CLN - Ask The Expert
Wearable devices are now generating clinically relevant data at an unprecedented tsunami scale. Patients with diabetes, hypertension, obesity, and cardiovascular disease produce endless data from continuous glucose monitors (CGM), wearable electrocardiography patches, blood pressure devices, pulse oximeters, and sleep trackers. Yet most of this data remains within proprietary apps and is not integrated into the electronic health record (EHR), resulting in a disconnect from clinical decision-making.
Labs should take deliberate steps to integrate wearable data into existing laboratory data systems. First, Food and Drug Administration (FDA)-approved devices should be treated as analytical tools that must be validated with the same rigor that we use to validate laboratory assays. Second, labs should use selective, threshold-based integration into existing laboratory information systems (LIS) rather than attempting to import raw continuous data streams.
If laboratories do not engage, wearable data will be interpreted without analytical oversight, and clinical errors will follow. Clinical laboratories are uniquely positioned to serve as the integration hub with device manufacturers because we already specialize in validating methodologies, managing longitudinal patient data, and defining clinical thresholds. This is not a disruption but a natural extension of what we already do well.
CGM extends clinical insight beyond HbA1c by capturing time in range and glycemic variability (Diabetes Care 2024; doi.org/10.2337/dci24-0073). The current consensus endorses coefficient of variation (CV) as a standardized measure of glycemic variability, with CV >36% indicating clinically significant instability (Diabetes Care 2017; doi: 10.2337/dc17-1600). For example, two patients may have identical HbA1c values yet completely different glycemic variability profiles. It is the patient with high variability who carries greater cardiometabolic burden, but this is not evident from a 3-month HbA1c.
CGM is well established for hypoglycemia prevention through real-time alerts. However, recent evidence suggests that time below range alone does not fully predict severe hypoglycemia and must be interpreted within clinical context (Diabetes Care 2026; doi.org/10.2337/dc25-2353). For hyperglycemia, CGM-integrated insulin systems improve glycemic stability and allow earlier detection of deterioration toward metabolic decompensation.
Beyond glucose, heart rate variability reflects autonomic dysfunction and cardiometabolic risk, while blood pressure wearables are moving toward broader clinical adoption. The laboratory’s role is to integrate these dynamic signals with traditional biomarkers such as lipid profiles, high-sensitivity C-reactive protein, and natriuretic peptides to create a more complete risk assessment.
Data integration must happen at three levels. At the data ingestion level, application programming interfaces that are compatible with fast healthcare interoperability resources allow device platforms to communicate seamlessly with LIS and EHR systems, ensuring interoperable data transfer. At the data processing level, continuous signals must be converted into clinically interpretable summary metrics using standardized validated algorithms. Finally, wearable metrics should be integrated alongside traditional laboratory results in unified, clinical dashboards.
Artificial intelligence (AI) holds promise in this area, but the validation gap remains real. Many models are trained on limited or nonrepresentative datasets, and bias remains an ongoing challenge. In addition, variability in CGM-derived metrics across platforms highlights the urgent need for standardization, similar to the historical harmonization of HbA1c.
Although FDA-approved wearables meet regulatory technical standards, their clinical value depends greatly on manufacturers collaborating with laboratory directors, EHR vendors, informaticists, and guideline committees to ensure standardization, rigorous clinical validation, and seamless data integration, all guided by a shared implementation roadmap.
Interested in learning more? Attend the ADLM 2026 roundtable, “AI-driven precision medicine: Integrating wearable data streams to stratify cardio-metabolic risk and management,” on Monday, July 27, in Anaheim, California.
Earnest J. P. Daniel, PhD, is a clinical chemistry fellow from Baylor College of Medicine and Texas Children’s Hospital in Houston. +Email: [email protected]