Background

Almost one billion individuals traveled internationally in 2011 [1]. Travelers’ diarrhea (TD) is the most common malady afflicting travelers, and several observational studies report an incidence of 50% after a 2-week travel period [2, 3]. High risk areas for experiencing travelers diarrhea according to the CDC include most of the developing areas of the world, particularly within the tropical and subtropical areas including SubSaharan Africa (excluding South Africa), SouthEast Asia, SouthCentral Asia, the Middle East, Latin America (South and Central America excluding Chile and Argentina) and Oceana [2, 4, 5]. Intermediate risk regions include the Caribbean nations, South Africa, Argentina, Chile, Eastern Europe, Russia, China and Portugal [2, 4, 5]. Bacteria account for up to 90% of identified infectious etiologies for acute TD, predominately enterotoxigenic Escherichia coli (ETEC), and enteroaggregative E. coli (EAEC), although there is regional variability [5, 6]. Of public health importance, travelers are recognized as an important vector for transmission of emerging and multi drug resistant (MDR) enteropathogens globally, mandating global public health surveillance [2].

An estimated 3–10% of travelers experience persistent diarrhea (diarrhea exceeding 2-weeks) [7] while upwards 4% of returning travelers experience chronic diarrhea (diarrheal exceeding 4-weeks) [5–9]. These estimates vary widely dependent on geographical location, travel duration, itinerary, population, and preceding utilization of pre-travel clinic education and counseling. As our research failed to identify the incidence of specific enteropathogenic etiologies of persistent or chronic diarrhea in returning travelers, and noted that most syndromic diagnoses were biased towards the definition of chronic diarrhea, we will consolidate both definitions as persistent/chronic diarrhea.

We propose to catalogue four categories of persistent/chronic diarrhea in the returning traveler (as referenced in [5, 7, 9, 10] as follows: 1) infectious [presumably biased towards parasitic [5] (although a host of bacterial enteropathogens contribute to the burden of chronic infectious diarrhea)] [7, 10–12]; 2) presumed infectious (tropical sprue, Brainerd’s diarrhea); 3) post-infectious sequelae [post infectious irritable bowel syndrome (PI-IBS), lactose intolerance, small intestinal bowel obstruction (SIBO)]; and 4) unmasked (presumably via an infectious disease exposure/trigger) [bile acid malabsorption (BAM), inflammatory bowel disease (IBD), celiac disease, microscopic colitis]—see Table 1. The terminology “unmasked” suggests that an infectious exposure may be a trigger in a cascasde of events which induces disease in a susceptible population.

Table 1 Categorization and brief exposition of known etiologies for persistent/chronic diarrhea in returning travelers
Full size table

Our current understanding of the etiology of persistent/chronic infectious-mediated diarrhea in the returning traveler is limited to case studies, case series, and cross-sectional studies. Given the large number of international travelers and the ever increasing geographical destinations, persistent/chronic diarrhea is likely to increase as a public health threat. Epidemiological data on the infectious etiology of persistent/chronic diarrhea are needed to develop evidence-based guidelines for disease management. Therefore, we conducted a systematic review of the published literature to summarize the current data on the incidence, etiology and regional variability of persistent/chronic diarrhea among returning travelers.

Methods

Search strategy

We conducted a search of electronic databases (Medline, Embase, and the Cochrane database of clinical trials) from 1990 to 2015, with the following terms: “chronic or persistent diarrh* and (returning) travel* [allowing for travel, traveler, and variable spelling (diarrhea or diarrhoea) and (traveler or traveller)]; GeoSentinel Surveillance and diarrh*; Geosentinel (based on review of references from the aforementioned search keywords Geosentinel Surveillance and diarrh*), chronic or persistent diarrh* and enteropathogen, and travel-associated infection (predicated upon idetnifying potential articles within the references of all aforementioned search keywords).

Inclusion/Exclusion Criteria

All articles were reviewed for eligibility criteria. To be included, studies were required to 1) report on adult (≥18 years) travelers presenting for travel-related illness at a health-care facility (excluding survey based data), 2) be published in the English language from 1990 to 2015, 3) report denominator data (extractable incidence rates) of persistent and/or chronic diarrhea among returning travel populations (experiencing travel duration for up to 3 months).

Data abstraction

The following data were abstracted and entered into a MS Excel® worksheet for analysis: author, publication year, study years, travel origination, population demographics, travel destination, number of travelers, PM, diarrheal etiology, and burden of persistent/chronic diarrheal disease relative to all cause travel related morbidity.

