Severe Dengue Virus Infection in Travelers

Abstract & Commentary

By Omega Edwards, MD, and Maria D. Mileno, MD

Omega Edwards is a Fellow in Infectious Diseases, Section of Infectious Diseases, Department of Medicine, Brown University School of Medicine.

Dr. Edwards reports no financial relationships relevant to this field of study.

Maria D. Mileno is Director of Travel Medicine, The Miriam Hospital, and Associate Professor of Medicine (Infectious Diseases) Director, International Travelers Clinic, Brown University School of Medicine, Providence, RI.

Dr. Mileno is a consultant for GlaxoSmithKline.

Synopsis: There is no uniformly useful definition for severe dengue virus infection and severe dengue may go undiagnosed in returning travelers, if we use the WHO classification scheme for strict diagnosis. Preexisting dengue antibodies were the most significant risk factor for spontaneous bleeding and other severe manifestations of dengue, according to these authors.

Source: Wichmann O, et al. Severe Dengue Virus Infections in Travelers: Risk Factors and Laboratory Indicators. J Infec Dis 2007:195 (15 April) 1089-1096.

Global travel continues to increase, with over 800 million tourist arrivals recorded worldwide in 2005. Increasing numbers of international travelers are returning with dengue fever. The GeoSentinel surveillance network reported that dengue virus infection was the leading cause of febrile illness among travelers returning from every geographic area except for sub-Saharan Africa and Central America. Wichmann et al described 219 patients with acute or recent dengue virus infection in their surveillance of dengue fever among returning European travelers, as reported by the European Network on Surveillance of Imported Infections Diseases.

Intensified study since May 2004 included an additional standardized questionnaire covering dengue-specific symptoms, serological results, hematological values, chemical blood constituents, and history of previous travel and immunizations. Fully 64% of the 219 travelers acquired their dengue infections in either southeast Asia or India. Central America accounted for another 17.8% of cases. There were confirmed diagnoses in 133 patients and probable diagnoses recorded in the remaining 86 individuals. Most cases had documented increasing dengue antibody titers using commercially available Panbio-ELISA assays. PCR confirmation alone was used in two patients. A combination of methods confirmed the diagnosis in 17 patients. There were no significant differences in clinical, laboratory or demographic characteristics between persons with probable and confirmed dengue virus infection. Most common clinical manifestations of dengue included fever (93%), headache (69%), fatigue (57%), rash (53%), muscle pain (50%), retroorbital pain (44%) and a positive tourniquet test in 44% of those tested. In terms of dengue antibody titers, 77% of patients had primary immune response and 23% had a secondary response. Forty-five persons had missing data, and 13 with a secondary response were excluded, due to history of prior flavivirus immunization. Seventeen percent had acquired a secondary dengue virus infection.

In further describing disease severity, 28 patients showed petechiae, 17 had spontaneous bleeding, most often from the nose or gums. One individual showed hematemesis and GI bleeding. Eighty-two patients underwent a tourniquet test and 44% showed positive results.

A positive tourniquet test was associated with occurrence of petechiae but not with spontaneous bleeding. No significant association was noted between a positive tourniquet test and the frequency of severe clinical manifestations, low platelet counts (less than 100,000/µL), non-European origin, secondary immune response or increased liver enzyme (AST) levels to greater than 3 times normal. Leukopenia and thrombocytopenia occurred between day 3 and 6 of illness. Abnormal AST, ALT, and LDH values occurred later than day 6.

Twenty-three percent of patients required hospitalization and the average hospital stay was 4 days. One individual with significant bleeding developed atrial fibrillation and required treatment in the ICU. On patient had dengue infection complicated by visual disturbances manifesting itself as black dots in the field of vision, which resolved gradually over 2 months. One individual had hypotension. A total of 11% (23 patients) demonstrated severe dengue related disease: four had internal hemorrhage; 2 had plasma leakage; one developed shock, and 18 had platelet counts less than or equal to 50,000/µL. In multivariate analysis, only secondary immune responses to dengue viruses and a greater than 3-fold increased serum AST level were independent predictors of severe dengue associated disease.


The WHO clinical case definition for Dengue Hemorrhagic Fever requires the presence of 4 criteria: fever, platelet count less than 100,000 cells/µL, hemorrhagic tendency in addition to evidence for capillary leakage as shown by either an hematocrit increase of greater than 20% from baseline, a pleural effusion, ascites or hypoproteinemia. Only 0.9% of patients in this study met the criteria for severe dengue infection based upon the WHO classification system - it failed to accurately categorize the spectrum of dengue virus infections observed. The authors found that returning travelers with fever, high liver enzymes, and thrombocytopenia are far more likely to have a severe form of dengue virus infection than would be predicted from these criteria. The tourniquet test, although visually stimulating and convincing when positive, also appears to have limited clinical utility. Preexisting dengue antibodies were the most significant predictor of severe disease.

Travel Medicine practitioners may take home a message that persons with prior travel to dengue endemic areas are at risk for a severe dengue infection, keeping in mind that 8 individuals in this cohort visited a dengue endemic country for the first time and yet they still presented with severe dengue. A heightened discussion of personal protection measures is just as necessary for travelers visiting dengue endemic regions and should be as rigorous as the measures recommended for travelers to malaria endemic regions.

Additional Sources:

Freedman DO, et al. Spectrum of Disease and relation to place of exposure among ill returned travelers. N Engl J Med2006:354:119-130.

Wilder-Smith, A and Tambyah, PA. Severe Dengue Virus Infection in Travelers. J Infect Dis 2007:195:1081-1083.