Illness in French Travelers to Senegal: Prospective Cohort Follow-up and Sentinel Surveillance Data

Transcription

1 I S T M 296 Illness in French Travelers to Senegal: Prospective Cohort Follow-up and Sentinel Surveillance Data Aïssata Dia, MD, Philippe Gautret, MD, PhD, Eric Adheossi, MD, Albane Bienaimé, MD, Catherine Gaillard, MD, Fabrice Simon, MD, Philippe Parola, MD, PhD, and Philippe Brouqui, MD, PhD Service des Maladies Infectieuses et Tropicales, Hôpital Nord, AP-HM, Marseille, France; Service de Pathologies Infectieuses et Tropicales, Hôpital d Instruction des Armées Lavéran, Marseille, France; Faculté des Sciences de la Santé, Université de Niamey, Niger DOI: /j x Objective. To investigate travel-associated illnesses in French travelers to Senegal. Methods. A prospective cohort follow-up was conducted in 358 travelers recruited at a pre-travel visit in Marseille and compared to data from ill travelers collected from the GeoSentinel data platform in two clinics in Marseille. Results. In the cohort survey, 87% of travelers experienced health complaints during travel, which most frequently included arthropod bites (75%), diarrhea (46%), and sunburns (36%). Severe febrile illness cases, notably malaria and salmonella, were detected only through the surveillance system, not in the cohort follow-up. Food hygiene was inefficient in preventing diarrhea. Arthropod bites were more frequent in younger patients and in patients with pale phototypes. Sunburns were also more frequent in younger patients. Finally, we demonstrate that mild travel-related symptoms and the lack of arthropod bites are significantly associated with poor observance of antimalarial prophylaxis. Conclusions. In this study, we suggest the complementary nature of using cohort surveys and sentinel surveillance data. Effective protection of skin from arthropod bites and sun exposure should result in significantly reduced travel-associated diseases in Senegal. Travelers to Senegal should be informed that diarrhea is extremely common despite preventive measures, but it is mild and transitory and should not lead to the disruption of malaria chemoprophylaxis. In recent years, the World Tourism Organization has estimated the growth in travel to be approximately 6% per year, and similar trends are expected in the future. Growth has been driven by emerging destinations in Asia, the Pacific, Africa, and the Middle East, increasing the risk of travel-associated diseases. 1 Different approaches for risk estimation and/or risk characterization in travel medicine may be used, including the use of notification data, case series and chart reports, cohort surveys, airport surveys, and data collected by sentinel surveillance networks for travelers. 2 We propose here a combination of two methods to investigate travel-associated illnesses in Aïssata Dia and Philippe Gautret equally contributed to this work. Corresponding Author: Philippe Gautret, MD, PhD, Service de Maladies Infectieuses et Tropicales, Hôpital Nord, AP- HM, Chemin des Bourrelys, F Marseille CEDEX 20, France. philippe.gautret@club-internet.fr travelers. We conducted a prospective cohort follow-up in travelers recruited at a pre-travel visit in one travel clinic in Marseille and compared the results to data on ill travelers who presented in two sentinel surveillance clinics in Marseille. Travel characteristics, specific health behaviors, and compliance with preventive measures were also assessed as probable risk factors. Senegal was elected as the travel destination in this study because it is a very popular destination for tourists, with around 900,000 foreign visitors per year ( Senegal is the most popular destination in sub-saharan Africa for French travelers, 3 and little data about travel-associated diseases in French citizens returning from Senegal are available in the published literature. 4 9 Patients and Methods Prospective Cohort Survey All patients aged >18 years, who were seeking pretravel advice at the Marseille Travel Medicine Centre 2010 International Society of Travel Medicine, Journal of Travel Medicine 2010; Volume 17 (Issue 5):