Analysis

Incidence rates and standard 95% confidence intervals (for all travelers (global) and region specific [Latin America, Africa and Asisa]) were estimated using a random-effects model (DerSimonian & Laird) [13]. Heterogeneity in study incidence rates was assessed using a χ 2 statistic, and graphically represented with Forest plots. Statistical analyses were performed with Stata Version 10 (StataCorp., College Station, TX).

Results

Our initial search resulted in 541 articles from which we identified 19 studies meeting the inclusion criteria. Of those, 18 reported on the incidence of persistent/chronic diarrhea as a syndromic diagnosis in returning travelers while one study reported adequate denominator data to enable estimates of pathogen-specific etiology (Table 2). The majority (287; 53%) of articles were excluded due to reporting on non-diarrheal travel related illness. Additional exclusionary criteria included duplicate articles (93; 17.2%), reviews (82; 15.2%), publication in foreign language (54; 9.9%), failure to provide denominator data (37; 6.8%), case studies (34; 6.3%), non-travel study populations (26; 4.8%), and pediatric populations (14; 2.6%) (Fig. 1). Odolini et al. reported data from two separate years of surveillance and data for each year were entered as separate observations [14]. Similarly, Gautret et al. [15] reported on separate cohorts stratified by age (young and elder cohorts). These data were extracted and entered as separate observations yielding a total of 20 observations for syndromic diagnoses [15]. The predominant reasons for failing eligibility in identifying enteropathogenic etiologies for persistent/chronic diarrhea in returning travelers were a lack of pathogen specific incidence reporting, and failing to partition acute and chronic diarrheal presentations.

Table 2 Studies identifying the syndromic diagnosis of chronic diarrhea with ranking relative to all assessed syndromic diagnosesa
Full size table
Fig. 1
figure 1

Flow Chart Depicting Search Methodology

Full size image

We identified a single study reporting upon the pathogen-specific etiology for persistent/chronic diarrhea in the returning traveler, delineating transparent denominator data enabling estimates of incidence rates [16]. A total of 116 consecutive patients experiencing persistent/chronic diarrhea post-travel were enrolled at two clinics between 1995 and 1996 (Netherlands and Belgium). Giardia was the most common infection in 16.4% of specimens followed by Campylobacter (6.1%); Shigella (3.5%); Cyclospora (3.5%); Salmonella spp. (0.9%); and Entamoeba histolytica (0.9%). Paschke et al. [17] conducted a similar study, but did not stratify the enterpoathogens across the acute and persistent/chronic presentations.

The incidence of persistent/chronic diarrhea for travelers from the included studies is shown in Fig. 2. The total incidence for persistent/chronic diarrhea ranged from 0.05 to 0.11. Of note, the highest incidence was identified in Freedman [18] reporting rates from all three regions (0.11 [0.11–0.12]), Hagman [19] reporting global rates (0.11 [0.10–0.11]), and Flores-Figuera [4] reporting rates from Central America (0.11 [0.11–0.12]) [4, 18]. The continental specific [Latin America, Africa, and Asia] number of travelers and incidence (95% CI) was [15816 (0.09 [0.07–0.11]), 42290 (0.06 [0.05–0.07]), and 27433 (0.07 [0.06–0.09]) respectively] depicted in Fig. 3a-c. From the above data we observe significant heterogeneity in incident rates across regions. There was a significant difference (p = 0.05) in the incidence of persistent/chronic diarrhea between Africa and Latin America. Persistent/chronic diarrhea ranked fourth as a syndromic diagnosis across all regions. Additionally, within the continental specific incident rates, we observe a trend toward decreasing diarrheal rates in contemporary reporting periods, more pronounced for Latin America and Asia.

Fig. 2
figure 2

Forest Plot depicting point estimates (95% CI) for the incidence (PM) persistent/chronic diarrhea in Global Returning Travelers

Full size image
Fig. 3
figure 3

a-c Forest Plot depicting point estimates (95% CI) for the incidence (PM) persistent/chronic diarrhea in Returning Travelers Stratified by Continent (Latin America, Africa, and Asia)

Full size image

Figure 4a-c delineates the persistent/chronic diarrhea incidence by continent and region. For Latin America, we observe a trend towards decreased incidence rates in South America (signifiant relative to the Carribean islands). In Asia, we observe a trend towards decreasing incidence rates observed between SE Asia and SC Asia. Finally, in Africa, we observe a significant difference between incidence rates observed between North Africa and Subsaharan Africa (SSA). Although study numbers are small, we do see significantly lower diarrheal rates between subsaharan Africa and [North Africa, South Central Asia, and Central America].

