Digital biomarkers promise to transform preventive medicine and clinical trials, but also raise issues of security, regulation and potential disparities in access. Deepika Khedekar, Clinical Trial Lead at IQVIA, provides an overview of the pros and cons, and suggests a possible solution.
The human body, a complex tapestry of over 100 trillion cells1 and 25,000 genes2, remains a biological marvel yet to be fully deciphered. However, this robust system can become alarmingly vulnerable when confronted with debilitating conditions like Alzheimer’s, cancer, and cardiovascular diseases. With a staggering 450,000 clinical trials3 registered globally to address these pressing health issues, and an average failure rate of 90%4, the call to accelerate and revolutionise clinical research is paramount. Digital biomarkers could be the cornerstone of this evolution.
The dawn of digital biomarkers in clinical research
Digital biomarkers represent the health data collected from digital devices, like wearables and smartphones, that can help track, predict, or indicate health conditions. They mark a significant shift in clinical research, utilising technology to capture real-time patient data. Beyond the scope of traditional biomarkers, which typically rely on blood tests or tissue samples, digital biomarkers encompass a diverse array of measurements, from heart rate variability and sleep patterns to speech characteristics, offering a comprehensive view of an individual’s health.
The potential of these digital tools is vast. The richness and frequency of data provided by digital biomarkers offer unprecedented insights into patient health and response to treatments, enabling researchers to identify subtle changes and trends that may have gone unnoticed with conventional methods. For instance, the Apple Heart Study by Stanford Medicine utilised the Apple Watch’s pulse sensor to detect atrial fibrillation in 34% of its 400,000 participants, showcasing the profound impact of such tools. This goes beyond mere data collection; it’s a testament to the transformative power of digital biomarkers in shaping the future of preventive medicine and clinical trials. As technology and healthcare merge more closely, it’s the individual success stories that highlight the true potential of this convergence.
Digital biomarkers in cardiovascular trials
The Apple Heart Study5 by Stanford Medicine illustrates the impact of digital biomarkers in clinical trials, specifically in detecting atrial fibrillation (AF). AF is a type of irregular heart rhythm that can lead to stroke and hospitalisation, often going undetected due to its sporadic symptoms. Traditional methods of detecting AF often rely on periodic medical check-ups, which might miss these sporadic symptoms, leading to undetected and untreated conditions. However, by utilising the Apple Watch’s pulse sensor, the study monitored over 400,000 participants for irregular heartbeats. The study’s app intermittently checked for irregular pulses and those flagged were further monitored through an ECG patch, confirming AF in 34% of flagged cases5. This not only demonstrates the efficacy of digital biomarkers in identifying critical health conditions but also signifies a shift towards integrating wearable technology for preventive and patient-centric care.
Just as digital biomarkers are reshaping cardiovascular research, they are also making strides in the realm of cognitive health.
Using AI to detect cognitive health anomalies
With artificial intelligence (AI) at its core, digital biomarkers are providing a novel approach to decipher the intricacies of Alzheimer’s disease, particularly through analyzing speech patterns. Researchers from UT Southwestern Medical Center6 employ a subset of AI systems such as machine learning and natural language processing models to detect subtle changes in a patient’s voice, providing an early indicator of cognitive impairment and Alzheimer’s before visible symptoms manifest. By capturing these otherwise elusive changes, AI-enabled digital voice biomarkers hold the potential to revolutionise the early detection of cognitive decline. The integration of such digital voice biomarkers into clinical trials empowers clinicians to make more informed decisions, allowing patients and families more time to plan for the future and providing greater flexibility in recommending promising lifestyle interventions.
The realm of digital biomarkers also holds tremendous potential in reshaping how we approach respiratory diseases.
A novel approach to detecting lung diseases
Interstitial lung disease (ILD) encompasses a group of disorders that cause scarring (fibrosis) of the lungs, making it harder for patients to breathe. Traditionally, diagnosing ILD has been a complex process. The primary diagnostic tools have been CT scans, which provide detailed images of the lungs, and lung function tests that measure how well the lungs are working. However, these tests, while informative, sometimes don’t provide a complete picture. In certain cases, to pinpoint the specific type of ILD and its severity, doctors might recommend a lung biopsy. This procedure involves removing a small piece of lung tissue for examination. Not only is it invasive, but it also comes with risks, such as potential complications from surgery, and can be quite costly.
