Published oncology studies most often use wearable sensors to capture physical activity, step count, heart rate, and sleep, then test how those measures relate to performance status, adverse events, hospitalization, and survival. Much of this work describes digital biomarkers and their associations with clinical outcomes. A digital endpoint takes more: a defined concept of interest, validation evidence, and a prespecified role in the trial.
Oncology sponsors are asking a practical question. Which digital measures have actually been studied in patients with cancer, and how close are they to use as trial outcomes? This article walks through published examples, the terms that separate a measurement from an outcome, and the steps that move one toward the other.
What Is the Difference Between a Digital Biomarker and a Digital Endpoint in Oncology?
The two terms are not interchangeable. A digital biomarker is “an objective, quantifiable physiological or behavioral measurement collected through a digital device.” A digital endpoint is “a trial outcome measure that uses one or more digital biomarkers to assess treatment effects, disease progression, or other study objectives.”
Put simply, digital biomarkers are the measurements and digital endpoints are the outcomes derived from those measurements. A wrist-worn sensor’s daily step count is a digital biomarker. A prespecified change in daily step count, analyzed against a defined population and context of use, could become a digital endpoint. For a fuller treatment, see Digital Biomarkers vs. Digital Endpoints.
| Digital biomarker | Digital endpoint |
|---|---|
| A measurement collected through a digital device | A trial outcome built from one or more digital biomarkers |
| Example: daily step count, sleep duration, heart rate | Example: change in a prespecified activity measure over a treatment period |
| Evidence question: does it measure what it claims? | Evidence question: is the outcome meaningful for this population and context of use? |
What Do Wearable Sensors Measure in Oncology Studies?
Physical function is a common focus of oncology work with wearables. Performance status already shapes eligibility, dosing decisions and prognosis in cancer care, and clinicians usually rate it during a short clinic visit. Continuous monitoring offers a view of function between those visits, in the patient’s own environment.
Measurement domains that appear in published oncology research include:
-
- Physical activity and daily step count
-
- Gait and mobility
-
- Sleep duration and efficiency
-
- Heart rate and heart rate variability
-
- Fatigue and energy levels, often paired with patient-reported outcomes
A VivoSense commentary published in January 2025 on measuring physical function in clinical research makes the underlying point. Physical function “is a critical determinant of health-related quality of life,” and digital health technologies (DHTs) allow “continuous, objective monitoring of patients in their natural environments.”
Which Published Oncology Studies Use Wearable-Derived Measures?
The examples below are research studies. Each one reports what its authors found in a specific population with a specific device. None establishes a digital endpoint on its own.
Activity Monitors and Performance Status in Advanced Cancer (2018)
A study published in npj Digital Medicine in July 2018, Wearable activity monitors to assess performance status and predict clinical outcomes in advanced cancer patients, followed 37 adults with advanced solid tumors who wore a consumer wrist-worn activity tracker. The team compared activity metrics such as steps, stairs climbed and heart rate with clinician-rated performance status, patient-reported outcomes, adverse events, hospitalizations and survival.
The authors described the work as a feasibility study. They concluded that activity monitors may help assess performance status and called for larger trials to confirm the findings. That framing matters. Feasibility in a small single-site cohort is a starting point for a digital biomarker, not evidence for an endpoint.
Apple Watch Data and Clinical Events in Patients With Cancer (2025)
VivoSense researchers published an analysis in the Karger journal Digital Biomarkers, summarized on the VivoSense site in March 2025 as Evaluating Digital Biomarkers From Apple Watch Data. This is published research on consumer-grade data, not a recommendation to use a consumer wearable in a trial.
The analysis estimated daily step counts from Apple Watch data in patients with cancer and examined their association with time to a clinical event, including emergency department visits, hospitalizations and palliative care referrals. Higher daily step counts on days with sufficient wear time were associated with a lower hazard of adverse events and mortality.
The finding most relevant to sponsors concerns data handling. Step data used without preprocessing did not show the association. Data that had been processed to classify gaps by duration and context, such as inactivity, non-wear and sleep, did. In that analysis, how the authors handled missing data affected the results.
A Review of Smartphone and Wearable Sensing in Cancer Research (2020)
A 2020 narrative review in npj Digital Medicine, Harnessing consumer smartphone and wearable sensors for clinical cancer research, surveyed the field. It argues that passively collected data could reveal changes in functioning, symptoms and risk of adverse outcomes between clinic visits.
