Journal of Infection (2009) 59, 19e27

www.elsevierhealth.com/journals/jinf

A global study of pathogens and host risk factorsassociated with infectious gastrointestinal diseasein returned international travellers*

Ashwin Swaminathan a,n, Joseph Torresi b,c,*,o, Patricia Schlagenhauf d,p,Karin Thursky e,q, Annelies Wilder-Smith f,r, Bradley A. Connor g,r,Eli Schwartz h,i,r, Frank vonSonnenberg j,r, Jay Keystone k,l,r,Daniel P. O’Brien e,m,o, for the GeoSentinel Network

a National Centre for Epidemiology and Population Health, Australian National University, Canberra, Australiab Austin Hospital, Heidelberg, Victoria, Australiac The University of Melbourne, Victoria, Australiad WHO CC for Travellers’ Health Division of Epidemiology and Communicable Diseases University of Zurich Centre forTravel Medicine, Zurich, Switzerlande Victorian Infectious Diseases Service, Royal Melbourne Hospital, Victoria, Australiaf Department of Medicine, National University Singapore, Singaporeg Weill Medical College of Cornell University, New York, USAh The Center for Geographic Medicine and Tropical Diseases, Chaim Sheba Medical Center, Tel Hashomer, Israeli Sakler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israelj Department of Infectious Diseases and Tropical Medicine, University of Munich, Germanyk Center for Travel and Tropical Medicine, Toronto General Hospital, Toronto, Ontario, Canadal Department of Medicine, Faculty of Medicine, University of Toronto, Ontario, Canadam Geelong Hospital, Victoria, Australia

Accepted 13 May 2009Available online 31 May 2009

* All listed authors have given final approval for this manuscript to be submitted, as has the Research Committee of the GeoSentinelSurveillance Network.

* Corresponding author. Department of Infectious Diseases, Austin Hospital, The University of Melbourne, 145 Studley Road, Heidelberg3084, VIC, Australia. Tel.: þ61 3 9496 6676; fax: þ61 3 9496 6677.

E-mail address: josepht@unimelb.edu.au (J. Torresi).n Data collection, analysis and manuscript preparation.o Data collection, study design, interpretation of analyses, manuscript preparation.p Data collection, interpretation of analyses, preparation of maps, critical review of manuscript.q Statistical analyses and critical review of results.r Data collection, interpretation of analyses, critical review of manuscript.

0163-4453/$36 ª 2009 The British Infection Society. Published by Elsevier Ltd. All rights reserved.doi:10.1016/j.jinf.2009.05.008

20 A. Swaminathan et al.

KEYWORDSGastrointestinaldiseases;Infection;Travel;Diarrhoea

Summary Objectives: Infectious gastrointestinal disease (IGD) is a significant cause of mor-bidity in returned international travellers. This study aims to elucidate host and travel charac-teristics associated with IGD presentation, and describe the broad spectrum of aetiologicalpathogens responsible by geographic region of acquisition and reason for travel.Methods: We analyzed demographic, clinical and microbiological data recorded for ill,returned international travellers presenting to GeoSentinel Surveillance Network sites globallyduring the period September 1996eDecember 2005.Results: A total of 25,867 returned travellers were analyzed, of whom 7442 (29%) patients hada total of 8273 IGD diagnoses. Multivariate analysis demonstrated that IGD presentation wasassociated significantly with female sex (OR: 1.11; p Z 0.001); younger age group; attendinga pre-travel medical appointment (OR: 1.28; p< 0.0001); and travelling for the reason of tour-ism. Travelling for longer periods (>28 days) was associated with lower risk (OR: 0.93;p Z 0.04). Of the 2902 clinically significant pathogens isolated, 65% were parasitic, 31% bacte-rial and 3% viral. Presentation of IGD by specific pathogen varied markedly dependent on geo-graphic region of recent travel, and reason for travel.Conclusions: Host characteristics, region of travel and category of traveller, significantlyimpact on the relative likelihood of presenting with a broad range of pathogen-specific IGD.ª 2009 The British Infection Society. Published by Elsevier Ltd. All rights reserved.

Introduction

Previous studies of gastrointestinal illness in returnedtravellers have focused on travellers’ diarrhoea and studiedpopulations presenting to clinics in single countries,1e5 or inthose travelling to specific destinations.6e9 Although travel-lers’ diarrhoea is common, occurring in up to 50% of thosetravelling from a developed to a developing country,10 itis only one of a number of infectious gastrointestinal com-plaints that travellers may acquire. The complete spectrumof infectious gastrointestinal pathogens causing illness intravellers, apart from the well-described causes of travel-lers’ diarrhoea, has not been previously well characterizedand reported.

