including those such as Uganda, for which community
carriage rates of ESBL-E were previously unknown.
We identifi ed several predictors (some new) for
ESBL-E acquisition, including factors specifi c to subregions.
Moreover, we were able to ascertain duration of
ESBL-E carriage and associated resistance genes, identify
predictors for sustained colonisation, and to model
transmission rates mathematically within households.
Implications of all the available evidence
High frequencies of ESBL-E acquisition during travel,
subsequent sustained carriage, and evidence of onward
transmission within households show that travellers
contribute to the emergence and spread of ESBL-E on a global
scale. Active screening for ESBL-E and CPE and adjustment of
empirical antimicrobial therapy should be considered for
returning travellers at increased risk of ESBL-E carriage.
However, implications for infection prevention and antibiotic
treatment policies will diff er locally because the degree of
consequence of acquisition and spread of ESBL-E by travellers
is highly dependent on local ESBL-E prevalence in the country
of origin.
See Online for appendix
Articles
80
www.thelancet.com/infection Vol 17 January 2017
genes with microarray, as described previously
(appendix). The presence of ESBL genes was confi rmed
by PCR with primers specifi c for CTX-M enzyme groups
1, 2, 8, 9, and 25 and in-house primer sets. Further
characterisation by sequencing was done for the most
prevalent and largest CTX-M groups, 1 and 9. PCR
confi rmation and sequencing of genes for TEM and SHV
ESBLs were limited to isolates that had negative
microarray results for all CTX-M genes. A generic CTX-M
PCR was done if no ESBL genes were detected by
microarray, and, if positive, was followed by specifi c PCR
and sequence confi rmation for the diff erent CTX-M
groups (appendix). Sequences were compared with those
in the NCBI GenBank and Lahey databases.
Acquisition was defi ned as the absence of ESBL-E in
faecal samples before travel and the presence of ESBL-E
in those obtained immediately after travel, as identifi ed
by phenotypic tests. Duration of carriage was defi ned by
the last positive sample harbouring an ESBL of the same
group (TEM, SHV, or CTX-M group 1, 2, 8, 9, or 25, or a
combination) as detected immediately after travel.
Participants with consecutive samples positive for
ESBL-E were classifi ed as being persistent carriers and
those with ESBL-E-positive samples interspersed with at
least one negative sample were classifi ed as being
intermittent carriers.
Statistical analysis
Incidence proportions and incidence per 100 person-days
of travel and accompanying 95% CIs for ESBL-E
acquisition were calculated for each subregion (appendix)
and country of destination. Incidence per 100 person-days
of travel was calculated with a maximum likelihood
method that was based on a constant acquisition rate
with right-censored and interval-censored data.
Predictors for ESBL-E acquisition were determined by
logistic regression models that were based on the method
proposed by Bursac and colleagues
13
(appendix) and
analysed with IBM SPSS Statistics (version 21.0). Results
are presented as odds ratios (ORs) and 95% CIs. We did
separate analyses for the subregions of southeast Asia,
southern Asia, and eastern Africa, as several dietary
variables (eg, consumption of chicken, barbecue meat, or
pork) interacted with specifi c travel destination subregions.
Time to decolonisation was assessed with Kaplan-Meier
survival analyses with right censoring for participants
whose last provided sample was ESBL-E positive.
Univariable and multivariable Cox’s regression analyses
were done to identify predictors associated with de-
colonisation (appendix). Results are presented as hazard
ratios (HRs) and 95% CIs
(HRs <1·00 indicate decreased
risk of decolonisation and, therefore, increased duration
of carriage).
