Wearables & digital measures

Digital measures for clinical trials, powered by wearables.

Raw signals from 25+ devices become one clinical data layer, with human support watching wear-time and sync so the endpoint stays intact.

Request a demo See the platform

Overview of wearable digital measures for clinical trials

70+Validated digital-endpoint algorithms

95%Device compliance with human support

3,500+Devices provisioned in 40+ countries

The wearables reality no one talks about.

Digital measures rarely fail because the sensors fail. They fail because no one operationalizes the devices.

Patients forget to charge and sync their devices.
Sites are overwhelmed and can’t troubleshoot every wearable.
CROs lack playbooks for device issues.
Data gaps grow quietly until it’s too late to recover them.

Operationalizing wearables at scale in clinical trials

From sensor to clean clinical data.

Connect any wearable or sensor, harmonize the signals into one clinical model, and keep the data clean with automated QC.

Universal device compatibility

Medical wearables, consumer sensors, cellular devices, or BYOD, all in one platform.

Explore integration → 02

Real-time signal harmonization

HR, HRV, SpO₂, sleep, steps, and respiratory data normalized into a consistent clinical data model.

Explore harmonization → 03

Automated QC and signal integrity

Outliers, missing samples, device drift, timestamp errors, and physiological anomalies caught early.

Explore signal QC → 04

Open endpoint APIs

Harmonized endpoints exported directly to EDC, analytics pipelines, or submission packages.

Explore APIs →

Digital measures built to hold up.

Sponsors don’t just need wearables. They need digital measures that are accurate, reproducible, and clinically relevant, with live visibility in one clinical dashboard.

Fit-for-purpose endpoint development, correlated to gold standards
Algorithm training, verification, and FDA-aligned documentation
Live dashboards for patient status, signal stability, and trends
Symptoms, diaries, and objective measures fused into one outcome

Digital endpoints → Regulatory-ready endpoints → Endpoint dashboards → Endpoint + ePRO fusion →

Delve study overview dashboard showing diary completion, wear time, sync health, and signal status

70+ validated algorithms across six modalities.

Every algorithm is documented, version-controlled, and aligned with FDA Digital Health Technology guidance, the DiMe V3 framework, and ICH E6(R3).

ECG & cardiac

Heart rate, HRV, QT and PR intervals, QRS and ST analysis, and AFib burden.

Glucose & CGM

Time in, above, and below range, GMI, glycemic indices, and variability measures.

Activity & movement

Cadence, stride length, distance, MVPA, sedentary time, METs, and energy burned.

Sleep & circadian

Rest-activity rhythm measures plus configurable sleep staging.

Frequency & signal

Spectral analysis and filtering that prepare raw signals for measurement.

General purpose & QC

Signal statistics, outlier and change-point detection, and gait speed, asymmetry, and variability.

The full algorithm library

ECG & cardiac (15)

Heart rate, SDNN (HRV), pNN50, RMSSD, P-wave detector, QRS morphology analyzer, QT interval, T-wave detector, PR interval, ST segment analyzer, CV% of RR intervals, QRS estimator, AFib burden, AFib episodes, average HR during AFib.

Glucose & CGM (10)

High Blood Glucose Index (HBGI), Low Blood Glucose Index (LBGI), M-value, GRADE, Glucose Management Indicator (GMI), J-index, CONGA, Time Above Range (TAR), Time in Range (TIR), Time Below Range (TBR).

Activity & movement (19)

Running and walking cadence, running and walking stride length, stride length estimator, running and walking distance, MVPA (moderate-vigorous activity), sedentary minutes, activity bouts detector, activity classifier, step detector, ENMO calculator, METs calculator, active and basal energy burned, M10 / L5 RA, intradaily variability, interdaily stability.

Sleep & circadian (3+)

M10 / L5 relative amplitude, intradaily variability (sleep), detrend signal, plus configurable sleep staging.

Frequency & signal (8)

PSD calculator (power spectral density), FFT magnitude calculator, spectrogram calculator, Butterworth lowpass, highpass, bandpass, and bandstop filters, notch filter.

General purpose & QC (22)

RMS calculator, signal energy, peak-to-peak, zero-crossing rate, standard deviation, coefficient of variation, array min/max, array average, interpolate missing values, normalize signal, median filter, moving average filter, Savitzky-Golay filter, Z-score and MAD outlier detection, change point detection (PELT), walking and running speed, walking and running asymmetry, walking and running variability.