Our group recently reported the syndromic presentationof gastrointestinal disorders according to geographicaldestinations of travel11 and in a previous study, a restrictedanalysis of selected pathogen-specific causes of acute andchronic diarrhoea in travellers.12 In this study, we haveused the GeoSentinel Surveillance Network database to de-scribe the host and travel characteristics of returned inter-national travellers presenting to a GeoSentinel-associatedmedical centre with infectious gastrointestinal disease(IGD), and to provide a comprehensive analysis of the broadspectrum of microbiologically confirmed pathogens respon-sible for IGD acquired by travellers according to geograph-ical regions of travel. It is important to note that thisanalysis examines the characteristics of IGD amongst a pop-ulation of returned travellers seeking medical attention,and therefore differs from previous studies (which havelargely focused on travellers’ diarrhoea) that have lookedat overall rates of gastrointestinal illness acquired duringtravel.

Methods

The GeoSentinel Surveillance Network currently comprisesover 42 specialized travel or tropical medicine sites locatedon six continents. The spectrum of patients served by these

centers varies from those presenting to tertiary hospitals toself-referred out-patients.

Information recorded on the GeoSentinel database be-tween September 1996 and December 2005 was analyzed.Patient data were included if they had crossed an interna-tional border in the previous 10 years and had been assignedat least one final diagnosis which was travel related. Patientswere classified as having an ‘infectious gastrointestinaldisease’ if the diagnosis was primarily related to thegastrointestinal system and was deemed to be attributableto an infectious agent. Our analyses of the IGD data in theGeoSentinel database are based on the destination wherepeople have travelled to and not the country wheretravellers are seen and focused on microbiologically con-firmed rather than syndromic diagnoses.

Recorded data included demographic information (age,sex, country of birth and residence), travel history, clinicalsymptoms and reason for most recent travel (tourism,business, missionary work/volunteering, education/re-search, or visiting friends and/or relatives (VFR)). Wherea causative pathogen was isolated from clinical specimens,this information was recorded. Isolates not consideredclinically significant, such as Blastocystis hominis andnon-pathogenic entamoebae were excluded. Helicobacterpylori was also excluded as it was difficult to ascertainwhether infection was travel related. Additionally, infec-tion with the blood fluke Schistosoma spp. was not analyzedin this study, although its’ epidemiology in returned inter-national travellers presenting to GeoSentinel sites is de-scribed in a recent publication.13 The treating clinician’sfinal diagnosis was used to assign a diagnostic code froma standardized list of over 500 possible diagnoses. All Geo-Sentinel sites contributed microbiologically confirmed databased on the best reference diagnostic tests available inthat country at that time.

Region of travel was broadly classified into the followingdivisions: North America (United States and Canada),Central America, South America, the Caribbean, northernand western Europe, southern and central Europe, Oceania,Australasia (Australia and New Zealand (NZ)), South-east

Infectious gastrointestinal disease in returned international travellers 21

Asia, south Asia, east and north Asia, the Middle East, NorthAfrica and sub-Saharan Africa. The composite entity‘‘Western Industrialized World’’ was used as a baselinecomparator and combined North America, Australasia andnorthern and western Europe.

Statistical analysis

Where summary statistics are presented, they are for thetotal number of travellers with information available forthat variable. To determine the crude association betweenvarious baseline demographic and travel-related variablesand returned travellers presenting with an IGD, c2 andt tests were utilized. Variables achieving statistical signifi-cance on univariate testing (two-tailed p value� 0.05)were included in a multivariate logistic regression modelwith backwards step-wise elimination of non-statisticallysignificant variables. An a priori decision was made toinclude age, sex, region of travel and reason for travel inthe final regression model as these were considered impor-tant likely confounders based on previous studies. Statisti-cal analysis was performed using Stata v10.0 software(TX, USA).

To determine the ‘‘proportionate morbidity’’ of present-ing with a specific pathogen by region, only travellers whohad travelled to a single region were analyzed. This wasexpressed as a rate (per 1000 returned travellers froma single region). Proportionate morbidity avoided potentialmisclassification of the region of IGD acquisition wheremultiple regions had been visited. Proportionate morbidityby reason for travel was similarly expressed as a rate (per1000 returned travellers by reason for travel).

Limitations

As we have analyzed data that describes only thosereturned travellers who have attended a GeoSentinel sitefor the reasons of illness (or screening), the conclusionsoutlined below may not be applicable to all travellers. Ourstudy population represents a subset of the total populationthat has travelled abroad and hence the morbidity rates wehave calculated refer only to those who reported withillness to a GeoSentinel site. Our data does not representthe incidence of IGD amongst all returned travellers.Furthermore, as our data are collected by academicinstitution-affiliated GeoSentinel sites, it is likely that theillnesses recorded are more severe or chronic than thatroutinely encountered in the primary care setting.

