Zhiwei Xu1, Hilary Bambrick1, Laith Yakob2, Gregor Devine3, Francesca D Frentiu4, Rina Marina5, Pandji Wibawa Dhewantara6, Roy Nusa7, R Tedjo Sasmono8, Wenbiao Hu9. 1. School of Public Health and Social Work, Queensland University of Technology, Brisbane, 4059, Australia; Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, 4059, Australia. 2. Department of Disease Control, London School of Hygiene and Tropical Medicine, London, WC1H 9SH, UK. 3. Mosquito Control Laboratory, QIMR Berghofer Medical Research Institute, Brisbane, 4006, Australia. 4. Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, 4059, Australia; School of Biomedical Sciences, Queensland University of Technology, Brisbane, 4059, Australia. 5. Center of Public Health Effort Research and Development, National Institute of Health Research and Development, Jakarta, 10560, Indonesia. 6. UQ Spatial Epidemiology Laboratory, School of Veterinary Science, University of Queensland, Gatton, 4343, Australia; Pangandaran Unit for Health Research and Development, National Institute of Health Research and Development, Ministry of Health of Indonesia, Pangandaran, 46396, Indonesia. 7. Indonesian Ministry of Health, Jakarta, 12950, Indonesia. 8. Eijkman Institute for Molecular Biology, Jakarta, 10430, Indonesia. 9. School of Public Health and Social Work, Queensland University of Technology, Brisbane, 4059, Australia; Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, 4059, Australia. Electronic address: w2.hu@qut.edu.au.
Abstract
BACKGROUND: Although the association between dengue in Bali, Indonesia, and imported dengue in Australia has been widely asserted, no study has quantified this association so far. METHODS: Monthly data on dengue and climatic factors over the past decade for Bali and Jakarta as well as monthly data on imported dengue in Australia underwent a three-stage analysis. Stage I: a quasi-Poisson regression with distributed lag non-linear model was used to assess the associations of climatic factors with dengue in Bali. Stage II: a generalized additive model was used to quantify the association of dengue in Bali with imported dengue in Australia with and without including the number of travelers in log scale as an offset. Stage III: the associations of mean temperature and rainfall (two climatic factors identified in stage I) in Bali with imported dengue in Australia were examined using stage I approach. RESULTS: The number of dengue cases in Bali increased with increasing mean temperature, and, up to a certain level, it also increased with increasing rainfall but dropped off for high levels of rainfall. Above a monthly incidence of 1.05 cases per 100,000, dengue in Bali was almost linearly associated with imported dengue in Australia at a lag of one month. Mean temperature (relative risk (RR) per 0.5 °C increase: 2.95, 95% confidence interval (CI): 1.87, 4.66) and rainfall (RR per 7.5 mm increase: 3.42, 95% CI: 1.07, 10.92) in Bali were significantly associated with imported dengue in Australia at a lag of four months. CONCLUSIONS: This study suggests that climatic factors (i.e., mean temperature and rainfall) known to be conducive of dengue transmission in Bali can provide an early warning with 4-month lead time for Australia in order to mitigate future outbreaks of local dengue in Australia. This study also provides a template and framework for future surveillance of travel-related infectious diseases globally.
BACKGROUND: Although the association between dengue in Bali, Indonesia, and imported dengue in Australia has been widely asserted, no study has quantified this association so far. METHODS: Monthly data on dengue and climatic factors over the past decade for Bali and Jakarta as well as monthly data on imported dengue in Australia underwent a three-stage analysis. Stage I: a quasi-Poisson regression with distributed lag non-linear model was used to assess the associations of climatic factors with dengue in Bali. Stage II: a generalized additive model was used to quantify the association of dengue in Bali with imported dengue in Australia with and without including the number of travelers in log scale as an offset. Stage III: the associations of mean temperature and rainfall (two climatic factors identified in stage I) in Bali with imported dengue in Australia were examined using stage I approach. RESULTS: The number of dengue cases in Bali increased with increasing mean temperature, and, up to a certain level, it also increased with increasing rainfall but dropped off for high levels of rainfall. Above a monthly incidence of 1.05 cases per 100,000, dengue in Bali was almost linearly associated with imported dengue in Australia at a lag of one month. Mean temperature (relative risk (RR) per 0.5 °C increase: 2.95, 95% confidence interval (CI): 1.87, 4.66) and rainfall (RR per 7.5 mm increase: 3.42, 95% CI: 1.07, 10.92) in Bali were significantly associated with imported dengue in Australia at a lag of four months. CONCLUSIONS: This study suggests that climatic factors (i.e., mean temperature and rainfall) known to be conducive of dengue transmission in Bali can provide an early warning with 4-month lead time for Australia in order to mitigate future outbreaks of local dengue in Australia. This study also provides a template and framework for future surveillance of travel-related infectious diseases globally.