Aligned with the frameworks regulators and sponsors use

FDA DHT guidance

Digital Health Technologies for Remote Data Acquisition (2024 final).

DiMe V3 framework

Verification, analytical validation, and clinical validation.

ICH E6(R3)

Data integrity, attributability, and version control.

21 CFR Part 11

Electronic records and signatures.

Request the algorithm validation pack See real-time compliance scoring

Automation plus real people, so every device keeps working.

Compliance isn’t a reminder problem. It’s an operational one. Automated triggers catch issues instantly, and human outreach closes the loop.

Missed sync detection and human outreach
Battery, connectivity, and signal monitoring
Device onboarding, troubleshooting, and escalation paths for sites
Global multilingual support, 24/7

Explore Concierge → Compliance engine → See outcomes →

95%Device compliance across global, multi-device studies

87%Retention at 6 months

63%Higher engagement vs. site-only workflows

95%Device uptime

Go deeper.

The detail behind the approach, for teams evaluating wearables in depth.

Wearable modalities supported

Fit-for-purpose device strategies aligned to endpoint sensitivity, patient burden, and operational feasibility.

Activity & mobility

Steps, gait proxies, activity tolerance, and functional change.

Cardiovascular

HR, HRV, ECG (device-dependent), and blood pressure.

Respiratory

SpO₂, respiratory rate, and symptom-linked physiology.

Sleep

Duration, efficiency, and staging where supported.

Metabolic

CGM, weight scales, and longitudinal trend analysis.

Temperature & vitals

Skin temperature and longitudinal vital trends.

The digital endpoint R&D pipeline

Delve is building a defensible pipeline for digital endpoints and AI-enabled study operations, focused on making sensor data submission-grade and reliable in the real world.

Stage 1: Device ingestion

Device ingestion, telemetry, and QC. In production.

Stage 2: Measures library

Signal harmonization and the measures library. In production and expanding.

Stage 3: Validation packs

Endpoint validation packs with gold-standard correlation and reproducibility. Active.

Stage 4: AI-assisted mapping

AI-assisted protocol mapping and risk flags. Active.

Stage 5: Real-world evidence

Post-market and real-world evidence workflows. Roadmap.

This approach supports cardiovascular, oncology, metabolic, respiratory, and chronic disease programs, and extends evidence generation across the full lifecycle.

Validation, deployment, and global scale

Digital measures only matter if they are defensible, deployable, and supported at scale.

FDA/EMA aligned

Fit-for-purpose validation and reproducibility checks for digital submissions.

Provisioning & logistics

Shipping, SIM activation, returns, replacements, and inventory management.

BYOD + medical-grade

Support for both patient-owned and sponsor-provided wearables.

24/7 global support

Technical, patient-facing, and site support teams available worldwide in 40+ countries.

The engagement and compliance engine

Human concierge support

Onboarding, troubleshooting, reminders, and missed sync recovery.

Automated issue detection

Hardware failures, low battery, and connectivity loss caught instantly.

Closed-loop compliance

Automated triggers plus human outreach keep evidence collection continuous.

95%+ compliance

Industry-leading adherence across global, multi-device clinical studies.

Questions sponsors ask about wearables.

How does Delve turn wearable data into a clinical trial endpoint?

Delve runs 70+ validated digital-endpoint algorithms across 25+ devices on one harmonized data layer, then applies signal quality control to confirm the data is complete and usable. Raw sensor signal becomes a defensible digital measure rather than an unvalidated data feed.

Which wearable devices and modalities does Delve support for clinical trials?

Delve supports 100+ device and sensor types across modalities including ECG, glucose (CGM), activity, sleep, respiratory, and cardiovascular signals, via BYOD, provisioned, cellular, and non-app devices, so data collection does not depend on a participant managing an app.

How does Delve protect wearable signal quality and wear-time?

An active signal quality control layer validates completeness, flags drift, and catches sync failures before data is lost, while Concierge-as-a-Service™ recovers wear-time in 120+ languages before participants disengage. Around 95% device compliance is sustained with this human-supported model.

Can Delve run wearables alongside eCOA and ePRO in the same study?

Yes. Delve unifies wearable data with eCOA/ePRO, analytics, and concierge support in one platform, so wearable signal and patient-reported outcomes are managed together rather than across separate vendors.