Importantly, although all sites used best availablereference diagnostics, there was considerable heterogene-ity between centers in terms of availability and sophistica-tion of diagnostic (e.g. molecular) testing. This may havelimited the identification of fastidious, uncommon or viralpathogens.

Results

There were 25,867 patients entered into the GeoSentineldatabase during the study period. Only returning interna-tional travellers were included. Of these 7442 (28.8%)patients had a total of 8273 IGD diagnoses recorded. Five

percent of returned travellers presenting with an IGDrequired admission to hospital.

Travel patterns

Ninety-three percent (n Z 6889) of patients presenting withIGD visited one geographic region only. The single regions vis-ited during recent travel for those with an IGD were south Asia(27%), sub-Saharan Africa (26%), south-east Asia (20%), SouthAmerica (13%), Caribbean (7%), Middle East (7%), southern orcentral Europe (5%), north or east Asia (3%), north Africa (3%),North America (1%), Oceania (1%), northernor western Europe(1%) and Australia/NZ (1%).

The largest proportion of returned travellers presentingwith IGD were reported from GeoSentinel sites in Europe(47%) and North America (41%), followed by the Middle East(6%), and Australia/NZ (4%). This reflected the largernumber of GeoSentinel sites in the United States andEurope, compared with other regions.

The stated reason for recent travel in those presentingwith an IGD were tourism (64%), business (13%), VFR (11%),missionary/volunteer (8%) and for research/education pur-poses (4%).

Host and travel characteristics affecting thelikelihood of an IGD diagnosis

The mean age and median duration of travel was 35.1 yearsand 22 days, for returned travellers with IGD, comparedwith 36.8 years and 23 days, respectively, for travellerspresenting with another complaint.

Univariate analysis (Table 1) demonstrated numerous in-dividual and travel-related variables with a strong associa-tion with IGD presentation in those travelling to a singleregion. Women were more likely to present with an IGDthan men (OR 1.16; p< 0.001), as were younger travellers.Travellers whose trips were longer (>28 days) were lesslikely to present with an IGD than those travelling forshorter periods (<28 days) (OR: 0.92; p Z 0.005). Fifty-eight percent (n Z 4285) of those presenting with IGDreported a pre-travel encounter with a health-care pro-vider. Those who attended a pre-travel encounter had sig-nificantly higher odds of presenting with IGD than thosewho had not (OR 1.33; p< 0.001). Presentation with anIGD occurred most frequently amongst tourists relative toall other ‘‘reason for travel’’ groups.

Similarly, after adjustment for potential confounding ina multivariate regression model (Table 1), being of a youn-ger age group, female sex, attending a pre-travel encoun-ter and travelling for the reason of tourism, were allfound to be significantly associated with IGD presentation.

Significant pathogens isolated from returnedtravellers with IGD

From the 7442 returned travellers diagnosed with an IGD,2902 pathogenic isolates were recorded. Of these, 1899(65%) were parasitic, 905 (31%) bacterial, and 98 (3%) viral.The rates of the 25 most commonly isolated pathogens areshown in Table 2. Rates of IGD presentation in returnedtravellers by specific parasitic, bacterial and viral/

Table 1 Likelihood of returned unwell travellers presenting with infectious gastrointestinal infectiona.

Variable Univariate Odds ratio p Value Multivariate regressionOdds ratiob

p Value

Age (years) <30 1.00 1.0030e60 0.85 (0.79e0.90) <0.0001 0.87 (0.82e0.94) <0.0001>60 0.70 (0.62e0.79) <0.0001 0.72 (0.63e0.83) <0.0001

Travel duration <28 days 1.00 1.00>28 days 0.92 (0.86e0.97) 0.005 0.93 (0.87e1.00) 0.04

Sex Male 1.00 1.00Female 1.16 (1.09e1.22) <0.0001 1.11 (1.04e1.18) 0.001

Pre-travel encounter No 1.00 1.00Yes 1.33 (1.26e1.41) <0.0001 1.28 (1.19e1.37) <0.0001

Reason for travel Tourism 1.00 1.00Business 0.79 (0.72e0.86) <0.0001 0.95 (0.86e1.05) 0.29Research/education 0.73 (0.63e0.84) <0.0001 0.79 (0.66e0.93) 0.005Volunteer/missionary 0.71 (0.64e0.78) <0.0001 0.82 (0.73e0.93) 0.002VFR 0.60 (0.55e0.65) <0.0001 0.70 (0.62e0.78) <0.0001

a Only travellers to a single region were included in this analysis.b Adjusted for age, sex, pre-travel encounter, travel duration, reason for travel and region of travel.

