Low Risk of Japanese Encephalitis in Short-Term Australian Travelers to Asia


Corresponding Author: Associate Professor Joseph Torresi, FRACP, Department of Infectious Diseases, Austin Hospital, 145 Studley Road, Heidelberg, Victoria 3084, Australia. E-mail: josepht@unimelb.edu.au


The risk of Japanese encephalitis (JE) in travelers is unknown. In this prospective study, we investigated the incidence of JE in 387 short-term Australian travelers visiting Asia over a 32-month period from August 2007 to February 2010 by performing pre- and post-travel antibody testing. No travelers were infected with JE virus during travel, indicating a low risk of infection for short-term travelers.

Japanese encephalitis (JE) is widespread in many countries within Asia and remains the leading cause of encephalitis in children from JE endemic countries.[1] However, the risk of infection for a nonimmune traveler who visits JE endemic destinations is unknown. A recent study reviewing published cases of JE in travelers reported an incidence estimate of 0.2 cases per million travelers.[2] A second study of JE in Swiss and British travelers reported an incidence of 1.3 cases per 7.1 million travelers.[3] For the general traveler who may only spend short periods of time in areas that put them at risk of acquiring JE, the need for vaccination remains questionable, and there are no published prospective studies of JE incidence in short-term travelers. In this report, we investigated the incidence of JE in short-term travelers to Southeast Asia by measuring seroconversion rates to JE virus.


We performed a multicenter prospective cohort study of Australian travelers over a 32-month period from August 2007 to February 2010. Travelers were consecutively enrolled if they were at least 16 years of age, intending to travel to Asia for a minimum duration of 7 days, and returning to Australia within the study period. Validated questionnaires were provided to travelers at recruitment before travel (pre-travel questionnaire) and after travel (post-travel questionnaire).[4] The questionnaires recorded data on gender, age, ethnicity, travel destinations, travel duration, health during travel, mosquito prevention strategies, receipt of JE vaccination, and prior history of flavivirus infection.[4] Baseline blood samples were taken at recruitment to assess for pre-existing exposure to JE virus. Travelers were followed up within 10 days of return from travel and a second blood sample was taken to assess for JE seroconversion.

Serological testing was performed at the Victorian Infectious Disease Reference Laboratory (VIDRL; North Melbourne, Victoria, Australia) using a JE-specific immunofluorescence assay that detected immunoglobulin G (IgG) antibodies to JE to assess JE seroconversion. Post-travel sera with JE antibody titer ≥80 were reported as positive and JE antibody titers >10 but <80 were reported as “low positives.”

Data were analyzed with Minitab statistical software, version 16. The incidence density of JE infection was calculated as number of infections per 10,000 traveler-days and exact Poisson 95% CIs were calculated around this estimate. There is no universal agreement on the best method for calculating CIs around zero incidence, so the upper limit should be taken as approximate only.[5]


In the study period, 681 eligible travelers were invited to participate and 467 travelers agreed to participate. Of the 467 travelers, 58 (12.4%) were lost to follow-up, 4 had missing sera and 18 were later excluded as they no longer fulfilled the inclusion criteria (travel to non-Asian destinations and/or did not return during the study period), leaving 387 travelers. The demographic characteristics of the 387 travelers in the study cohort have been described previously.[6] A majority of travelers (75.5%) had traveled to Asia on a previous occasion. There were no travelers infected with JE virus during travel to Asia as assessed by JE IgG seroconversion or clinical disease. As a result, the incidence density was zero cases of JE infection per 10,000 traveler-days (95% CI 0–3.9). During a 1-year period (2007–2008) of the study, JE vaccine was unavailable in Australia. Only 35 travelers (9%) were given JE vaccine prior to travel and they were excluded from the incidence-density analysis.

The potential risk factors for JE infection were considered. Twenty-seven percent of travelers had a trip duration of 30 days or more and 55% (n = 214) reported one or more overnight stays in rural destinations. Peak travel periods generally coincided with the rainy season for several Southeast Asian (SEA) countries (May to October). Of all the traveler-days in the study (n = 11,840), only 16% of the traveler-days were spent doing outdoor activities (hiking, camping, rock climbing, fishing, water skiing, and diving), 55% of travelers stayed overnight in a rural location, and 1% reported camping outdoors.