2 Illness in Travelers to Senegal 297 (Tropical and Infectious Disease Ward, University Hospital, Hôpital Nord) before traveling to Senegal for less than 3 months, were prospectively screened for inclusion between January and December Overall, 6,000 travelers seek pre-travel advice at the Travel Medicine Center each year. A verbal questionnaire was administered on each individual by a physician addressing baseline demographics, socioeconomic status, and travel characteristics. Questionnaires were pilot tested among travelers at the Marseille Travel Medicine Centre. Because the evaluation of travelassociated sunburn occurrence was one of the objectives of the study, the phototype of individuals was assessed during the pre-travel encounter by observing skin appearance and assessing sunburn and tanning history according to the Fitzpatrick classification. 10 Briefly, phototype I burns easily and never tans; II burns easily and tans minimally with difficulty; III burns moderately and uniformly; IV burns minimally and tans moderately and easily; V rarely burns and tans profusely; and VI never burns but tans profusely. During the consultation, each individual was provided with extensive scripted advice about major travelassociated risks (arthropod bites, food and drinking water-related risk, sun exposure, environmental hazard, and animal-related injuries) and related preventive measures. Travelers were systematically prescribed malaria prophylaxis, self-treatment for diarrhea, and immunization against yellow fever, hepatitis A, and typhoid fever based on risk assessment by the physician. Travelers were subsequently contacted by telephone within a week of their return to minimize recall bias. Individuals were considered lost to follow-up after three unsuccessful calls at 1-week intervals. Data regarding risk behaviors, the occurrence of health problems during travel, and malaria chemoprophylaxis observance were recorded. Data regarding insect bite prophylaxis, sun exposure, food and drink consumption, freshwater bathing, sport activities, wet sand exposure, and animal contact were documented. The occurrence of health problems during travel was recorded. Systematically, investigation was conducted for the following: fever, cough, nose and throat diseases, diarrhea, vomiting, dehydration, heat stress, chronic disease decompensation, lower limb venous problems, trauma, psychological disorders, genitourinary symptoms, and skin diseases, including insect bites and sunburns. Data were analyzed with the SPSS v15.0 (SPSS, Inc., Chicago, IL, USA) software package. Chi-square tests were used to compare proportions of travelers who reported specific symptoms to those who did not. A p value <0.05 was considered significant. All p values were determined by two-tailed t-test. Factors associated with poor compliance to malarial prophylaxis were explored using logistic regression models. Factors with p values below 0.20 in univariate models were considered eligible for multivariate analysis, as suggested in the classical work of Mickey and Greenland. 11 Astepwise procedure based on likelihood ratio criteria was used to obtain the best criteria with the lowest Akaike criteria For the final model, a two-tailed p value <0.05 was considered significant. Sentinel Surveillance Data Data were prospectively collected from the GeoSentinel data platform, using standard GeoSentinel data fields, 15 for patients presenting to the two sites in Marseille (Infectious Diseases and Tropical Medicine wards, Hôpital Nord and Hôpital Lavéran) from March 2003 to December 2008 with a travel-associated illness following travel to Senegal. The GeoSentinel Surveillance Network consists of specialized travel/tropical medicine clinics on six continents where ill travelers are seen during or after traveling to a wide range of countries and where information on travelers is prospectively recorded using a standardized format ( Information collected included demographic data (age, sex, and country of birth), reason for most recent travel, duration of travel, pre-travel encounter, and time to presentation. Patients whose reason for traveling was their initial migration trip from Senegal to France were excluded from the study. Results Prospective Cohort Survey Among the 392 individuals enrolled during pre-travel consultation, nine canceled their journey (2.3%), 25 were lost in follow-up (6.4%), and 358 were administered a post-travel questionnaire. Comparisons between populations of respondents and nonrespondents to the post-travel questionnaire revealed no significant differences in terms of demographic, socioeconomic status, or travel characteristics (data not shown). Pre-Travel Data The ratio of males to females in the study was 0.9 with a median age of 43.3 years (range 19 79). Most travelers were French-born executives, professionals, and nonmanual employees. Tourism was the main reason for visiting Senegal and most individuals traveled