Table 2 Rates (per 1000 returned unwell travellers) of themost frequently isolated pathogens.

Clinically significantIGD pathogen

n % Total Rate per 1000returnedtravellersa

Giardia 810 27.9 31.3Campylobacter 384 13.2 14.8Entamoeba histolytica 363 12.5 14.0Shigella 182 6.3 7.0Strongyloides 176 6.1 6.8Salmonella spp. other 134 4.6 5.2Dientamoeba fragilis 116 4.0 4.5Ascaris 110 3.8 4.3S. typhi 99 3.4 3.8Hookworm 71 2.4 2.7Tapeworm 71 2.4 2.7Hepatitis A virus 67 2.3 2.6Trichuris trichura 52 1.8 2.0S. paratyphi 47 1.6 1.8C. difficile 38 1.3 1.5Enterobius 36 1.2 1.4Cryptosporidium 32 1.1 1.2Cyclospora 31 1.1 1.2Hepatitis E virus 31 1.1 1.2Yersinia spp. (non-pestis) 20 0.7 0.8Clonorchis 19 0.7 0.7Fasciola 5 0.2 0.2Trichomonas intestinalis 4 0.1 0.2Isospora 3 0.1 0.1V. cholerae 1 0.0 0.0

Total 2902 100 112.2a From single or multiple regions.

22 A. Swaminathan et al.

protazoal pathogen are shown in Fig. 1aec, respectively.There were significant regional differences in the rates ofpresentation of pathogen-specific IGD. For example, therates of giardiasis and amoebiasis were highest in travellersreturning from South Asia, the Middle East and South Amer-ica; Campylobacter and Salmonella typhi/paratyphi fromSouth Asia and South-east Asia; and Hepatitis A from Oce-ania, the Middle East and North Africa.

Rates of IGD presentation caused by specific pathogensin travellers with different reasons for travel are shown inGraph 1aec. ‘‘Tourists’’ had the highest rates for giardiasisand Campylobacter. ‘‘Business’’ travellers had the highestrate of Shigella infection whereas ‘‘Missionary/volunteer-ing’’ travellers had the highest rates for amoebiasis. Travel-lers ‘‘visiting friends and/or family’’ (VFRs) demonstratea lower rate of IGD presentation for all pathogens exceptstrongyloides infection, S. typhi (but not paratyphi) andtapeworm infection.

Discussion

We have performed a comprehensive analysis of microbio-logically confirmed infectious gastrointestinal disorders,which to our knowledge, is the largest reported in a pop-ulation of returned travellers. This study extends ourgroup’s previous analyses of IGD in returned travellerswhich had primarily focused on the geographic risk of acuteand chronic diarrhoea with regard to a restricted number ofspecific pathogens,11 and IGD acquisition according to anessentially syndromic description of IGD by selectedregions.12 This study provides an abundance of new infor-mation regarding the characteristics of returned travellerspresenting with IGD that is not restricted to travellers’ diar-rhoea. It describes the relative rates of acquiring a broadspectrum of microbiologically confirmed pathogens byregions of travel and reason for travel.

TapewormHookwormDientamoeba

fragilis

AscarisStrongyloidesEntamoeba

Histolytica

GiardiaKey:

83

21

19

63

12122

1411 8

1

25

1483

35

17

8 1

22

9 7 92

10610

1

22

10103

27

158 10

2

29

18

8 62

1696 8

26

Other (<5.5)b

a

other (<5.5)Key: Clostridium

Difficile

Salmonella

(other)

Salmonella

typhi

ShigellaCampylobacter

27

15 16

10

3 29

610

2

82

106

2

17

10

15

62

7 62

63

16 139

2

10

1

<1

106

14

1

Salmonella

paratyphi

b

Figure 1 (a) Rates of parasitic pathogens per 1000 returned unwell travellers from a single regiona, (b) rates of bacterial path-ogens per 1000 returned unwell travellers from a single region, (c) rates of viral/protozoal pathogens per 1000 returned unwelltravellers from a single region. aWestern Industrialized World not shown. b‘‘Other’’ refers to the mean rate of other isolated path-ogen/s (where each individual pathogen’s rate is less than 5.5/1000 returned travellers) (note: also relevant for Fig. 1b,c).