Fig. 4
figure 4

a-c Forest Plots depicting point estimates (95% CI) for the incidence (PM) persistent/chronic diarrhea in Returning Travelers Stratified by Continent (Latin America, Africa, and Asia) and Region

Full size image

Discussion

We estimated 60 cases of persistent/chronic diarrhea per 1000 travelers in over >300,000 global travelers, comparable to a prior report [7]. Although interpretations are limited by the dearth and heterogeneity of studies and variability in outcomes reported, we identified lower published rates of chronic diarrhea from Sub-Saharan Africa relative to North Africa, South Central Asia, and Central America. Of note, persistent/chronic diarrhea was consistently one of the leading syndromic diagnoses across all regions in returning travelers confirming its prominence as a significant public health issue.

The considerable heterogeneity observed in the incidence rates (reflected in the elevated I2 values depicted in all forest plots) best scrutinized within the continental and regional forest plots may be attributed to 1). the historical cohort effect (generally rates have decreased in recent years); 2). varying study designs (passive vice active surveillance); 3) variable enrollment (population sizes and demographics (age, gender, travel origin); 4) heterogeneous travel durations and itineraries; and 5) no control over antecedent travel education and counseling. This heterogeneity contributes to the wide confidence intervals identified upon pooled estimates.

The majority of data identified in our search reporting upon the etiology of travel-related chronic diarrheal infections stem from the GeoSentinel Global Surveillance Network. This surveillance network comprises 57 specialized international travel and tropical medicine clinics in >25 nations on 6 continents contributing sentinel longitudinal surveillance data on all ill travellers, representing the largest repository of travel-related data [20]. This repository provides epidemiologic information on infectious disease burden (and its gradient) in travelers including chronic diarrhea. This network enables communication of novel or emerging disease and outbreaks including diarrheal enteropathogens. As these clinics are referral centers, exploiting passive case ascertainment, accurate epidemiological descriptions (disease and pathogen incidence) may be biased [20] as diagnoses are limited to more chronic, severe or complex diseases leading to underreporting and underrepresentation of the full spectrum and burden of illness [21]. Despite these limitations, the travelers included in this analysis comprise a sentinel cohort facilitating insight into the complex epidemiology of travel-associated chronic diarrhea.

The eligible study reporting enteropathogenic etiologies for persistent/chronic diarrhea in returning travelers [16], coupled to the ineligible studies surveyed, and the unpublished data from the Geosentinel Surveillance Network suggest Giardiasis (and other enteropathogenic parasites) comprises an appreciable percentage of infectious mediated etiology [20].

Our study highlights the relative dearth of published data characterizing chronic diarrheal incidence, and enteropathogenic etiologies in infectious-mediated chronic diarrhea in travelers. Although we identified several studies which identified specific enteropathogens associated with chronic diarrhea in returning travelers [11, 17, 22–24], these studies did not sufficiently report denominator, or incidence data. Many studies reported enteropathogenic etiologies for diarrhea across the spectrum of diarrheal acuity without stratifying into chronic (vice acute) categories limiting data interpretation [25, 26].

Despite the lack of etiologic data, a host of case studies affirm a breadth of enteropathogens should be considered [9]. For example, Swaminathan et al. reported travel-associated enteropathogenic etiologies for gastrointestinal disease in a survey of over 25,000 international travelers from 1996 to 2005 exhibiting acute and persistent/chronic diarrhea [26]. Notably, they identified 29% of travelers presenting with infectious gastrointestinal disease (encompassing acute and chronic durations) of which 65% were attributed to parasites, 31% bacterial and 3% viral with significant geographical variation. Giardia was the most common pathogen identified (27.9%) followed by Campylobacter (13.2%), E. histolytica (12.5%), Shigella (6.3%), and Strongyloides (6.1%) [26]. Soonawala et al. reported on asymptomatic post-travel parasitic carriage in 556 Dutch travelers (median travel 12 weeks, minimum 2 weeks) from 2007 to 2009 to the subtropics. Many of these travelers did report episdoes of acute diarrhea during travel that had resolved by the post-travel evaluation. Giardia (4%), Cryptosporidium spp (1%), Schistosoma spp (6%) (only from travelers to Africa), and Strongyloides stercoralis (0.2%) were identified post-travel [25]. Another investigation exploited multiplex PCR for four parasites in fecal specimens acquired from over 2500 Belgium travelers post-travel (regardless of symptoms). They noted the following pathogen distribution: Giardia lamblia (6%), Cryptosporidium spp (1.3%), S. stercoralis (0.8%), E. histolytica (0.5%)] [27]. Although these data can not be directly extrapolated to incidence, they do support these parasites as common etiologic agents of aute and chronic TD.

