A climate-based prediction model in the high-risk clusters of the Mekong Delta region, Vietnam: towards improving dengue prevention and control
MetadataShow full item record
Objective To develop a prediction score scheme useful for prevention practitioners and authorities to implement dengue preparedness and controls in the Mekong Delta region (MDR). Methods We applied a spatial scan statistic to identify high-risk dengue clusters in the MDR and used generalised linear-distributed lag models to examine climate–dengue associations using dengue case records and meteorological data from 2003 to 2013. The significant predictors were collapsed into categorical scales, and the β-coefficients of predictors were converted to prediction scores. The score scheme was validated for predicting dengue outbreaks using ROC analysis. Results The north-eastern MDR was identified as the high-risk cluster. A 1 °C increase in temperature at lag 1–4 and 5–8 weeks increased the dengue risk 11% (95% CI, 9–13) and 7% (95% CI, 6–8), respectively. A 1% rise in humidity increased dengue risk 0.9% (95% CI, 0.2–1.4) at lag 1–4 and 0.8% (95% CI, 0.2–1.4) at lag 5–8 weeks. Similarly, a 1-mm increase in rainfall increased dengue risk 0.1% (95% CI, 0.05–0.16) at lag 1–4 and 0.11% (95% CI, 0.07–0.16) at lag 5–8 weeks. The predicted scores performed with high accuracy in diagnosing the dengue outbreaks (96.3%). Conclusion This study demonstrates the potential usefulness of a dengue prediction score scheme derived from complex statistical models for high-risk dengue clusters. We recommend a further study to examine the possibility of incorporating such a score scheme into the dengue early warning system in similar climate settings.
Tropical Medicine and International Health
Public Health and Health Services not elsewhere classified