Infectious gastrointestinal disease in returned international travellers 23

other (<0.5)Key: CyclosporaCryptosporidiumHepatitis A

42 1

13<1

2 2

1 <1

10

4

4 3<1

2 3 2

1 13

9

13

5

1

c

Figure 1 (Continued)

24 A. Swaminathan et al.

Our analysis revealed that of 25,867 travellers reportingfor medical assessment following international travel, overa quarter (29%) were diagnosed with an IGD. This iscomparable to findings of other studies, where IGDaccounted for 36% of infective cases in returned travellersadmitted to hospital in the United Kingdom,5 are frequentlyassociated with post-travellers diarrhoea irritable bowelsyndrome14,15 and accounted for 22% of those admittedwith a febrile illness in Australia.16

The overall rate of presentation with IGD in returnedtravellers in this analysis (29%) is lower than reportedrates of travellers’ diarrhoea in persons travelling froma developed to a developing region (up to 50%).10 This isbecause most cases of travellers’ diarrhoea occur duringtravel or are not severe enough to require attendance ata dedicated travel clinic on return. The range of IGD diag-noses covered by our analysis therefore are of a morechronic or severe nature and differ compared with studiesthat focus on travellers’ diarrhoea alone. Therefore,although our study does not quantify overall the risk ofacquiring an IGD during travel, it does provide strongevidence that in travellers presenting for medical assess-ment in clinics worldwide, IGD is a commonly diagnosedillness.

Although other studies did not find that gender in-fluences the risk of travellers’ diarrhoea,1,2,4,6,8 our studysuggests that female returned travellers are more likelyto present with an IGD. Similarly this is the case for age,where we have found younger travellers present with IGDat higher rates than older travellers. This may reflect thefact that younger travellers and/or females are more likely

to undertake travel behaviour that put themselves at risk ofacquiring gastrointestinal pathogens (e.g. females mayhave greater exposure to children), or that they seek med-ical help for gastrointestinal symptoms at a relatively lowerthreshold compared with other groups or for other medicalcomplaints.

In our population, short-term travellers (<28 days) weremore likely to present with an IGD than longer-termtravellers (>28 days). These findings are similar to thosefound in previous studies where reductions in incidencerates of travellers’ diarrhoea have been noted in long-termtravellers.7,17 This finding may reflect the fact that long-term travellers develop illness and recover while travelling,may be better prepared or have more time to acquire localknowledge about risk avoidance behaviour, or possibly havedeveloped immunity to local pathogens.

Interestingly, those who had a pre-travel medical con-sultation had a higher likelihood of presenting with an IGD(OR: 1.28; p< 0.0001) than those who did not. Likely, trav-ellers who seek pre-departure advice do so because theyperceive themselves to be travelling to high-risk areas forhealth that include a higher risk for acquiring IGD. Alterna-tively, and perhaps less likely, the results may reflecta lower threshold for seeking medical advice post-travelin those who had sought out medical advice before theydeparted.

Traveller category also influenced the likelihood ofpresenting with an IGD relative to other illnesses. Travelfor reasons of tourism posed the highest risk, while travelfor reasons of visiting friends and relatives (VFR) was leastlikely. Business travel has previously been associated with

0

2

4

6

8

10

12

14

16

18

20

Tourist

Business

Education

VolunteerVFR

Rat

e pe

r 10

00 R

etur

ned

Unw

ell T

rave

ller

0

5

10

15

20

25

30

35

40

Tourist

Business

Education

VolunteerVFR

Rat

e pe

r 10

00 R

etur

ned

Unw

ell T

rave

ller

0

0.5

1

1.5

2

2.5

3

3.5

4

Tourist

Business

Research/Educatoin

Missionary/Volunteer

VFR

Hepatitis ACryptosporidiumCyclospora

Rat

e pe

r 10

00 R

etur

ned

Unw

ell T

rave

ller

Campylobacter

Shigella

Salmonella (other)

Salmonella typhi

Salmonella paratyphi

Clostridium difficile

Giardia

Entamoeba Histolytica

Strongyloides

Ascaris

Dientamoeba fragilis

Hookworm

Tapeworm

Trichuris trichura

Enterobius

a b

c

Graph 1 (a) Rates of presentation with bacterial pathogens per reason for travel. (b) Rates of presentation with parasitic path-ogens per reason for travel. (c) Rates of presentation with viral/protozoal pathogens per reason for travel.

Infectious gastrointestinal disease in returned international travellers 25

a lower risk of travellers’ diarrhoea.2 Although VFR travelincreases the likelihood of presenting with other seriousinfections such as malaria,18,19 Hepatitis A20 and typhoidfever,18,19 this seems not to be the case for IGD. Certainly,travelling for the reason of VFR ostensibly representshigher risk travel because of lower attendance rates forpre-travel medical advice, lower vaccination rates, con-sumption of high-risk food and drink, and more prolongedand remote international stays. Thus the decreased likeli-hood of VFRs presenting with IGD in our study and others,6

may reflect: (i) protection acquired from immunity devel-oped as a result of immigration from or prolonged travelto, highly endemic areas for IGD, (ii) better risk avoidancestrategies, or (iii) differences in health-care seeking be-haviour for gastrointestinal symptoms compared withnon-VFRs.