The medical community has long sought ways to improve the diagnostic process for ILD, aiming for methods that are less invasive yet equally, if not more, accurate. This is where the promise of digital biomarkers comes into play. Leveraging the advancements in technology, tools such as Fibresolve7 can be used. Fibresolve uses AI algorithms to detect specific patterns in the scans that are indicative of ILD. This aims to provide a more precise diagnosis, without the need for invasive procedures.
Challenges and ethical considerations
Data security and regulatory compliance
With the rise of digital biomarkers in clinical trials, there are critical questions that we will need to answer. For example, how secure is the data being collected? Could a simple breach in a wearable device’s security expose an individual’s personal health data to malicious actors? And as this field rapidly evolves, are there clear regulatory guidelines in place to ensure the accuracy and clinical relevance of these biomarkers? The integration of technology in healthcare not only promises advancements but also raises concerns about security, regulation and potential disparities in access.
Disparities in access to technology and bias and discrimination
How inclusive are clinical trials that rely on smartphone apps or other advanced technologies? Could they inadvertently exclude participants from underserved communities, leading to biased results? And as we lean on algorithms, is there a risk of them introducing unintentional biases? If an AI model is predominantly trained on data from one demographic, could it misdiagnose others, perpetuating healthcare inequalities?
Impact on physician-patient relationship and patient autonomy and control
As digital biomarkers become more prevalent, will they diminish the invaluable face-to-face interactions between physicians and patients? How will this reliance on remote monitoring affect the quality of care and patient satisfaction? And in this digital age, do patients truly have control over their personal health data, especially when shared with third parties? The balance between technological innovation and preserving the sanctity of the physician-patient relationship becomes a pressing concern.
The widespread adoption of digital biomarkers in clinical research and healthcare has highlighted the need for a comprehensive framework to address associated concerns effectively. Though currently theoretical the ‘BioGuard Framework’ could provide a strategy to address these issues and tackle the multifaceted challenges posed by digital biomarkers.
Robust data security protocols
The BioGuard Framework recommends the adoption of the Advanced Encryption Standard (AES-256), currently the gold standard in data encryption techniques. Coupled with Transport Layer Security (TLS) for data in transit and Zero-Knowledge Proofs8 for data at rest, these protocols ensure that patient data remains inaccessible to unauthorised entities. Regular penetration testing, as per the OWASP Top Ten vulnerabilities, will further fortify the system against potential breaches.
Comprehensive regulatory and compliance roadmap
The dynamic and evolving landscape of digital biomarkers calls for adaptive and forward-thinking regulatory guidelines. The BioGuard Framework proposes the establishment of the Digital Biomarker Regulatory Council (DBRC). This council, acting as a collaborative consortium, would involve key stakeholders from global health organisations, tech innovators and policy-makers. Working in tandem with established bodies like the FDA and EMA, the DBRC would develop and regularly update standardised protocols. These guidelines would be akin to the Clinical Laboratory Improvement Amendments (CLIA)9 but specifically tailored for digital biomarkers, ensuring they remain clinically relevant, ethically sound, and adhere to stringent international standards.
Democratising access and inclusive R&D
The BioGuard Framework could create a future where the transformative potential of digital biomarkers is universally accessible, transcending technological, linguistic, and cultural barriers. Prioritising compatibility, it will champion tools that seamlessly integrate with both iOS and Android, the dominant global smartphone operating systems. But true accessibility goes beyond technology. The framework will emphasise the creation of tools that resonate globally, offering multilingual support and cultural adaptability. Collaborations with organisations like GSMA, known for bridging mobile technology gaps, will ensure that even underserved regions benefit from these advancements. Furthermore, recognising the spectrum of health literacy worldwide, the framework will ensure tools are intuitive for both health professionals and the general populace.