A Markov model was used to calculate the probability
of transmission within households. For computational
reasons, this model was based on ESBL-E as defi ned by
phenotypic confi rmation, and only data from households
consisting of at most fi ve people were included, but these
accounted for 98% of households. The model took into
account false-negative results, missing culture results,
and unobserved colonisation times. The method of
calculation was as follows. ESBL-E-positive people
Travellers
(n=2001)*
Non-travelling
household
members
(n=215)†
Sex
Male 920 (46·0%) 80 (37·2%)
Female 1081 (54·0%) 135 (62·8%)
Age (years) 50·5
(32·8–60·7)
46·9
(25·7-55·8)
Education level
No education, elementary school, or
prevocational secondary education
243 (12·4%) 78 (36·4%)
Vocational secondary education 280 (14·2%) 37 (17·3%)
Senior general secondary education
or education up to university
200 (10·2%) 45 (21·0%)
Higher professional education 642 (32·7%) 53 (24·7%)
Academic (university) education 595 (30·3%) 38 (17·8%)
Antibiotic use in previous 3 months
No 1760 (90·1%) 189 (88·3%)
Yes 194 (9·9%) 25 (11·7%)
Travel in past year
None 185 (9·5%) 27 (12·6%)
In Europe 915 (46·9%) 124 (57·7%)
Outside Europe 852 (43·6%) 64 (29·8%)
Chronic disease‡
No 1500 (77·2%) 173 (82·0%)
Yes 443 (22·8%) 38 (18·0%)
Chronic bowel disease‡
No 1912 (97·4%) 212 (99·1%)
Yes 51 (2·6%) 2 (0·9%)
Continent visited during travel§
Asia 1016 (50·8%) NA
Africa 633 (31·6%) NA
America 326 (16·3%) NA
Europe 21 (1·0%) NA
Oceania 5 (0·2%) NA
Duration of index travel (days) 20
(15·0–25·0)
NA
Purpose of index travel
Holiday 1655 (84·2%) NA
Work or internship 161 (8·2%) NA
Visiting family or relatives 82 (4·2%) NA
Other reason 66 (3·4%) NA
Data are number (%) or median (IQR). NA=not applicable. *Some numbers do not
add up to 2001 because of missing data. †Some numbers do not add up to
215 because of missing data. ‡Self-reported by traveller or household member.
§If travellers visited multiple continents, only the main continent visited is
presented in this table.
Table 1: Baseline characteristics of travellers and non-travelling
household members
Articles
www.thelancet.com/infection Vol 17 January 2017
81
(travellers or non-travelling household members)
transmit ESBL-E to household members with rate β.
Transmission from other sources was incorporated by
the background transmission parameter α. De-
colonisation of ESBL-E occurred with rate γ. Negative
cultures could be false negative and aff ect the estimate of
the sensitivity (φ). The specifi city of culture was assumed
to be 100%. Thus, the probability of transmission from
an ESBL-E-positive to an ESBL-E-negative person, given
that the ESBL-E-negative household member did not
acquire ESBL-E via another route, could be calculated as
β/(β+γ). Model parameters were simultaneously
estimated with a maximum likelihood method in
Mathematica version 9.0. This study is registered with
ClinicalTrials.gov, number NCT01676974.
Role of the funding source
The funder of the study had no role in the study design,
data collection, data analysis, data interpretation, or
writing of the report. The corresponding author had full
access to all the data in the study and had fi nal
responsibility for the decision to submit for publication.
Results
2737 travellers were screened for eligibility, of whom
2001 were included in the study (appendix), with median
age 50·5 years (IQR 32·8–60·7) and good health before
travelling in most (table 1). 49 travellers were lost to
follow-up.
The main purpose for travel was tourism (1655 [84·2%]
of 1965 travellers) and the median travel duration was
20 days (IQR 15·0–25·0; table 1). The subregions most
frequently visited were southeast Asia (n=650), eastern
Africa (n=287), South America (n=228), and southern
Asia (n=217). 122 (6·1%) of 2001 travellers were carrying
ESBL-E before travel, leaving 1879 at risk of ESBL-E
acquisition. 1847 (98·3%) of these submitted faecal
samples after travel, among whom 633 had acquired at
least one ESBL-E during travel (table 2), giving an
acquisition rate of 34·3% (95% CI 32·1–36·5). From
these 633 travellers, 859 morphologically diff erent
ESBL-E strains were isolated (759 Escherichia coli, 67
Klebsiella pneumoniae, and 33 other species). CTX-M-15
was the most frequently acquired ESBL gene, being
found in 338 (53·4%) of 633 travellers (appendix).