The bacterial agents most frequently reported were Cam-pylobacter, Salmonella spp., Shigella and Clostridium diffi-cile. However, there were no diagnostic codes specificallyfor enterotoxigenic Escherichia coli, the most frequentcause of travellers’ diarrhoea,4,6,21 or other commonly de-scribed agents such as Enteroaggregative E. coli, Aeromonasand Plesiomonas. It is important to note that Vibrio choleraewas only reported in one case, highlighting the fact that thisis an uncommon cause of IGD in returned travellers.10,22 Inaddition, it has been stated that typhoid and paratyphoid fe-ver (S. typhi and S. paratyphi, respectively) are rare amongtravellers10; however, we found that these made up w5% ofpathogens isolated. This reinforces the importance of travelpractitioners recommending the use of typhoid vaccinationprior to travel to endemic regions, especially for VFRtravellers.20,23

26 A. Swaminathan et al.

In general, parasites have been found to cause 0e12% ofcases of travellers’ diarrhoea, with some regions such asEastern Europe and Russia having higher rates.21,22,24 In ourstudy of IGD they accounted for 65% of microbiologicallyconfirmed pathogens. Furthermore the rates of giardiasisand amoebiasis were highest in travellers returning fromSouth Asia, the Middle East and South America. This illus-trates a difference between IGD and travellers’ diarrhoea,since most of the pathogens described in our study (e.g.Strongyloides, Ascaris and other helminths) produce symp-toms other than diarrhoea. Giardia lamblia was the mostfrequent isolate identified in our study; it has previouslybeen described as the most common intestinal parasiteidentified in travellers.25 Coccidian parasites such as Cyclo-spora, Cryptosporidium, Microsporidia and Isospora areincreasingly recognised as causes of travellers’ diar-rhoea26,27; in our study they accounted for w2% of isolates.However, this may be an underestimate, as these diagnosesrequire specific staining techniques that are not routine inmany laboratories.

Only 98 (3.0%) of pathogens isolated with IGD were ofviral origin in this study. Previous studies have reportedviral aetiology rates of up to 12% for travellers’ diar-rhoea.6,21,28 However, most centers do not routinely per-form testing for viral pathogens known to causetravellers’ diarrhoea, such as Rotavirus and Norovirus vi-rus. In addition, these agents are more likely to causea short illness that resolves before the traveller receivesmedical attention. Of note, Hepatitis A (67 cases; 2.3% ofall isolates) continues to be reported in travellers despitethe availability of a safe and highly effective vaccine thatis recommended for all travel to endemic countries.

Our study clearly demonstrates geographic regionaldifferences in presentation with specific pathogens. Forexample, enteric fever secondary to S. typhi and S. paraty-phi infection, occurs in returned travellers from south Asiaat rates fourfold more than the next most likely region,a finding supported by numerous published studies.23,29,30

Rates of Hepatitis A proportionate morbidity amongst re-turned travellers were highest in those from Oceania, theMiddle East and North Africa in our study, that were consis-tent with those in previous reports.20,23

Among different categories of traveller, there were alsodifferences in presentation with specific pathogens. AmongVFR returned travellers presenting with IGD, rates ofchronic (non-travellers’ diarrhoea) pathogens such as tape-worm, hookworm and strongyloides were prominent. En-teric fever was found in a higher proportion in this group aswell, which reflects the greater relative proportion of VFRsreturning from south Asia e a finding supported by bothCanadian and US study.20,23

Conclusions

This study represents the largest analysis of infectiousgastrointestinal disease among international returned trav-ellers to date. We have shown that IGD is a major cause ofmorbidity in returned travellers (beyond travellers’ diar-rhoea), presentation rates are influenced by travellers’characteristics, and that a broad range of pathogens isimplicated. In addition, region of travel and category of

traveller impact significantly on the relative likelihood ofpresenting with a pathogen-specific IGD. Ongoing analysisof GeoSentinel data at regular intervals will allow for anassessment of the relative change in the epidemiology oftravel-related IGD.

Conflict of interest

All authors have no conflict of interest.

Financial support

GeoSentinel, the Global Surveillance Network of the In-ternational Society of Travel Medicine, is supported byCooperative Agreement U50/CI000359 from the Centers forDisease Control and Prevention.

Acknowledgments

We would like to thank all contributors to the GeoSentinelSurveillance Network. Thanks to Mr. H Jauss for assistancewith graphics.

