Forecast of dengue cases in 2025: what experts say
In 2024, an exceptional number of dengue cases was reported worldwide. In Brazil, according to the Ministry of Health, the disease affected more than 6.5 million people by mid-November, with over 5,700 confirmed deaths. It was a year in which dengue spread to southern regions and higher altitudes, areas previously unaffected or rarely experiencing outbreaks.
Experts from Oswaldo Cruz Foundation (Fiocruz), Universidad del Valle, Fundação Getúlio Vargas (FGV), and the Ministry of Health of Brazil gathered on 31 October in a webinar to discuss the upcoming 2025 dengue season using statistical predictive models. The event was organized by the Vice-Presidency of Education, Information, and Communication at Fiocruz (VPEIC/Fiocruz), via the Fiocruz Program for Scientific Computing (PROCC), in partnership with The Global Health Network Latin America and Caribbean Consortium (TGHN LAC).
During the online seminar, they discussed the current dengue situation in Brazil and shared statistical model predictions for the upcoming year, as well as answering participants’ questions. The event had over 1,200 registrants and more than 600 attendees.
Missed the webinar? Watch the event recording to stay informed about the 2025 dengue season.
What experts say about dengue in 2025
The experts explained that many uncertainties remain regarding dengue in 2025, given the atypical year experienced in 2024 both climatically and epidemiologically. The models do not predict an outbreak of the same magnitude as 2024, but warnings remain for certain regions, such as the south of the country, which may experience an even greater outbreak.
The researcher from the Barcelona Supercomputing Center and event mediator, Rachel Lowe, explained that in June 2024, Infodengue and Mosqlimate launched a dengue forecast sprint focused on the 2025 season. The initiative aimed to standardize predictive model training to develop a joint forecast for dengue in Brazil.
The coordinator of Infodengue, Claudia Codeço, described Infodengue as an early warning system for dengue, Zika, and Chikungunya in Brazil, led by Fiocruz and FGV, with support from the Ministry of Health of the country. She highlighted the spread of dengue transmission to municipalities previously unaffected by the arbovirus and noted how the uncertain 2025 climate could influence epidemic risks. She also emphasized the relatively high incidence of dengue notifications during the winter season.
The leader of the Mosqlimate platform and of the dengue predictive modelling task force, Flávio Coelho, presented results from the modelling task force, which brought together national and international expert groups.
The researcher from Universidad del Valle in Colombia, Mauricio Vegas, discussed how climate and environmental changes affect the transmission of arboviruses like dengue.
The technical consultant from the Ministry of Health of Brazil, Marcela Lopes Santos, emphasized the ongoing collaboration with Infodengue, highlighting the technology transfer and use of nowcasting (adjusting delayed data) to improve disease monitoring in Brazil—a model that could be replicated globally.
How to minimize dengue impacts in Brazil
The researchers demonstrated that achieving greater forecast accuracy requires investments in expanding surveillance, diagnostics, and locally focused attention to factors such as seasonality and the circulation of new virus types. These measures could enhance model calibration and the design of more effective health responses to mitigate dengue's impact on the country.
Many webinar participants raised important questions about predictive models and arbovirus monitoring, asking about the availability of models incorporating climate and entomological data to improve local forecasts.
The experts suggested using data platforms like Mosqlimate and reinforced the need for customized models, as factors like collective immunity and socioeconomic variables can influence outcomes.
Entomological surveillance
Some participants questioned the need for a standardized national system to monitor mosquito populations. While a uniform data collection system is not yet in place nationwide, researchers noted that some municipalities already use trap-based strategies, which can provide important data for predictive models.
They explained that integrating entomological and epidemiological data efficiently is crucial for incorporation into predictive models, even at the local level. They also mentioned ongoing efforts to include entomological data in their Infodengue analyses.
Monitoring other arboviruses
Among the main questions raised was the potential underreporting of arboviruses like Mayaro virus (MAYV) and Oropouche virus, which could be mistaken for dengue. The experts acknowledged that this presents a significant challenge for surveillance, given the presence of multiple circulating viruses. They emphasized the importance of good laboratory test coverage considering the broader circulation of these arboviruses.
Application of predictive models
Participants expressed curiosity about the predictive models presented by the speakers, especially the variables used, which include climatological data and case notifications.
Researcher Flávio Coelho responded that the models use both climate data and case notifications. He explained that more detailed models at the intramunicipal level are possible but may require distinct methodologies, such as spatiotemporal models. Flávio recommended consulting Mosqlimate’s GitHub and the task force's technical report.
Want to learn more about InfoDengue? Click here to explore the platform.
By Luisa Picanço (Fiocruz)