An open challenge to advance probabilistic forecasting for dengue epidemics.

A wide range of research has promised new tools for forecasting infectious disease dynamics, but little of that research is currently being applied in practice, because tools do not address key public health needs, do not produce probabilistic forecasts, have not been evaluated on external data, or do not provide sufficient forecast skill to be useful. We developed an open collaborative forecasting challenge to assess probabilistic forecasts for seasonal epidemics of dengue, a major global public health problem. Sixteen teams used a variety of methods and data to generate forecasts for 3 epidemiological targets (peak incidence, the week of the peak, and total incidence) over 8 dengue seasons in Iquitos, Peru and San Juan, Puerto Rico. Forecast skill was highly variable across teams and targets. While numerous forecasts showed high skill for midseason situational awareness, early season skill was low, and skill was generally lowest for high incidence seasons, those for which forecasts would be most valuable. A comparison of modeling approaches revealed that average forecast skill was lower for models including biologically meaningful data and mechanisms and that both multimodel and multiteam ensemble forecasts consistently outperformed individual model forecasts. Leveraging these insights, data, and the forecasting framework will be critical to improve forecast skill and the application of forecasts in real time for epidemic preparedness and response. Moreover, key components of this project-integration with public health needs, a common forecasting framework, shared and standardized data, and open participation-can help advance infectious disease forecasting beyond dengue.

Climate services for health: predicting the evolution of the 2016 dengue season in Machala, Ecuador.

BACKGROUND: El Niño and its effect on local meteorological conditions potentially influences interannual variability in dengue transmission in southern coastal Ecuador. El Oro province is a key dengue surveillance site, due to the high burden of dengue, seasonal transmission, co-circulation of all four dengue serotypes, and the recent introduction of chikungunya and Zika. In this study, we used climate forecasts to predict the evolution of the 2016 dengue season in the city of Machala, following one of the strongest El Niño events on record. METHODS: We incorporated precipitation, minimum temperature, and Niño3·4 index forecasts in a Bayesian hierarchical mixed model to predict dengue incidence. The model was initiated on Jan 1, 2016, producing monthly dengue forecasts until November, 2016. We accounted for misreporting of dengue due to the introduction of chikungunya in 2015, by using active surveillance data to correct reported dengue case data from passive surveillance records. We then evaluated the forecast retrospectively with available epidemiological information. FINDINGS: The predictions correctly forecast an early peak in dengue incidence in March, 2016, with a 90% chance of exceeding the mean dengue incidence for the previous 5 years. Accounting for the proportion of chikungunya cases that had been incorrectly recorded as dengue in 2015 improved the prediction of the magnitude of dengue incidence in 2016. INTERPRETATION: This dengue prediction framework, which uses seasonal climate and El Niño forecasts, allows a prediction to be made at the start of the year for the entire dengue season. Combining active surveillance data with routine dengue reports improved not only model fit and performance, but also the accuracy of benchmark estimates based on historical seasonal averages. This study advances the state-of-the-art of climate services for the health sector, by showing the potential value of incorporating climate information in the public health decision-making process in Ecuador. FUNDING: European Union FP7, Royal Society, and National Science Foundation.