Contributed Commentary by Dr. Niel Starksen, Vivalink, and T.Y. Alvin Liu, M.D., Johns Hopkins Medicine 

June 20, 2025 | In recent years, AI in healthcare has attracted growing interest and investment, with new roles like Chief AI Officer and AI-specific policies underscoring its transformative potential. However, one critical factor is often overlooked: the quality of the data used to train AI algorithms. While AI’s impact is undeniable, its effectiveness depends on the depth, diversity, and granularity of the data it processes.  

Traditional sources like electronic health records (EHRs) offer only snapshots of a patient’s health. Patient-generated health data (PGHD), however, paints a more comprehensive picture, capturing vital signs, daily behaviors, and lifestyle factors that provide the full context needed to understand a patient’s health. Recent technological advancements, such as remote patient monitoring (RPM) devices, now make it possible to collect this data at scale, providing AI algorithms with the robust datasets needed to improve diagnostics and predictive accuracy. 

Redefining Health with Patient Data 

The terms "real-world data" (RWD) and "real-world evidence" (RWE) have gained significant traction in healthcare. RWD refers to a broad range of health information collected outside controlled clinical trials—including data from registries, public health records, and routine care activities—not necessarily linked to direct clinical visits. 

PGHD, a subset of RWD, is clinically-relevant data generated directly by patients outside of traditional care settings, such as data from wearables, remote monitoring devices, and patient-reported outcomes. RWE emerges from analyzing this data, providing valuable insights that inform clinical practice, policy decisions, and therapeutic development. 

The growing use of PGHD in decision-making reflects a growing recognition of the importance of insights from patients' everyday experiences. In fact, 80% of a person’s health is influenced by factors outside traditional care settings, such as social determinants of health. To fully understand a patient’s health, capturing environmental influences—such as living conditions, activity levels, and climate factors—is essential. 

Technology is bridging this gap. Wearable sensors now track a large volume of daily life data, measuring physical activity, sleep patterns, and environmental conditions like ambient temperature. As these technologies evolve, AI will be able to process increasingly complex and nuanced information, enabling more personalized and precise medical advice—similar to the insights of a highly experienced physician. 

The Role of Continuous Monitoring in Modern Healthcare 

Continuous monitoring is changing the way we care for patients. With devices that track vital signs, activity levels, and environmental factors in real time, healthcare teams get a clearer picture of a patient's health. This constant flow of data helps catch issues early, giving doctors and nurses the chance to act before issues escalate.  

For example, diagnosing hypertension during routine doctor visits can be imprecise, as patients typically measure their blood pressure only a few times daily. This sporadic data often fails to reflect true blood pressure status and can lead to misdiagnosis or inadequate treatment. Continuous monitoring tracks blood pressure fluctuations in real time, offering a more accurate evaluation and supporting earlier interventions. 

Likewise, traditional temperature monitoring for conditions like neutropenic fever and cytokine release syndrome (CRS) relies on intermittent readings taken during clinic visits or at home. These snapshots may overlook critical fluctuations related to infections or fevers. Continuous temperature monitoring provides reliable, real-time data, reducing the risk of false positives or negatives and ensuring no symptoms are missed. 

This constant flow of PGHD helps enhance patient care and can help aid AI development, enabling more accurate diagnostic and predictive models. The future of AI in healthcare depends on the continuous capture, integration, and analysis of PGHD. AI systems must evolve beyond the static confines of hospital records, being constantly updated and fine-tuned to reflect the dynamic complexities of daily life. 

The Future of AI-Driven Medicine  

Algorithms that learn from continuous monitoring data gain deeper insights into disease progression, patient variability, and subtle environmental influences—whether it’s seasonal allergens, daily routine changes, or shifts in ambient temperature. These insights enable AI to anticipate complications, predict outcomes with greater accuracy, and offer more personalized recommendations. 

For healthcare providers, this shift means embracing a more holistic view of patient care, moving beyond diagnoses based solely on clinical visits. Environmental data, such as pollen counts or sedentary versus active states, provides context that can shape treatment plans. AI's ability to process these variables and understand their interplay holds immense potential for improving patient outcomes. 

As AI models advance, they will predict adverse events before they occur. For example, AI-driven analysis of continuous monitoring data can forecast when a patient with congestive heart failure is approaching decompensation, enabling proactive care adjustments to prevent hospitalization. In oncology, AI can detect early signs of immune-related adverse events from cancer therapies by continuously tracking physiological markers that routine exams might miss. 

Success, however, requires collaboration between data scientists, clinicians, patients, and regulatory agencies. Healthcare providers should expect a future where AI enhances human expertise with greater precision and scale. The future of medicine is about blending advanced data, continuous monitoring, and the expertise of healthcare professionals. 

Niel F. Starksen, MD, FACC is a cardiologist, entrepreneur, inventor and investor who’s spent his career innovating breakthrough medtech and digital health technologies. Dr. Starksen is a co-founder and advisor to Vivalink, as well as serving as an advisor to medtech, biotech and digital companies. He serves on the board of governors of the Johns Hopkins Medicine Wilmer Eye Institute, and advisor to the JHU School of Medicine’s James P. Gills Jr. M.D. and Heather Gills Artificial Intelligence Innovation Center. Dr. Starksen is also the founder and previous CEO and Director of Guided Delivery Systems, now Ancora Heart. He can be reached out at mbnfs@protonmail.com.  

Dr. T. Y. Alvin Liu is a professor with a dual faculty appointment at the Johns Hopkins University School of Medicine and School of Engineering. He is also the Inaugural Director of the James P. Gills Jr. M.D. and Heather Gills Artificial Intelligence Innovation Center. As an interdisciplinary strategist at the intersection of venture capital, startup companies and health systems, he specializes in the implementation and scaling of healthcare AI technologies in both clinical and operational domains. At Johns Hopkins Medicine, he is a co-chair of the AI and Data Trust Council, a leadership team that oversees all AI initiates across the entire health system in the imaging, clinical and operational domains.