Adherence with mosquito repellent use was reported in 298 (81%) travelers, and 231 (61%) used either one or more of mosquito coils, nets, and long-sleeved clothing. Approximately 15% used no preventive measures.

Of the travelers who completed the follow-up consultation, 363 travelers had no evidence of immunity to JE (post-travel antibodies ≤10). Low to moderate positive stationary (pre and post) antibody titers for JE (titers 40–80) were observed in 11 travelers of whom one had pre- and post-travel antibody titer levels of 80. Two of these 11 travelers recalled past vaccination for JE prior to travel. Nonspecific levels of antibodies (>10–20) were observed in 13 travelers of whom 8 recalled past JE vaccination.


There were no seroconversions for JE infection or clinical illnesses consistent with JE infection reported in this prospective cohort of Australian travelers. Interpretation of the 95% CIs around the estimate of zero cases of JE per 10,000 traveler-days should be done with care. Travelers have been infected in the past, so the true population risk is not zero. However, the upper bound of 3.9/10,000 traveler-days calculated is best thought of as indicative only, as it is affected by the sample size and the method of calculation. In addition, the CI is difficult to compare with the previous World Health Organization (WHO) crude estimate of one JE infection per million travelers as the latter estimate does not account for duration of exposure. Additionally, the epidemiology of JE infection worldwide and of travelers has changed considerably since the derivation of the WHO estimate. A high proportion of the earlier published cases of JE have been in the United States and allied military personnel stationed in SEA regions. From 1945 to 1972, 131 cases of JE were reported in military personnel. In the years 1978 to 1992 and 1992 to 2008, 24 cases and 21 cases were reported, respectively. Rates of 0.1–2.1 per 10,000 per week have been observed in nonimmunized US military personnel in Asia.[7] JE vaccination is recommended for this group of travelers.

JE infection has been reported in short-term travelers who have traveled outside of the rainy season with minimal travel to rural locations.[3, 8-10] This has raised concerns about the limits in our current understanding of the risk of JE infection in short-term travelers. Characterizing the current risk of JE in general travelers and the uncertainty limits around this risk provides valuable information to travel medicine practitioners advising prospective travelers.

In our cohort of predominantly short-term travelers, travel was more common in periods of the year where JE transmission is higher and whilst nearly half of the travelers visited or stayed in a rural area overnight, only a small proportion of travel-days were spent on “outdoor” activities. The risk of JE infection is linked to outdoor exposure in the dusk or evening times in rural destinations where JE transmission occurs. In terms of adherence to pre-travel advice, most travelers utilized some form of mosquito preventive behavior, although consistency of use was not documented. Only a small proportion of travelers (9%) were vaccinated for JE, which probably reflects the current recommendations for JE vaccine, in that a majority were short-term travelers and not spending a great amount of time in rural areas. Low-level antibodies at baseline were noted in 2.8% of travelers with possibilities of previous JE, given the presence of JE in Northern Australia, or other flavivirus vaccination or infection as possible explanations.

A limitation of this study is the potential impact of a small sample size on the likelihood of observing an infrequent infection such as JE (clinical or subclinical) in travelers. A further limitation is the generalizability of the findings from a travel-clinic attendee cohort who may be different to general travelers. Data were also incomplete for the travelers who did not complete the study. Although unlikely, it is also possible that some seroconversions were missed given the timing of the second bleed (day 10).

Several considerations relating to risk factors for infection, adverse effects and costs of vaccine, and individual personal preference regarding vaccination, need to be considered when discussing indications for or against vaccination. The threshold for JE vaccination is generally still based on historical risk-benefit considerations that may no longer be valid now as we have a safer vaccine. However, the absence of any cases or seroconversions for JE supports the current advice against vaccination for short-term travelers to JE endemic countries who will engage in low-risk activities, and provides novel quantitative risk data which can be used as a basis for recommendations.


This study was funded by an investigator initiated unrestricted grant from Sanofi-Pasteur.

Declaration of Interests

C. L. is an employee of Sanofi-Pasteur. J. T. has received a speaking honoraria from Sanofi-Pasteur. The other authors state they have no conflicts of interest to declare.