in pairs. Within the cohort, 68.4% of individuals traveled during the dry season, which lasts from November to the end of May, and stayed in high-quality hotels in Petite Côte (69.8%) and Dakar (16.2%). The median travel duration was 8 days (range 3 92). The predominant phototype of the individuals was type III (Table 1). Immunization and antimalarial prescription details are indicated in Table 2. The median time between travel clinic visit and planned date of travel departure was 21 days (range days). Post-Travel Questionnaire Risk Behaviors. A large majority of travelers protected themselves against arthropod bites, mainly with insect repellent. Most of the travelers had at-risk attitudes

3 298 Dia et al. Table 1 Demographics, socioeconomic standings, travel characteristics, and phototypes of 358 travelers returning from Senegal (January December 2008) Number of travelers (%) Gender Male 172 (48.0) Female 186 (52.0) Age group (y) (missing data = 3) (20.6) (31.3) (36.3) (11.8) Socio-professional categories (missing data = 9) Unemployed 25 (7.2) Retired 30 (8.6) Students 22 (6.3) Farmers 1 (0.3) Executives and 130 (37.2) professionals Nonmanual employees 94 (26.9) Manual workers 20 (5.7) Artisans and traders 17 (4.9) Artists and related 10 (2.9) Country of birth France 318 (88.8) Western Europe 11 (3.1) North Africa 14 (3.9) Senegal 10 (2.8) Other countries (United 5 (1.4) States, Vietnam = 2, Guinea, and Mauritania) Reason for travel (missing data = 2) Tourism 273 (76.7) Visiting friends and 44 (12.4) relatives Business 24 (6.7) Missionary 12 (3.4) Study 3 (0.8) Group size (missing data = 12) Alone 37 (10.7) Pair 148 (42.8) Family 73 (21.1) Group 88 (25.4) Accommodation during travel Good quality hotel (tour 250 (69.8) operator or self-organized hotel) Basic (backpackers, local 108 (30.2) resident housing) Duration of travel 1 2 wk 280 (78.2) 2 4 wk 62 (17.3) >4 wk 16(4.5) Phototype (missing data = 9) I 20 (5.7) II 109 (31.2) III 171 (49.0) IV 29 (8.3) V 6 (1.7) VI 14 (4.0) Table 2 Reported behavioral risks in 358 travelers returning from Senegal (January December 2008) Behavior Number of travelers (%) Protection against 333 (93.0) arthropod bites Wearing long clothes 158 (44.1) Use of insect repellent 298 (83.2) Use of insecticide 159 (44.4) Use of air-conditioning 177 (49.4) Use of bed net 152 (42.5) No protection against 25 (7.0) arthropod bites Consumption of 305 (85.2) potentially contaminated food or drinking water Drinking tap water 48 (13.4) Eating uncooked 262 (73.2) vegetables/salads Eating ice cream 85 (23.7) Eating pastry 131 (36.6) Eating raw 36 (10.1) meat/fish/shellfish Consumption of safe food 53 (14.8) and drinking water only Sand beach activities 298 (83.2) Lying on sand without 69 (19.3) towel Walking barefoot 275 (76.8) Sport activities 124 (34.6) Wading or swimming in 29 (8.1) freshwater lakes and rivers Physical contact with 65 (18.2) animals Dogs 44 (12.3) Cats 36 (10.1) Monkeys 2 (0.6) Sun exposure 288 (80.4) No protection at all 5 (1.4) Use of sunscreen cream 262 (73.2) Wearing a hat 175 (48.9) Wearing long clothes 64 (17.9) Avoided 70 (19.6) Immunization Yellow fever 358 (100.0) Hepatitis A 178 (49.7) Typhoid 135 (37.5) Recommended antimalarial Atovaquone/Proguanil 272 (76.0) Doxycycline 60 (16.8) Mefloquine 22 (6.1) Patient refusing 4 (1.1) antimalarial Compliance with antimalarials Correct intake 257 (71.8) Mistakes and oversights 67 (18.7) No antimalarial at all 34 (9.5)

4 Illness in Travelers to Senegal 299 regarding food and drinking water consumption, barefoot walking, and sun exposure (Table 2). Common Health Hazards. A total of 313 (87.4%) travelers presented at least one health problem during their trip; eight (2.2%) consulted a doctor during travel, 25 (7.0%) consulted one after travel, and one individual was hospitalized for bleeding. A large proportion of travelers reported dermatological (74.9%) and (48.9%) diseases (Figure 1). Arthropod bites (62.3% of travelers) and sunburns (35.7%) accounted for the majority of skin problems, while diarrhea was the main complaint (45.5%). Among the travelers suffering symptoms, 37.1% thought it was due to Figure 1 Reported symptoms experienced in 358 individuals during travel to Senegal (January December 2008) and reported diagnoses of patients seen at GeoSentinel clinics in Marseille after returning from Senegal (March 2003 December 2008). (NB/Prospective cohort study. Other dermatological disorder: skin infection (8 cases), rash (5), photosensitivity (3), myiasis (2), animal-related injury (2), and pruritis (1); other disorder: vomiting (27), abdominal pain (6), constipation (1), and gastroesophageal reflux disease (1); respiratory: cough (26) and nose/throat disease (14); neuro/psychological disorder: anxiety reaction (4), headaches (4), vertigo (3), insomnia (2), depression (1), and mourningassociated stress (1); other: heat stress (9), dehydration (5), lower limb venous problem (6), trauma/injury (4), vaginitis (2), chronic illness de-compensation (2), urinary-tract infection (1), conjunctivitis (1) tachycardia (1), and articular pain (1). Sentinel surveillance data. Febrile systemic illness: falciparum malaria (10), malariae malaria (1), malaria species unknown (2), Q fever (1), salmonella infection (4), viral syndrome (2), dengue (1); relapsing fever (1), and unspecified (6); dermatological disorder: myiasis (5), cutaneous larva migrans (3), animalrelated injury (3), skin infection (2), leishmaniasis (1), filariasis (1), and rash (1); disorder: diarrhea (7), Entamoeba histolytica liver abscess (1), hookworm (1), acute hepatitis A (1), and acute hepatitis unspecified (1). antimalarial medication. The median time between the beginning of the trip and the first diarrheal symptoms was 5 days (range 0 86) and the mean duration of diarrheal episodes was 2 days (range 1 30). Most travelers suffering from diarrhea self-treated themselves (82.8%), two consulted a doctor during travel (0.6%), and 12 consulted one after travel (3.3%). Respiratory disease was also a significantly reported health hazard. Younger individuals, phototype I and II travelers, individuals traveling during the wet season, and those who used insect repellent and mosquito bed nets were significantly more likely to report arthropod bites. Individuals who exposed themselves to sun and younger travelers were significantly more likely to report sunburns (Table 3). Drinking tap water was associated with a higher frequency of diarrhea as was eating ice cream; however, these results were not statistically significant. Compliance and Side Effects With Antimalarial Medication. Most travelers (71.8%) were compliant with malaria prophylaxis recommendations (Table 2). The main reasons for not taking medications were as follows: 47.1% of individuals found it useless and 44.1% feared the side effects. Among the 324 travelers who declared having taken malarial prophylaxis, 65 (20.1%) reported side effects, eight of whom stopped medication. Individuals who reported at least one symptom (assigned or not to antimalarials) were more likely to be noncompliant regarding malaria prophylaxis compared to other travelers. Individuals using doxycycline compared to those using atovaquone/proguanil were also more likely to be noncompliant regarding malaria prophylaxis. In the multivariate model, reporting at least one symptom was found to be independently associated with a poorer compliance of antimalarial treatment, as well as not reporting arthropod bites (Table 3). Sentinel Surveillance Data From March 2003 to December 2008, 55 patients were included in the database (Table 4). The ratio of males to females in the study was 1.4 with a median age of 39 years (range 4 71). Most patients were born in France. Tourism was the main reason for travel (54.5%), followed by visiting friends and relatives (21.8%) and then business (16.4%). The median travel duration was 18 days (range 2 382). The median time between the end date of the trip and the clinic visit was 10 days (range 0 1,018). A proportion of 29.1% of patients had a pre-travel encounter with a health care provider and 34.5% were seen as inpatients after their return from Senegal. Compared to the travelers of the cohort study, those included in the Sentinel Surveillance database were more likely to be born in Senegal (p = 0.01), to be younger (p = 0.01), and more likely to travel

5 300 Dia et al. Table 3 Significant odds ratios of reported symptoms (arthropod bite or sunburn) and noncompliance with antimalarials in 358 travelers returning from Senegal (January December 2008) p-value OR 95% CI Reporting arthropod bites Aged y (vs >60 y) < Phototype I or II (vs phototype VI) Traveling during wet season (vs dry season) Use of repellent (vs no < use of repellent) Use of mosquito bed net (vs no use of mosquito bed net) Reporting sun burn Exposure to sun (vs no exposure to sun) Aged y (vs >60 y) Noncompliance with antimalarials Univariate model Reporting at least one complaint (vs, non-reporting complaints) Use of doxycycline (vs use of atovaquone/ proguanil) Multivariate