Bias minimisation and ethical AI deployment
AI holds immense transformative potential in healthcare, but its effectiveness and fairness are deeply tied to the data it’s trained on. To ensure that AI doesn’t inadvertently perpetuate or introduce biases, the framework will champion the use of Fairness Indicators. These are specialised tools meticulously designed to assess, identify and rectify any fairness discrepancies in AI models. By actively training AI systems on comprehensive datasets, such as the Global Health Data Exchange, the framework will ensure that the insights and predictions generated are representative of varied demographics. This holistic approach, when combined with continuous learning algorithms and a strong commitment to ethical AI principles, guarantees that AI models not only evolve with the times but also consistently deliver unbiased and equitable health outcomes for all demographics.
Reinforcing the sanctity of physician-patient interactions
In the heart of healthcare lies the deep bond and trust between a patient and their physician. With this in mind, the BioGuard Framework will promote a harmonious blend of technology and traditional medical consultations. By integrating cloud-based Electronic Health Records (EHR) platforms such as Epic Systems or Cerner with digital biomarker tools, the framework can provide physicians with a comprehensive view of patient data. This amalgamation of traditional medical records and digital insights ensures that care is both data-driven and human-centric. The framework will further promote the use of telemedicine platforms like Teladoc or Amwell for routine virtual check-ins, while emphasising the importance of reserving in-person consultations for more in-depth discussions, preserving the invaluable essence of the physician-patient bond.
Empowering patient autonomy in the digital age
In today’s digital age, the power to control one’s personal health information is not just a right but a deeply personal journey of trust and understanding. Recognising this, the BioGuard Framework will be deeply rooted in patient-centric principles. It will support transparent data policies, ensuring every individual has clear and unambiguous control over their health data. Through the use of Personal Health Information Exchange (PHIE) systems, patients are empowered to decide who can access their information, fostering a sense of ownership and trust. To further guide this journey, the framework will promote educational initiatives. By leveraging online educational platforms, patients will be able to access courses on digital health literacy, ensuring they are not only in control but also fully informed and confident in their choices in the digital healthcare landscape.
Navigating the digital biomarker landscape
The integration of digital biomarkers into clinical trials represents a significant leap in healthcare innovation. The BioGuard Framework could offer a comprehensive roadmap for this integration, ensuring that technological advancements align with ethical considerations, inclusivity and the irreplaceable human touch. From championing robust data security to emphasising the sanctity of the physician-patient relationship, the framework underscores the importance of a balanced approach, where technology enhances, rather than replaces, the core values of healthcare. As we stand at this intersection of technology and healthcare, a pressing question remains unanswered: In our pursuit of digital excellence, how do we ensure that the core tenets of trust, empathy and human connection remain at the forefront of medical evolution?
About the author
Deepika Khedekar is a Clinical Trial Lead at IQVIA, where she spearheads clinical trial monitoring programmes for major pharmaceutical companies. In her 12-plus years in the industry, she has led clinical trial programmes for companies such as Gilead Sciences and Macleods Pharma. She holds a Master’s degree in Pharmacy from the University of Mumbai.
- Mapping the 100 trillion cells that make up your body, University of Florida, Oct 2018
- What is a gene? Medline Plus, National Library of Medicine
- Total number of registered clinical studies worldwide since 2000, Statista
- Sun D, Gao W, Hu H, Zhou S. Why 90% of clinical drug development fails and how to improve it? Acta Pharm Sin B 2022;12(7):3049-3062.
- Through Apple Heart Study, Stanford Medicine researchers show wearable technology can help detect atrial fibrillation, Stanford Medicine, Nov 2019
- AI can spot early signs of Alzheimer’s in speech patterns, study shows, UTSouthwestern Medical Center, April 2023
- Digital Biomarker Predicts Risk of Death in Interstitial Lung Diseases, RT Magazine, May 2023
- What are Zero-Knowledge Proofs, Forbes, Feb 2023
- Clinical Laboratory Improvement Amendments (CLIA), Center for Disease Control & Prevention