The review is candid about limits. Most supporting studies are small, and the author calls for larger longitudinal work and for “multidisciplinary collaboration” on data management, patient engagement and privacy. That gap between promising associations and trial-ready outcomes is where most oncology programs still sit.
Why Is It Hard to Turn a Digital Biomarker Into an Oncology Endpoint?
Oncology adds pressures that general wearable guidance often doesn’t anticipate.
-
- Wear time changes with the disease. Fatigue, hospital stays, and treatment side effects reduce wear. Missing data is rarely random, so the study needs a plan before it starts.
-
- Context of use is narrow. A sensor that performs well in healthy adults may not perform the same way in patients with cancer-related fatigue or limited mobility. Validation has to be evaluated within the intended context of use.
-
- The concept of interest must matter to patients. Physical function is a strong candidate because patients and clinicians both care about it. A measure chosen because a device happens to capture it is a weaker one.
-
- Regulatory expectations are still developing. The U.S. Food and Drug Administration (FDA) guidance on digital health technologies for remote data acquisition, finalized in December 2023 and issued with the Oncology Center of Excellence among others, sets out nonbinding recommendations on fit-for-purpose selection, verification, validation and usability.
How Do Sponsors Build Toward a Digital Endpoint in an Oncology Trial?
The path follows the same five published stages used across therapeutic areas.
Identify a Meaningful Clinical Concept
Start with the question the trial must answer, such as whether patients maintain physical function during treatment. The concept guides every later choice.
Select Appropriate Technologies
Choose a sensor based on the disease state and the patient population. Wear location, battery life and burden on patients who are already managing treatment all count. Our guide on how to select wearable sensors for a clinical trial covers the criteria.
Develop Digital Biomarkers
Define how raw sensor data becomes a digital biomarker. Specify wear-time rules, gap handling and the algorithm version before data collection begins.
Establish Validation Evidence
Show analytical validity, meaning the measure accurately captures its target, and clinical validity, meaning it relates to outcomes that matter in this population.
Build Digital Endpoints
Prespecify how the measure will be analyzed as a trial outcome and how it fits alongside conventional assessments.
What Mistakes Should Oncology Sponsors Avoid?
-
- Treating an association from a small cohort as evidence that a measure is ready to serve as an endpoint
-
- Choosing the device before defining the concept of interest
-
- Leaving missing-data rules until analysis
-
- Using “biomarker,” “measure” and “endpoint” as synonyms in the protocol
-
- Assuming a measure that performed well in one population transfers to another
How Does VivoSense Support Oncology Sponsors?
VivoSense is a wearable sensor contract research organization (CRO). It works alongside the sponsor’s CRO and covers the digital measurement workstream end to end, from study startup through closeout.
Starting with the sponsor’s clinical question, this work includes helping the sponsor choose the right device based on the disease state and patient population, deciding which measures to capture, shipping devices, training sites, monitoring wear compliance in real time, and cleaning and analyzing data from multiple sensors. VivoSense researchers have published on oncology directly, including the Apple Watch analysis above. To see how the category works, read What Is a Wearable Sensor CRO?
Frequently Asked Questions
What are examples of digital endpoints in oncology trials?
Published oncology research has studied wearable-derived measures of physical activity, step count, heart rate and sleep. These are digital biomarkers. They become digital endpoints only when a trial prespecifies them as outcome measures with supporting validation evidence for that population and context of use.
What is a digital biomarker in oncology?
A digital biomarker is an objective, quantifiable physiological or behavioral measurement collected through a digital device. In oncology, common examples are daily step count and sleep duration recorded by a wearable sensor.
Can consumer wearables be used in oncology trials?
Published research has analyzed consumer-grade wearable data in patients with cancer. That is different from selecting a device for a trial. Sponsors should choose a sensor based on the disease state, the patient population and the evidence needed for the intended context of use.
Why does missing wearable data matter in cancer studies?
Patients with cancer may wear a sensor less on days when they feel worse, so gaps are rarely random. Published work shows that how gaps are classified and handled can change whether a measure shows any association with clinical outcomes.
What does FDA guidance say about wearables in oncology trials?
FDA guidance on digital health technologies for remote data acquisition applies across therapeutic areas, including oncology. It carries nonbinding recommendations on fit-for-purpose selection, verification, validation, usability and data management.
Does VivoSense replace a sponsor’s CRO in an oncology trial?
No. VivoSense works alongside the sponsor’s existing CRO and covers the digital measurement workstream.
Let’s Talk
Schedule a consultation to explore designing and using validated digital measures in your clinical trials.