Appendix. The GeoSentinel SurveillanceNetwork

In addition to the authors, members of the GeoSentinelSurveillance Network whocontributed data (in descendingorder) are Kevin C. Kain, University of Toronto, Toronto,Canada; Louis Loutan, University of Geneva, Geneva,Switzerland; DeVon C. Hale and Stefanie S. Gelman,University of Utah, Salt Lake City, UT, USA; GiampieroCarosi and Francesco Castelli, University of Brescia, Bres-cia, Italy; Graham Brown, Royal Melbourne Hospital, Mel-bourne, Australia; Prativa Pandey, CIWEC Clinic TravelMedicine Center, Kathmandu, Nepal; Phyllis E. Kozarskyand Carlos Franco-Paredes, Emory University, Atlanta, GA,USA; Robert Steffen, University of Zurich, Zurich, Switzer-land; Lin H. Chen and Mary E. Wilson, Mount AuburnHospital, Harvard University, Cambridge, MA, USA; MarcShaw, Worldwise Travellers Health and Vaccination Centre,Auckland, New Zealand; Alejandra Gurtman, Mount SinaiMedical Center, New York City, NY, USA (October 2002eAugust 2005 only); Hiroko Sagara, Yokohama MunicipalCitizen’s Hospital, Yokohama, Japan; Philippe Parola andJean Delmont, Hopital Nord, Marseille, France; Thomas B.Nutman and Amy D. Klion, National Institutes of Health,Bethesda, MD, USA; N. Jean Haulman and Elaine C. Jong,University of Washington, Seattle, WA, USA; David O.Freedman, University of Alabama at Birmingham, Birming-ham, AL, USA; Carmelo Licitra and Antonio Crespo, OrlandoRegional Health Center, Orlando, FL, USA; R. Bradley Sackand Robin McKenzie, Johns Hopkins University, Baltimore,MD, USA (December 1997eDecember 2005 only); MurrayWittner, Albert Einstein School of Medicine, Bronx, NY, USA;Robert Muller, Travel Clinic Services, Johannesburg, SouthAfrica (May 2004eJune 2005 only); Michael W. Lynch,Fresno International Travel Medical Center, Fresno, CA,USA; Poh Lian Lim, Tan Tock Seng Hospital, Singapore;Elizabeth D. Barnett, Boston University, Boston, MA, USA;

Infectious gastrointestinal disease in returned international travellers 27

Susan MacDonald, Beijing United Family Hospital andClinics, Beijing, Peoples Republic of China; Susan McLellan;Tulane University, New Orleans, LA, USA (December 1999eAugust 2005 only); William M. Stauffer and Patricia F.Walker, University of Minnesota, Minneapolis, MN, USA;Cecilia Perret and Francisca Valdivieso, Pontificia Universi-dad Catolica de Chile, Santiago, Chile; Nancy Piper Jenks,Hudson River Health Care, Peekskill, NY, USA.

References

1. Steffen R, van der Linde F, Gyr K, Schar M. Epidemiology ofdiarrhea in travelers. JAMA 1983 Mar 4;249(9):1176e80.

2. Kollaritsch H. Traveller’s diarrhea among Austrian tourists inwarm climate countries: I. Epidemiology. Eur J Epidemiol1989 Mar;5(1):74e81.

3. Cobelens FG, Leentvaar-Kuijpers A, Kleijnen J, Coutinho RA.Incidence and risk factors of diarrhoea in Dutch travellers: con-sequences for priorities in pre-travel health advice. Trop MedInt Health 1998 Nov;3(11):896e903.

4. Peltola H, Kyronseppa H, Holsa P. Trips to the South e a healthhazard. Morbidity of Finnish travellers. Scand J Infect Dis 1983;15(4):375e81.

5. McKendrick M. Infectious diseases and the returning traveller e

experience from a regional infectious diseases unit over 20years. J Appl Microbiol 2003;(94 Suppl.):25Se30S.

6. Steffen R, Collard F, Tornieporth N, Campbell-Forrester S,Ashley D, Thompson S, et al. Epidemiology, etiology, andimpact of traveler’s diarrhea in Jamaica. JAMA 1999 Mar 3;281(9):811e7.

7. Ericsson CD, DuPont HL, Mathewson IJ. Epidemiologic observa-tions on diarrhea developing in U.S. and Mexican students livingin Guadalajara, Mexico. J Travel Med 1995 Mar 1;2(1):6e10.

8. von Sonnenburg F, TornieporthN, Waiyaki P, Lowe B,Peruski Jr LF,DuPont HL, et al. Risk and aetiology of diarrhoea at various touristdestinations. Lancet 2000 Jul 8;356(9224):133e4.

9. Hill DR. Occurrence and self-treatment of diarrhea in a largecohort of Americans traveling to developing countries. Am JTrop Med Hyg 2000 May;62(5):585e9.