model Reporting at least one complaint (vs non-reporting complaints) Non-reporting arthropod bites (vs reporting arthropod bites) Table 4 Basic characteristics of 55 ill travelers who were included in the GeoSentinel database upon returning from Senegal (March 2003 December 2008) Number of travelers (%) Gender Male 32 (58.2) Female 23 (41.8) Age group (y) <18 2 (3.6) (34.5) (18.2) (27.3) 60 9 (16.4) Country of birth France 42 (76.4) Western Europe 2 (3.6) North Africa 2 (3.6) Senegal 6 (10.9) Other countries 3 (5.5) Reason for travel Tourism 30 (54.5) Visiting friends and 12 (21.8) relatives Business 9 (16.4) Missionary 2 (3.6) Study 2 (3.6) Duration of travel 1 2 wk 21 (38.9) 2 4 wk 10 (18.5) >4 wk 23 (42.6) Pre-travel encounter Yes 16 (29.1) No 35 (63.6) Unknown 4 (7.3) Time to presentation to GeoSentinel clinic 1 wk 26 (47.3) 2 4wk 23(4.0) >4 wk 6 (10.9) Inpatients 19 (34.5) to visit friends and relatives (p = 0.05) or for business (p = 0.02). In addition, their travel duration was longer (p < 10 4 ). They were also more likely to be admitted to the hospital as inpatients upon return from Senegal (p < 10 4 ). Febrile systemic illnesses accounted for most of the cases (47.3%). Among etiologic diagnosis, malaria was the most frequent diagnosis followed by salmonella infections. Dermatological disease was the second most frequent cause of travel-associated disease (30.1%) and included mainly parasitic infections, such as myiasis, larva migrans, filariasis, and leishmaniasis. Among disorders (20.0%), diarrhea accounted for the most cases followed by hepatitis (Figure 1). During 2008, the Sentinel Surveillance system captured three cases of travel-related illnesses involving individuals from the cohort survey with diagnoses of diarrhea (Entamoeba histolytica), myiasis, and animal-related injury. Discussion Our survey gives a picture of common health hazards occurring during travel to Senegal as well as more severe diseases seen at specialized travel clinics and could serve as a basis for the adaptation of pre-travel advice. However, some limitations must be acknowledged. For instance, sample size is limited and conclusions cannot be generalized to all travelers to Senegal. Recruitment in the prospective cohort study of participants seeking pre-travel consultation, where the physician informs the participant of the risk factors, may prevent them from having at-risk behavior and therefore may introduce possible bias. In addition, as patients were sampled within 1 week after return, we were unable to identify diseases with a long incubation period and the antimalarial observance data analysis was restricted to during travel nonobservance. Finally, our study is based on self-reported data and therefore focused on syndrome rather than on specific etiological diagnoses. However,

6 Illness in Travelers to Senegal 301 to our knowledge, this is the only existing prospective study on travel-associated illnesses in travelers to Senegal. Data collected from the Sentinel Surveillance system suffer from selection and reporting biases in the types of patients who present at specialized sentinel clinics and the diagnoses that are made in these clinics. In addition, the collected data fields are relatively limited. Sentinel data do not concern all travelers, only patients who seek medical treatment. Therefore, it does not estimate incidence rates or provide a numerical risk for travelers to Senegal. However, combining the analysis of the two methods reduces the limits of each method. While all travelers were immunized against yellow fever, only half were immunized against hepatitis A and one third against typhoid fever. This results, in part, from the fact that French travelers tend to decline hepatitis A and typhoid fever vaccines for short-term travel. A high follow-up rate was obtained in our survey, with only 6.4% lost to follow-up. A proportion of 87.4% of travelers experienced some health complaints during travel, which is consistent with other recent studies However, the median travel duration was shorter in our survey. Arthropod bites, diarrhea, and sunburns were the most common complaints. A comparison of travelrelated diseases in other prospective cohort studies is problematic as none focused on travelers to Western Africa and included populations of travelers with distinct characteristics. Our cohort survey is mainly representative of short-term tourist travelers using travel-industry infrastructure in the context of pre-arranged or organized travel. Arthropod bite prevalence was