One Digital Health: New Protocols for Integrated Surveillance

The health of humans, domestic and wild animals, plants, and the environment is closely interconnected. Large-scale ecological and societal changes have amplified the complexity of infectious disease emergence and spread, exposing limitations in traditional surveillance systems in integrating cross-domain information, detecting early signals, and supporting timely public health decision-making. The One Health paradigm, together with its digital extension, One Digital Health, provides a conceptual and methodological framework to address this complexity through data integration and advanced analytical approaches. This doctoral thesis aims to contribute to the development of One Digital Health surveillance protocols by integrating classical epidemiological approaches with statistical modelling and artificial intelligence for the analysis of multi-domain data across the human–animal–environment interface. The thesis presents a series of complementary studies applying and evaluating the One Digital Health framework across different pathogens and surveillance contexts to explore methodological approaches and assess their transferability. First, a scoping review on zoonotic virus spillover, combined with an evidence-based risk-ranking approach, prioritised pathogens with potential epidemic or pandemic relevance, supporting targeted and integrated surveillance planning. Second, a systematic review of One Health surveillance systems analysed operational aspects, enabling factors, barriers, and evaluation practices in integrated surveillance, highlighting key challenges in implementation across domains. Third, a modelling study on tick-borne encephalitis in the Veneto Region demonstrated how the integration of environmental and climatic data can support the identification of risk drivers and the exploration of spatial and temporal changes under different climate scenarios, informing adaptive and territorially oriented surveillance strategies. Fourth, two systematic reviews and meta-analyses assessed the application of artificial intelligence in surveillance-related contexts, including early diagnosis of mosquito-borne diseases and surveillance of healthcare-associated infections, highlighting gaps between high analytical performance and real-world impact evaluation. Finally, a modelling study on respiratory syncytial virus hospitalisations in older adults illustrates how statistical approaches can be used to estimate underdetection of disease burden by integrating multiple data sources, complementing routine surveillance data where direct observation is incomplete. Across the qualitative and quantitative analyses included in this thesis, the findings highlight both persistent gaps in current surveillance systems and the potential contribution of the One Digital Health framework to address them. While advanced analytical approaches, including multi-domain data integration, artificial intelligence, and modelling, can strengthen infectious disease surveillance, their effective adoption remains contingent on clearly defined surveillance objectives, robust governance structures, data interoperability, and methodological transparency. Overall, this thesis contributes methodological and conceptual elements to the development of pragmatic and scalable One Digital Health surveillance protocols, guiding the design and evaluation of integrated surveillance approaches across contexts and offering a basis for further research toward operational implementation.