Digital Biomarkers in Telehealth: Development, Validation, and Clinical Utility in Chronic Disease Management

Abstract

Digital biomarkers—quantifiable physiological and behavioural data collected through connected sensors and digital devices—have emerged as a foundational element in modern telehealth ecosystems. Their integration into chronic disease management offers significant advancements in early diagnosis, continuous monitoring, personalized treatment, and predictive risk modelling. This research systematically analyses the development, validation, and clinical utility of digital biomarkers for chronic diseases including cardiovascular disorders, diabetes, respiratory conditions, and neurodegenerative diseases. By evaluating sensor technologies, signal analytics, machine learning pipelines, clinical validation frameworks, and regulatory considerations, this study provides a comprehensive multi-dimensional view of digital biomarker deployment. Using a mixed-method case study of 150 telehealth patients and multi-model evaluation (Random Forest, XGBoost, LSTM, Transformer), the research highlights how digital biomarkers enable more accurate interventions, reduce hospital dependency, and improve long-term outcomes. Findings indicate that transformer-based biomarker models outperform conventional algorithms in early risk detection, while LSTM models exhibit superior temporal accuracy. Ultimately, the study proposes a policy roadmap for integrating validated digital biomarkers into national telehealth programs to enhance chronic disease management.