shown to be age-dependent, which correlates with mosquito bite studies conducted under field conditions, 20 and skin phototype-dependent, which has not been previously described. The finding that intrinsic host factors may account for the variability in biting by arthropods is of special relevance for attempts to target subpopulations of travelers for persuasive pre-travel advice about arthropod bite preventive measures. The association between arthropod bite prevalence and use of repellent and bed nets in our survey reinforces this view. This apparently paradoxical result likely indicates that anti-arthropod measures were used mostly by individuals following arthropod bites, rather than as a systematic preventive measure. It may also be due to recall bias of bites in more careful travelers. We also observed that lack of arthropod bites was associated with a poorer compliance of antimalarials, as previously reported, indicating that more education is needed regarding the fact that an absence of observable arthropod bite does not mean an absence of malaria risk. 21 Poor adherence to recommendations regarding dietary restrictions was observed, which is consistent with most recent studies in Swiss and German travelers. 18,22 However, this is in contrast with another study conducted in Italians. 23 Diarrhea prevalence was high in our survey and not significantly influenced by food or water consumption patterns of travelers, as already observed in several recent works. 18,22 25 Increasing age was shown to be protective against diarrhea in several other studies, 22,26 which was not observed in our work. The inefficiency of food hygiene in preventing diarrhea stresses the need to clearly inform travelers to Senegal about the actual risk of diarrhea during their trip. Travelers should be prescribed medication for self-treatment of diarrhea. In addition, we demonstrate here for the first time that mild travel-related symptoms are associated with a poor compliance in the use of antimalarials, and therefore may account indirectly for a higher risk of severe infectious diseases. The association between disturbance and poor compliance to malaria chemoprophylaxis may be due to a general attitude toward poor compliance to preventive measures and the assumption by travelers that diarrhea was a side effect of the antimalarial. In such a case, it needs to be reinforced that mild, self-limiting diarrhea is not a reason to cease antimalarials. Finally, most travelers declared having experienced sun exposure during travel and having used sunscreen products. This is similar to a large study conducted in French expatriates and corroborates the sunscreen paradox hypothesis, which proposes that most people do not use sunscreen as protection but rather as a way to stay longer in the sun. 27 Sentinel Surveillance data identified Plasmodium falciparum as the most frequent cause of febrile illness in patients returning from Senegal, followed by salmonella infections, and myiasis as the most frequent cause of dermatological problems. Rare diagnoses were also reported, such as Q fever, dengue, relapsing fever, cutaneous larva migrans, cutaneous leishmaniasis and filariasis, hepatitis A, and hookworms. Both methods identified dermatologic and disease as frequent causes of illness in travelers to Senegal, but severe febrile illnesses and notably malaria were not captured by the cohort study. This is likely due to the differences in the demographics and travel characteristics of individuals studied by each method. The sentinel patients were more likely to be immigrants from Senegal visiting friends and relatives, business travelers, and more young travelers <30 years. The cohort survey was limited to short-term travelers (<3 months), whereas sentinel surveillance data were not. Indeed, the median travel duration was 8 days in the cohort study, compared to 18 days in the sentinel study, introducing a possible bias in comparing results. In addition, the time to presentation at sentinel clinics was longer than the interval time between the end of travel and the telephone call in the cohort survey. Finally, the low proportion of travelers who sought pre-travel advice may account for a higher proportion of severe diseases in the patients presenting to sentinel clinics. Diarrhea was underrepresented in sentinel surveillance data compared to the cohort survey data, which reflects the fact that the vast majority of patients suffering from diarrhea used self-treatment and did not consult a specialized sentinel clinic.