ESBL-E were most frequently acquired in southern
Asia (75·1%, 95% CI 68·4–80·9), followed by central and
eastern Asia (48·8%, 38·4–59·3; table 2, fi gure 1), but the
frequency of acquisition varied widely between countries.
Among the 22 most frequently visited countries,
acquisition was highest in India (88·6%, 95% CI
79·8–93·9) and lowest in Suriname (3·6%, 1·0–12·1;
appendix). Acquisition was also common after travel to
eastern African countries, such as Uganda (44·4%,
27·6–62·7, appendix).
In the multivariable logistic regression, antibiotic use
during travel was the strongest independent predictor for
ESBL-E acquisition (table 3). To assess the eff ects of
diff erent antibiotic classes in the model, we exchanged
the variable antibiotic use during travel (no vs yes) for a
variable indicating antibiotic class (no antibiotics vs
β-lactam, or quinolone, or other). Quinolone use was
most strongly associated with ESBL acquisition (adjusted
OR 6·0, 95% CI 2·9–12·4), whereas associations were
non-signifi cant for use of β-lactam (2·2, 0·95–5·14) or
other antibiotics (1·7, 0·59–2·35). We also detected
strong associations between ESBL-E acquisition and
diarrhoea during travel and, particularly, traveller’s
diarrhoea that persisted on return (table 3). Travellers
who had occasionally consumed food from street vendors
Number of
travellers
(n=1847)*
Number of
travellers who
acquired ESBL-E
(n=633)†
ESBL-E incidence
proportion (95% CI)‡
Number of
travel-days
Mean (SD)
duration of
travel (days)
ESBL-E incidence
per 100 person-
days of travel
(95% CI)§
Southern Asia 181 (9·8%) 136 (21·5%) 75·1 (68·4–80·9) 3727 20·6 (11·0) 7·2 (5·9–8·6)
Central and eastern Asia 84 (4·5%) 41 (6·5%) 48·8 (38·4–59·3) 1712 20·4 (10·8) 3·5 (2·5–4·7)
Western Asia 28 (1·5%) 12 (1·9%) 42·9 (26·5–60·9) 305 10·9 (7·5) 5·8 (3·0–9·9)
Northern Africa 81 (4·4%) 34 (5·4%) 42·0 (31·8–52·9) 981 12·1 (5·7) 4·5 (3·1–6·2)
Southeastern Asia 540 (29·2%) 200 (31.6%) 37·0 (33·1–41·2) 12 493 23·1 (11·6) 2·1 (1·8–2·4)
Caribbean and Central America 86 (4·7%) 24 (3·8%) 27·9 (19·5–38·2) 1653 19·2 (12·4) 1·7 (1·1–2·5)
Middle and eastern Africa 205 (11·1%) 57 (9·0%) 27·8 (22·1–34·3) 4060 19·8 (14·3) 1·6 (1·2–2·1)
Western Africa 106 (5·7%) 20 (3·2%) 18·9 (12·6–27·4) 1638 15·5 (11·1) 1·4 (0·8–2·0)
South America 180 (9·7%) 33 (5·2%) 18·3 (13·4–24·6) 4778 26·5 (14·7) 0·8 (0·5–1·1)
Southern Africa 116 (6.3%) 7 (1·1%) 6·0 (2·5–12·0) 2522 21·7 (8·6) 0·3 (0·1–0·6)
Northern America, Europe, and Oceania 17 (1·0%) 1 (<1·0%) 5·9 (1·1–27·0) 292 17·2 (11·3) 0·4 (0–1·6)
ESBL-E=extended-spectrum β-lactamase-producing Enterobacteriaceae. *Numbers do not add up to 1847 because 221 travellers visited more than one subregion (66 with
ESBL-E acquisition) and destination information was missing for two. †Numbers do not add up to 633 because 66 travellers visited multiple subregions and destination
information was missing for two. ‡Based on binomial distribution (Wilson’s score interval). §Calculated with the maximum likelihood estimation method based on a
constant acquisition rate with right-censored and interval-censored data.
Table 2: Incidence proportion and incidence per 100 person-days of travel for ESBL-E acquisition in Dutch travellers, by subregion