10. Ericsson CD. Travellers’ diarrhoea. Int J Antimicrob Agents2003 Feb;21(2):116e24.

11. Greenwood Z, Black J, Weld L, O’Brien D, Leder K, VonSonnenburgF, etal.Gastrointestinal infectionamong internationaltravelers globally. J Travel Med 2008 JuleAug;15(4):221e8.

12. Freedman DO, Weld LH, Kozarsky PE, Fisk T, Robins R, vonSonnenburg F, et al. Spectrum of disease and relation to placeof exposure among ill returned travelers. N Engl J Med 2006Jan 12;354(2):119e30.

13. Nicolls DJ, Weld LH, Schwartz E, Reed C, von Sonnenburg F,Freedman DO, et al. Characteristics of schistosomiasis in trav-elers reported to the GeoSentinel Surveillance Network 1997e

2008. Am J Trop Med Hyg 2008 Nov;79(5):729e34.

14. Ilnyckyj A, Balachandra B, Elliott L, Choudhri S, Duerksen DR.Post-traveler’s diarrhea irritable bowel syndrome: a prospec-tive study. Am J Gastroenterol 2003 Mar;98(3):596e9.

15. Okhuysen PC, Jiang ZD, Carlin L, Forbes C, DuPont HL. Post-di-arrhea chronic intestinal symptoms and irritable bowel syn-drome in North American travelers to Mexico. Am JGastroenterol 2004 Sep;99(9):1774e8.

16. O’Brien D, Tobin S, Brown GV, Torresi J. Fever in returned trav-elers: review of hospital admissions for a 3-year period. ClinInfect Dis 2001 Sep 1;33(5):603e9.

17. Tjoa WS, DuPont HL, Sullivan P, Pickering LK, Holguin AH,Olarte J, et al. Location of food consumption and travelers’diarrhea. Am J Epidemiol 1977 Jul;106(1):61e6.

18. Leder K, Tong S, Weld L, Kain KC, Wilder-Smith A, vonSonnenburg F, et al. Illness in travelers visiting friends and rel-atives: a review of the GeoSentinel Surveillance Network. ClinInfect Dis 2006 Nov 1;43(9):1185e93.

19. Fenner L, Weber R, Steffen R, Schlagenhauf P. Imported infec-tious disease and purpose of travel, Switzerland. Emerg InfectDis 2007 Feb;13(2):217e22.

20. Behrens RH, Collins M, Botto B, Heptonstall J. Risk for Britishtravellers of acquiring hepatitis A. BMJ 1995 Jul 15;311(6998):193.

21. Jiang ZD, Lowe B, Verenkar MP, Ashley D, Steffen R,Tornieporth N, et al. Prevalence of enteric pathogens amonginternational travelers with diarrhea acquired in Kenya (Mom-basa), India (Goa), or Jamaica (Montego Bay). J Infect Dis 2002Feb 15;185(4):497e502.

22. Peltola HL. Textbook of travel medicine and health. In:DuPont HL, editor. Hamilton: Decker; 1997. p. 78e86.

23. Provost S, Gagnon S, Lonergan G, Bui YG, Labbe AC. HepatitisA, typhoid and malaria among travelers e surveillance datafrom Quebec (Canada). J Travel Med 2006 JuleAug;13(4):219e26.

24. Ericsson CD. Travelers’ diarrhea. Epidemiology, prevention,and self-treatment. Infect Dis Clin North Am 1998 Jun;12(2):285e303.

25. Ryan ET, Wilson ME, Kain KC. Illness after international travel.N Engl J Med 2002 Aug 15;347(7):505e16.

26. Okhuysen PC. Traveler’s diarrhea due to intestinal protozoa.Clin Infect Dis 2001 Jul 1;33(1):110e4.

27. Muller A, Bialek R, Kamper A, Fatkenheuer G, Salzberger B,Franzen C. Detection of Microsporidia in travelers with diar-rhea. J Clin Microbiol 2001 Apr;39(4):1630e2.

28. Johnson PC, Hoy J, Mathewson JJ, Ericsson CD, DuPont HL.Occurrence of Norwalk virus infections among adults in Mexico.J Infect Dis 1990 Aug;162(2):389e93.

29. Mermin JH, Townes JM, Gerber M, Dolan N, Mintz ED,Tauxe RV. Typhoid fever in the United States, 1985e1994:changing risks of international travel and increasing antimi-crobial resistance. Arch Intern Med 1998 Mar 23;158(6):633e8.

30. Connor BA, Schwartz E. Typhoid and paratyphoid fever in trav-ellers. Lancet Infect Dis 2005 Oct;5(10):623e8.