7 302 Dia et al. We evidence here the complementary nature of using a cohort survey and sentinel surveillance data. We demonstrate that information regarding the incidence of common but mild health problems is better collected through a prospective cohort survey although there are inherent biases even in this approach because only people presenting for pre-travel advice will be included. However, it is very difficult to study the incidence of severe but relatively uncommon travelrelated illnesses prospectively because a very large sample size is required to observe adequate numbers of infrequent health outcomes. This underpins the logic of employing different methodological approaches to answer questions about morbidity during travel. The observation period for the surveillance data was significantly longer than for the cohort survey. A cohort survey extending over many years would obviously not be feasible, which is another advantage of combining approaches. Such an innovative approach paints a clearer picture of the overall health risks for a specific destination and allows the design of evidence-based recommendations for travelers. In the case of Senegal, our results suggest that effective protection of skin from arthropod bites, animal-related injuries, sun exposure, and contact with wet soil or non-ironed clothes should result in a significant reduction of travel-associated diseases. Acknowledgments This document (C76F-C7D8-191C-6D83-35FF) was edited by American Journal Experts (support@journal experts.com). Declaration of Interests The authors state that they have no conflicts of interest. References 1. Chen LH, Wilson ME. The role of the traveler in emerging infections and magnitude of travel. Med Clin N Am 2008; 92: Leder K, Wilson ME, Freedman DO, Torresi J. A comparative analysis of methodological approaches used for estimating risks in travel medicine. J Travel Med 2008; 15: Armand L. French tourist abroad in 2006: results from the survey follow-up of tourists preference. Bull Epid Hebdo 2007; 25 26: Parola P, Niang M, Badiaga S, Brouqui P. Acute Q fever in patient from the tropics. Postgrad Med J 2000; 76: Parola P, Soula G, Gazin P, et al. Fever in travelers returning from the tropical areas: prospective observational study of 613 cases hospitalised in Marseille, France, Travel Med Infect Dis 2006; 4: Tarantola A, Nabeth P, Tattevin P, et al. Incident management Group. Lookback exercise with imported Crimean-Congo hemorrhagic fever, Senegal and France. Emerg Infect Dis 2006; 12: Legros F, Bouchaud O, Ancelle T, et al. Risk factor for imported fatal Plasmodium falciparum malaria, France Emerg Infect Dis. 2007, 13: Dupouy-Camet J, Lecam S, Talabani H, Ancelle T. Trichinellosis acquired in Senegal from warthog ham. Euro Surveill 2009; 14: Pistone T, Ezzedine K, Boisvert M, et al. Cluster of chikungunya virus infection in travelers returning from Senegal, J Travel Med 2009; 16: Astner S, Anderson RR. Skin phototypes, J Invest Dermatol 2004; 122:xxx xxxi. 11. Mickey RM, Greenland S. The impact of confounder selection criteria on effect estimation. Am J Epidemiol 1989; 129: Zucchini W. An introduction to model selection. J Math Psychol 2000; 44: Agresti A. Categorical data analysis. San Francisco: Wiley, Hosmer DW, Lemeshow S. Applied logistic regression. San Francisco: Wiley, Freedman DO, Weld LH, Kozarsky PE, et al. Spectrum of disease and relation to place of exposure among ill returned travelers. N Engl J Med 2006; 354: Hill DR. Health problems in a large cohort of Americans traveling to developing countries. J Travel Med. 2000; 7: Winer L, Alkan M. Incidence and precipitating factors of morbidity among Israeli travellers abroad. J Travel med 2002; 9: Rack J, Wichmann O, Kamara B, et al. Risk and spectrum of diseases in travellers to popular tourist destinations. J Travel Med 2005; 12: Redman CA, MacLennan A, Wilson E, Walker E. Diarrhea and respiratory symptoms over travelers to Asia, Africa and South and Central America from Scotland. J Travel Med 2006; 13: Michael E, Ramaiah KD, Hoti SL, et al. Quantifying mosquito biting pattern on humans by DNA fingerprinting of bloodmeals. Am J Trop Med Hyg 2001; 65: Ropers G, Du Ry van Beest Holle M, Wichmann O, et al. Determinants of malaria prophylaxis among German travelers to Kenya, Senegal, and Thailand. J Travel Med 2008; 15: Stephen R, Tornieporth N, Costa Clemens SA, et al. Epidemiology of traveler s diarrhea: details of a global survey. J Travel Med 2004; 11: Laverone F, Boccaline S, Bechini A, et al. Traveler s compliance to prophylactic measures and behaviour during stay abroad: results of a retrospective study of subjects returning to a travel medicine center in Italy. J Travel Med 2006; 13: Shlim DR. Looking for evidence that personal hygiene precautions prevent traveler s diarrhea. Clin Infect Dis 2005, 41(Suppl 8):S531 S Cabada MM, Maldonado F, Quispe W, et al. Risk factors associated with diarrhea among international visitors to Cuzco, Peru. Am J Trop Med Hyg 2006; 75: Hill DR. Occurrence and self treatment of diarrhea in a large cohort of Americans travelling to developing countries. Am J Trop Med Hyg 2000; 62: Ezzedine K, Guinot C, Mauger E, et al. Expatriates in high-uv index and tropical countries: sun exposure and protection behaviour in 9,416 French adults. J Travel Med 2007; 14:85 91.