to treatment effect?

For example, using step count as a digital measure of functional status is meaningful in COPD or oncology trials, but may be less applicable in acute care settings.

Device Selection and Validation

Choose a wearable that balances accuracy, patient usability, and regulatory acceptance:

• Regulatory Grade: FDA-cleared or CE-marked devices preferred for pivotal trials
• Sensor Specifications: Ensure relevant metrics (e.g., 3-axis accelerometer, PPG, ECG) match endpoint needs
• Comfort and Wearability: Evaluate patient burden and likelihood of long-term compliance
• Data Format & Exportability: Device should support raw data access, timestamping, and EDC integration

Collaborate with tech vendors early to confirm software stability, firmware update protocols, and available APIs for data transfer.

Patient Onboarding and Site Training

CROs must ensure seamless patient onboarding:

• Provide patients with illustrated user guides and multilingual app interfaces
• Implement in-clinic simulations to confirm device usage understanding
• Establish support channels for troubleshooting and resupplies

Site staff should be trained in:

• Device setup and sync procedures
• Data collection SOPs and documentation
• Audit trail generation for compliance with 21 CFR Part 11

Internal SOPs should clearly define responsibilities for device dispatch, collection, and data quality review.

Data Capture Integrity and Audit Readiness

Maintaining data quality is paramount. Recommended practices include:

• Automatic time-stamping and geotagging for each data point
• Real-time data sync with alert triggers for data gaps & anomalies
• Redundant cloud storage to prevent loss during site transitions
• Data hash tagging for secure audit trail validation

Data governance SOPs must align with FDA’s Digital Health Policies and include backup strategies, eConsent integration, and version control for device firmware.

Signal Extraction and Data Cleaning

Wearable data is often noisy and voluminous. Extracting meaningful endpoints requires:

• Predefined Algorithms: Use validated algorithms to derive metrics like step count, HRV, or sleep efficiency
• Handling Missing Data: Establish thresholds for acceptable missingness (e.g., 10% of expected daily signal)
• Drift Detection: Monitor for signal degradation or baseline shifts due to battery, skin impedance, or sensor displacement

Employ centralized analytics hubs or CRO partners who specialize in wearable signal processing to ensure data integrity. It’s critical to document the version and parameters of every algorithm used in the statistical analysis plan.

Regulatory Alignment and Documentation

For trials using wearable-derived endpoints, sponsors should proactively engage regulators:

• FDA: Use the pre-submission process for feedback on digital biomarkers and their proposed context of use
• EMA: Refer to the EMA’s qualification opinion procedures for novel methodologies
• ICH: Align wearable usage with ICH E6(R3) and E8(R1) quality frameworks

Documentation must include:

• Device manuals and version history
• Raw and derived dataset structures
• Data flow diagrams from device to analysis dataset
• Audit trail exports and SOPs

See also this detailed guidance on SOPs for Wearable Compliance in GxP settings.

Case Study: Decentralized Trial in Rheumatoid Arthritis

A mid-sized sponsor conducted a 12-week decentralized trial in RA patients, using wearables to monitor flare frequency via activity dropouts. Key best practices from the trial:

• HRV drops >20% preceded flare episodes in 68% of patients
• Missing data above 30% was linked to non-compliance; reminder notifications reduced this by half
• Wearable adherence >80% was associated with improved ePRO correlations

These insights not only supported exploratory endpoint inclusion but also reduced reliance on clinic-based PRO collection.

Common Pitfalls and How to Avoid Them

• Assuming Device Validity: FDA clearance ≠ validated clinical endpoint; validate context of use
• Neglecting Site Burden: Sites need centralized dashboards and reporting tools for efficient monitoring
• Forgetting Firmware Management: Auto-updates can create versioning inconsistencies in signal analysis
• Ignoring Device Drift: Regular calibration checks must be built into the monitoring plan

Conclusion: Standardization and Vigilance Are Key

Wearable data collection in clinical trials is no longer novel—it’s becoming a mainstream approach for capturing digital biomarkers, patient activity, and adherence data. However, successful implementation requires rigorous planning, device vetting, site and patient support, data integrity assurance, and continuous monitoring.

With growing expectations from regulators, sponsors and CROs must treat wearable-derived data with the same GxP rigor as lab or imaging data. Those who adopt best practices will unlock richer insights, enhance patient-centricity, and future-proof their trials for the next wave of digital innovation.