Precision Healthcare’s Next Act: Digital Biomarkers and Real-Time Personalization
How continuous data, AI, and patient trust are driving precision medicine beyond the clinic
Healthcare is facing a pivotal moment. Outdated, “one-size-fits-all” medical approaches are colliding with modern realities: chronic diseases affect nearly half of adults, and even standard treatments often meet only a fraction of patient needs. For millions, this means navigating a complex, frustrating cycle of missed diagnoses, unnecessary side effects, or simply not feeling heard. The gap between episodic care and people’s real, continuous health journeys can lead to overlooked crises—as in the case of undetected depression, missed early warnings for falls in the elderly, or the day-to-day struggles of those with complex conditions like Long COVID.
This challenge was at the heart of a rich discussion on “The Future of” webcast, where I sat down with Dr. Gareth Sessel, Chief Growth & Product Officer at innovahealth.ai. Dr. Sessel, an Oxford-trained physician-engineer, brings hands-on experience at the intersection of digital biomarkers, clinical innovation, and health AI. Together, we explored why now is a turning point for precision healthcare, the practical tools emerging to bridge persistent care gaps, and what industry leaders must know to keep up with this accelerating future.
Why Healthcare Must Move Beyond “One-Size-Fits-All”
Historically, most of medicine has relied on broad protocols and population averages, delivering care that is “evidence-based” but insufficiently tailored to the individual. As Dr. Sessel pointed out, “Your doctor will treat you according to these guidelines or these protocols... but essentially they’re still designed around a population average rather than the actual individual sitting in front of you.” This generalized approach often leads to trial-and-error cycles that are both inefficient and, in some cases, risky.
Consider chronic disease management: In diabetes or depression, the standard practice involves progressing through a first-line, second-line, and third-line treatment sequence. My own multi-year journey with Long COVID mirrored the experiences of many, mired in subjective interviews and generic interventions - a frustrating process that clearly illustrates the drawbacks of healthcare’s rigid playbooks. This is supported by research in The New England Journal of Medicine, which noted that many drugs are effective for less than 60% of patients, underscoring the urgent need for more personalized approaches.
The Rise of Digital Biomarkers: Practical, Passive, and Scalable
Precision medicine’s initial promise centered on genetic and biochemical tests, but obstacles of cost, episodic measurement, and limited accessibility kept these tools from everyday practice. Improvements in genome sequencing—dropping from $3 billion and 13 years to under $200 in a single day—are remarkable, yet such approaches remain largely siloed and clinic-bound.
Digital biomarkers break down these barriers. Defined as “calculated scores” drawn from a tapestry of digital signals—including wearables, smartphones, smart home devices, and even behavioral data—digital biomarkers provide a living, breathing picture of individual health. Their defining features:
Non-invasive and passive: No labs or clinician visits needed; everyday activities generate useful data.
Continuous, real-world measurement: Health status is tracked in real time, reflecting real life, not the stress of a clinical setting.
Affordable and accessible: From smart scales used in rural homes to wearables in developing countries, digital biomarkers reach where traditional medicine can’t.
A real-world narrative: Traditional monitoring of depression depends on patients’ self-reporting and infrequent check-ins, often missing subtle early warning signs. With digital biomarkers, changes in sleep quality, communication patterns, or social engagement—all easily tracked with consumer devices—can serve as sensitive indicators, prompting timely support before crises arise.
Why This Moment Is Different: Data, Compute, and Validation
The explosion of digital health data would be meaningless without the concurrent leaps in AI-driven analysis and robust data sharing platforms. As Dr. Sessel shared,
“It’s a combination of now we have compute and storage capabilities, the data sharing capabilities, and the model training capabilities that we didn’t have before.”
Key advances:
Mass Adoption of Connected Devices: Wearables and smart devices now collect diverse physiological signals continuously, scaling far beyond the periodic “snapshot” of clinic visits.
Deep Learning and Flexible Models: Machine learning systems can process massive streams of often noisy, real-world data, and are adaptable enough to make sense of incomplete or varying inputs.
Interoperable Platforms and Federated Learning: The rise of secure, distributed training approaches means insights can be developed from diverse sources without centralizing sensitive data - addressing both utility and privacy.
A healthcare startup working with older adults, for example, used wearable step counters and local business transactional data to identify early declines in mobility and lifestyle. When a diner owner noticed a regular hadn’t visited in a week, it prompted a wellness check that uncovered a serious fall - an example of how subtle deviations can trigger life-saving interventions.
Closing the Loop: From Lab to Life, and Back
Digital biomarkers also pave the way for more inclusive, participatory research. Traditional clinical trials are expensive and often inaccessible to many due to travel or complexity. By gathering data passively via consumer devices - say, tracking heart rate or activity with smartwatches - studies can reach underserved populations and reflect the full diversity of real-world living.
