Disease Identification and Profiling Using Linked Health Data

Abstract The increasing availability of big data in healthcare-particularly routinely collected healthcare data-offers new opportunities for the identification, detection, and profiling of chronic diseases. This session will focus on how these data can be harnessed to improve our understanding of complex chronic conditions, with a focus on neurological disorders. Healthcare big data are invaluable due to their scale, granularity, and ability to capture patterns of healthcare resource utilization. When linked with clinically detailed sources such as disease registries or electronic health records, these data support both traditional epidemiological analyses and more advanced methods, including machine learning. Such integration enables more accurate identification of chronic conditions and more comprehensive profiling of individuals’ health trajectories. Importantly, these linked datasets can reveal conditions that may otherwise remain undetected or misclassified in routine clinical practice. An essential component of this work involves addressing undiagnosed cases and improving disease ascertainment, especially in high-burden areas like neurodegenerative diseases. Phenotyping algorithms-developed using both classical biostatistical techniques and machine learning-play a key role in identifying cases from complex, often incomplete data. Ensuring the quality and validity of these algorithms is critical, and methods such as capture-recapture analysis and assessment of missing data patterns are important tools for evaluating data robustness. This session will highlight how the combination of routinely collected data and advanced analytical tools can enhance chronic disease monitoring and support more efficient, evidence-informed healthcare planning. Use cases will focus on index conditions such as Alzheimer's disease, other dementias, and multiple sclerosis, demonstrating the real-world impact of these methods on disease surveillance, policy development, and clinical decision-making.