5.Y.2. PechaKucha: Advancing public health epidemiology: applications of machine learning and causal inference methods

Public health epidemiology is continuously evolving, driven by the integration of innovative methodologies that address the complex nature of health challenges. With the increased availability of high-dimensional, population-level, and cross-national data, advanced methodologies offer new opportunities for modeling the complex interactions between risk factors, health outcomes, and disparities across populations. This workshop will focus on the application of cutting-edge methods, such as causal inference models, machine learning, and artificial intelligence (AI)-driven analysis, in the context of epidemiology. These methodologies enable novel insights into disease mechanisms, improve risk prediction, and identify vulnerable populations, ultimately guiding more effective public health interventions. Topics will include causal decomposition techniques to assess interventions targeting cardiovascular health disparities, AI-driven analyses to improve hypertension control strategies across diverse populations, and federated machine learning methods for privacy-preserving cardiovascular risk prediction. Moreover, a key focus of the workshop will be the importance of reproducibility in public health research. Ensuring that findings are transparent, reproducible, and open to further development is critical for advancing scientific knowledge. Special attention will be given to the role of causal graphs, which help make causal assumptions explicit and promote the reuse and modification of epidemiological models. The session aims to foster critical discussion around these methods, encouraging a proactive exchange of ideas among epidemiologists, public health professionals, scientists, and policymakers. Following an interactive format, the session will facilitate open discussion and exchange of ideas among participants. Attendees will have the opportunity to actively engage with speakers to explore advanced and emerging methods in epidemiology, fostering a collective debate on methodological innovation and its implications for public health policy. In conclusion, this workshop aims to foster a deeper understanding of the latest methodologies in public health epidemiology, with a focus on their application to cardiovascular disease. By facilitating discussions on innovative methods and reproducibility, the workshop will provide a platform for sharing knowledge, fostering collaboration, and advancing the field of public health epidemiology. Key messages • Advanced analytical techniques can uncover subtle epidemiological patterns, deepening scientific understanding and guiding more equitable and effective public-health interventions. • Transparency and reproducibility are critical for generating credible evidence that informs policy and practice.