Continuous metrics aren’t just useful for trials. They’re also transforming regular care: For example, FDA-cleared AI tools for diabetic retinopathy screening now review eye images quickly and safely, triaging only those patients who actually need to see an ophthalmologist - freeing human experts for more complex cases.
The Power - and Risk - of Proxy Data
Innovators are learning to make use of proxy signals—nonmedical data streams that offer early clues to vulnerability. Shopping behavior, online activity, or financial changes can flag cognitive decline or shifts in mental health, useful for aging populations who want to “age in place.” The nervous system, as Sessel noted, is affected by many diseases; trends in basic physiological metrics can offer fingerprints of emerging problems well before they manifest as symptoms.
Action for Executives: Seek partnerships across sectors—retailers, financial institutions, technology platforms—to responsibly incorporate new proxy data streams into predictive care models. Start small; pilot programs in high-need populations to demonstrate value and navigate early privacy concerns.
Privacy, Trust, and Data Stewardship: The Brand Imperative
Perhaps no topic looms larger for executives than trust. Digital biomarkers and the increasing breadth of personal data make privacy and data governance non-negotiable. “You don’t want sensitive information leaked out there,” said Dr. Sessel, emphasizing the growing public scrutiny around data practices.
Meaningful progress will require:
Transparent patient consent, visible privacy engineering, robust de-identification, and interoperable controls.
Federated learning—where models are trained locally on siloed data—allowing multiple institutions to collaborate without ever exposing raw records. Open source initiatives like ML Commons’ MedPerf and Nvidia Flare are quickly setting benchmarks in responsible health AI.
Brand alignment with patient values: As consumer research in Harvard Business Review and Deloitte shows, organizations known for privacy leadership (such as Apple) earn more trust and, as a result, more willingness from users to share their health data.
Simply put: The ability to access and use sensitive health and behavioral data will increasingly depend on a company’s reputation for trust and ethical stewardship.
Synthetic Data: Fuel for Accelerated, Secure Innovation
Innovation in AI and healthcare shouldn’t come at the cost of real patient privacy. The emergence of synthetic data - algorithmically generated datasets that mimic real patient statistics without ever exposing true identities - solves a major bottleneck in model development and regulatory compliance. As Deloitte highlights, synthetic data now underpins many pharmaceutical R&D trials for rare and high-risk diseases, enabling robust model training and faster patient impact.
From Data Burden to Clinical Partnership
The explosion in available health data - while promising - is creating a risk of information overload for providers. Traditional dashboards often require clinicians to sift through large amounts of data manually, risking missed signals and fatigue.
The solution? Seamless, workflow-embedded recommendations powered by validated AI and digital biomarkers, with “human-in-the-loop” oversight. Dr. Sessel argued,
“A physician aided by these technologies will always outperform a physician who resists these technologies.”
The model is not to replace the expert, but to increase their productivity and reach, especially as diagnostic and treatment pathways become more granular and personalized.
Redefining Disease: From Labels to Personal Trajectories
With richer, more sensitive measurement, we are moving away from umbrella disease terms like “type 2 diabetes” and towards sub-classification, allowing targeted interventions and smarter resource allocation. Recent studies (Nature Medicine, 2018; JAMA 2023) have shown that within the “type 2 diabetes” cohort, certain subgroups are much more likely to develop specific complications; tailored care pathways—enabled by digital biomarkers and analytics—can both improve outcomes and reduce unnecessary interventions.
Five Leadership Imperatives for Executives
Move Beyond the Clinic: Build or partner for platforms that actively gather, integrate, and analyze real-world, patient-generated data alongside traditional health records.
Embed Privacy by Design: Invest in secure federated learning, clear patient consent journeys, and robust de-identification—aligning data strategy directly to brand promise.
Bridge Clinical and Data Science: Cross-train teams or build interdisciplinary groups that include data scientists and practicing clinicians. Encourage pragmatic validation and ongoing “human-in-the-loop” criteria for all system deployments.
Pilot for Outcomes, Not Just Insights: Launch tightly focused pilot programs where new models or biomarkers must demonstrate measurable benefit: improved early detection, reduced hospitalizations, or streamlined care pathways.
Earn and Safeguard Trust: Communicate openly about data use and benefit; involve patients in design; and regularly audit fidelity to values that matter most to your communities and customers.
Closing Perspective
Healthcare stands at a crossroads. The technical building blocks for precision medicine have arrived, but only those organizations that blend innovation with responsibility will lead. As Dr. Sessel put it,
“If these are validated technologies, it’s unethical to not use them.”
Executives who make privacy, trust, and interdisciplinary teamwork part of their execution strategy will define the next era in health—one that is genuinely personal, equitable, and impactful.
For deeper insights, follow leaders like Dr. Gareth Sessel and connect with emerging interdisciplinary communities at innovahealth.ai and similar platforms. The future of health is not just about better data—it’s about delivering better care, with and for real people, at scale.