Dataset for Scaling up the Early Warning and Response System for climate-sensitive disease outbreaks (EWARS-csd) in the Dominican Republic.

Background: Dengue is considered the fastest-spreading mosquito-borne disease in the world affecting tropical and sub-tropical countries and causing periodic waves of devasting outbreaks. With the purpose of seeking tailored strategies for outbreak control, EWARS-csd has become the most promising strategy to predict dengue outbreaks and trigger the deployment of early response activities.
Methods: Mixed-methods study of two components: the evaluation of the statistical performance of the EWARS-csd tool and stakeholders’ assessment of its pilot implementation process in the Dominican Republic. Hospitalised dengue cases from 12 provinces of three endemicity levels and meteorological data from 2017 to 2023 were used for the validation of the model. Semi-structured interviews and a focus group discussion were conducted to gather stakeholders’ insights. ATLAS.ti for Mac (V.24.1.1) for data analysis and thematic analysis method was performed.
Results: The statistical performance of EWARS-csd across endemicity levels was optimal. Overall, the low endemicity provinces presented the best statistical performance with all the statistical measurement metrics above 0.7, except for one. The high endemicity provinces showed a comparatively less optimal statistical performance; however, meaningful information was produced in all 12 pilot provinces regarding predicted outbreak probability and predicted disease incidence between 5 to 12 weeks following the observation period. In most provinces the predicted incidence correlated with the observed incidence. As 2024 was an epidemic year, the observed incidence rates were well above the endemic channel. The stakeholder perspective showed that the biggest challenges are the availability of meteorological and entomological data, scarcity of workforce and timely allocation of a budget for outbreak response.
Conclusions: Nation-wide EWARS-csd implementation is feasible. Optimal statistical performance of the EWARS-csd tool was observed and multiple favourable factors to fully implement the tool throughout the country were identified.