Extreme weather events linked to climate change, such as severe flooding and extended drought periods, frequently lead to serious outbreaks of diarrheal diseases, particularly in developing nations, where these diseases rank as the third most prevalent cause of death among young children. Recently, a study published on October 22, 2024, in Environmental Research Letters, conducted by a global team of researchers led by Amir Sapkota from the University of Maryland’s School of Public Health (UMD SPH), reveals a method to forecast the likelihood of such harmful outbreaks using AI modeling. This advancement could provide public health officials with weeks or even months to get ready and potentially save lives.
“The frequency of extreme weather events due to climate change is expected to rise in the coming years. We need to evolve as a society,” stated Sapkota, who leads the SPH Department of Epidemiology and Biostatistics. “The early warning systems proposed in this research represent a move toward enhancing community resilience against health risks stemming from climate change.”
The diverse team worked across various institutions, utilizing data on temperature, rainfall, past disease occurrences, El Niño climate patterns, and other geographic and environmental conditions from three countries—Nepal, Taiwan, and Vietnam—over a span from 2000 to 2019. This information enabled them to train AI-driven models capable of predicting disease burdens in specific areas weeks to months in advance.
“Having an idea of the expected disease load weeks to months ahead gives public health officials essential time to prepare. This preparedness allows for a more effective response when the time arises,” noted Sapkota.
Although the study primarily focused on Nepal, Vietnam, and Taiwan, lead author Raul Curz-Cano, an Associate Professor at Indiana University School of Public Health in Bloomington, commented, “Our discoveries are quite relevant to other global regions, especially those where communities do not have access to safe drinking water and adequate sanitation facilities.”
Sapkota pointed out that AI’s capability to analyze large data sets positions this study as an initial step towards developing more precise predictive models for early warning systems. He envisions this will empower public health entities to help communities shield themselves from an increased threat of diarrheal disease outbreaks.
The research team consisted of experts from various fields, including atmospheric and oceanic sciences, community health research, and water resources engineering. The contributors hailed from UMD, which includes the Departments of Epidemiology and Biostatistics as well as Atmospheric and Oceanic Science, alongside collaborators from Indiana University School of Public Health in Bloomington, the Nepal Health Research Council, Hue University of Medicine and Pharmacy in Vietnam, Lund University in Sweden, and Chung Yuan Christian University in Taiwan.
Funding for this research was provided by the National Science Foundation via the Belmont Forum (award number (FAIN): 2025470), the Swedish Research Council for Health, Working Life and Welfare (Forte: 2019-01552), the Taiwan Ministry of Science and Technology (MOST 109-2621-M-033-001-MY3 and MOST 110-2625-M-033-002), and the National Science Foundation National Research Traineeship Program (NRT-INFEWS:1828910).