Despite the wealth of data from the GeoSentinel Surveillance Network databases, data on enteropathogen-specific etiology for persistent/chronic diarrhea are lacking. Furthermore, laboratory support was not structured and often limited [18]. Freedman et al. stated that parasitic etiologies accounted for the majority of enteropathogens identified in chronic infectious diarrheal cases presenting to GeoSentinel based clinics from all regions except SE Asia in which bacterial etiologies predominated. We note that many of the eligible studies reported syndromic diagnoses, yet consistently isolated the specific diagnosis of “Giardiasis” as an appreciable etiology of all cause travel-related morbidity. For example, Chen [28] reported Giardiasis (PM: 36) as the second most common diagnosis in long-term travelers. As travelers evaluated within the GeoSentinel surveillance networks are biased toward persistent/chronic cases, and coupled to the results depicted by Schultsz et al. [16] above, we may infer that Giardiais likely a common etiologic agent of infectious persistent/chronic diarrhea [20, 28].

Although not a standardized syndromic diagnosis, many of the studies reported PI-IBS as a major etiology of travel-related morbidity. Recent data culled from the GeoSentinel network (1997–2011) quote a PI-IBS incidence of 2–5% [19, 29] while others have estimated a rate of 5–10% [30, 31]. There was generally a lack of clarity as to whether the PI-IBS diagnosis was rendered in travelers presenting with chronic diarrhea. However, as PI-IBS wasn’t included in the diarrheal estimates, and as PI-IBS cases are biased towards diarrheal presentations [7], we may infer that the incidence of chronic diarrhea in returning travelers is higher than tabulated. Our review provides support for persistent/chronic diarrhea as an important medical issue for the returning traveler and a significant public health issue. We limited our search to publications in the English language. Although we may not have captured literature published in non-English, given that the bulk of the contemporary data derives from the GeoSentinel Global Surveillance Nework whose results are published in the English language, we feel confident we captured the majority of data published.

Despite the limitations cited above, it appears that based on the unpublished and published data supplied by the Geosentinel network, coupled to the eligible and ineligbile studies reported above, that Giardiasis and parasites in general comprise an appreciable percentage of infectious mediated persistent/chronic diarrhea in returning travelers [16, 18, 25–27].

To improve our insight into the epidemiologic data on the etiologic agents of travel-associated persistent/chronic diarrhea, systematic investigations utilizing standardized exposure histories, laboratory evaluation and complementary endoscopic evaluation are needed. The use of molecular methods, including multiplex PCR assays on stool specimens, may increase pathogen identification [25, 27]. This would also improve characterization of the non-infectious causes of travel-associated chronic diarrhea, while potentially elucidating the triggers (infectious and non-infectious) and cascades of events precipitating disease unmasking. This information is paramount to developing optimal diagnostic, preventive, and treatment algorithms. The travel clinic is well positioned to conduct these studies contingent upon pursuing active surveillance, and implementing harmonized evaluations across participating clinics.

Conclusions

Persistent/Chronic diarrhea is a leading syndromic diagnosis globally and across all regions for travel-associated morbidity. The 6% incidence (PM of 60) of persistent/chronic diarrhea observed in over >300,000 global travelers is comparable to prior estimates. We identified lower rates of chronic diarrhea from Sub-Saharan Africa relative to North Africa, South Central Asia, and Central America. Parasites, most notably Giardia lamblia, comprise an appreciable percentage of the enteropathogenic etiology of infectious mediated persistent/chronic diarrhea. Our study highlights the relative dearth of published data characterizing chronic diarrheal incidence and pathogen etiology. Ideally, active surveillance investigations desigend to capture incidence data on persistent/chronic diarrhea exploiting the exisitng Travel clinic networks, marshalling standardized exposure histories, and exhaustive and advanced diagnostic methods with delineation of diarrheal duration in returning travelers would fill a significant gap in our understanding of this important public health issue.