Digital biomarkers in chronic disease management: a systematic review of personalization and ethical challenges.

Abstract

Introduction: Chronic diseases remain leading contributors to global mortality, yet digital biomarkers offer transformative potential for personalized management. This systematic review evaluates the clinical performance of digital biomarkers across multiple chronic disease areas and the ethical, privacy, and equity barriers to their adoption.

Methods: Following PRISMA guidelines, we searched PubMed (January 2010-August 2024) for studies evaluating digital biomarkers in chronic disease management. English-language, peer-reviewed studies reporting original empirical data were included; reviews and protocols were retained as contextual references. Study quality was assessed qualitatively given heterogeneity in designs and outcome metrics, precluding meta-analysis.

Results: Sixty-eight studies (33 primary empirical, 35 contextual) were included, spanning cardiovascular disease, diabetes, respiratory conditions, neurological disorders, and mental health. Key findings included 100% atrial fibrillation detection sensitivity (KardiaMobile), 92-94% seizure detection sensitivity (Embrace Watch), significant HbA1c improvement with smart insulin pens (p = 0.006), and 87% versus 74% CPAP adherence with smart monitoring. However, device costs ($80-$1,690), regulatory gaps, limited population diversity in validation studies, and poor data interoperability hinder equitable adoption.

Conclusions: Digital biomarkers demonstrate strong clinical utility but require harmonized regulatory standards, diverse-population validation, algorithmic transparency, and expanded reimbursement to achieve equitable, personalized chronic disease management. Evidence remains preliminary, necessitating large-scale, multi-site trials.