Speech by Robert M. Califf, M.D. to the National Health Council's 2023 Science for Patient Engagement Symposium -- Patient Empowerment in the Digital Health Era
May 8, 2023
- Speech by
Robert M. Califf, M.D., MACC
(Remarks as prepared for delivery)
I’m delighted to be with you as part of this important meeting to discuss patient engagement and digital health. Both are priorities for the FDA, and both have also been areas of medicine and healthcare to which I have devoted much of my own work. In pairing the two topics in this symposium, you underscore not only their individual importance but also the potential for synergy as the digital world and understanding of how to engage people converges in many ways. At the same time this same convergence raises serious questions about our approach to clinical research, health and healthcare.
Your agenda includes an exciting set of “red hot” topics in this area, so what I’d like to do is make some general opening comments and then take questions and have a discussion to help kick off what should be a very stimulating symposium.
I’ll start with the concept of patient engagement. We are lucky to be living and working in a time of unprecedented scientific, medical, and technological breakthroughs. We have a deepening understanding of the underlying mechanisms of disease and human biology, of genetics and genomics, and of the enormous opportunities afforded through the development and application of new technologies, including the expanded ability to digitally collect, review and share data. These technological advances allow us to envision, explore, and realize unprecedented new approaches for preventing and treating a wider range of previously untreatable diseases.
In combination with these more biologically and mechanistically oriented scientific advances, we also are breaking new ground in understanding human behavior, social interactions and how they are inextricably intertwined with health, disease and human outcomes. As in many fields experiencing a rapid growth in technology and knowledge, uncomfortable questions are raised that will test our mettle as we delve into an exciting area with so much potential and also with so much risk.
One of the most important aspects of our work at the FDA is to incorporate the perspectives and experiences of patients to make sure the evidence reflects what matters most to, and is most effective for, patients and their caregivers.
It seems obvious to say that if we want essential insights about what it is like to live with a disease, the outcomes that matter, and the adequacy of treatment options, we should ask people who are living with the problem or at risk of experiencing the problem—both patients and caregivers. As we focus in on this issue, especially with improved digital technology, we should be able to design a more accurate, more complete benefit-risk framework that helps us to better evaluate the safety and effectiveness of a medical product across the vast expanse of patients and consumers with different values, educational levels and living and healthcare environments. We need to assure that the experience of living with the disease is factored explicitly into development programs for all medical products, including measures of benefit and harms within clinical trials. We are also entering a new era of ability to assess and improve the fit between the product and people who could benefit from its use—otherwise known as human factors.
We’re getting better at this as we gain experience with explicit integration of the patient voice into our work. For example, our Patient-Focused Drug Development (PFDD) initiative engages patients and caregivers in a facilitated dialogue with FDA on a variety of individual diseases to develop rigorous and objective methods for measuring patients’ experiences and perspectives in a reliable and representative way. This can include the development of methods for identification of key impacts and elements of disease experience that matter most to patients, and the translation of those elements into measurement tools that are clear, register meaningful change, and prove valid and reliable in capturing patients’ reported experience in clinical trials and clinical outcome studies.
Our Patient Representative Program brings the patient voice to discussions about new and already approved drugs and devices, as well as to those that are currently being considered for approval. It also allows for input earlier in the regulatory medical product development and review process. Patient representatives can serve on FDA Advisory Committees, as consultants for the review divisions in the determination of whether a medical product's benefits outweigh its potential risks, and as presenters at FDA meetings and workshops on disease-specific or regulatory and health policy issues. These and many other models of patient engagement are being used across the FDA and are making a real difference.
But this is where the rapid advance of digital technology is likely to disrupt our current methods. We have enormous potential as the world enters the 4th industrial revolution in which rich and diverse sources of digital data are available at scale in real time with potentially unlimited storage capacity, and these data are becoming widely available as part of the clinical care system. Large language models are the next step that appears to be ushering in the revolution that many of us were hoping for.
Throughout my career, I’ve focused on strengthening systems for generating and gathering new and better data and analyzing those data to provide reliable evidence to inform and improve the many decisions that we make as consumers, patients, families, clinicians, and regulators. This kind of information supports the nuts and bolts of the FDA’s work.
The ultimate goal of gathering all of this data and evidence is to better inform our knowledge about the benefits and risks of health interventions and to assure that they can be used to improve health relative to the risk of harm. Of course, our focus at FDA is on medical products and nutritional interventions, but as we expand our capabilities with digital technologies, the bright line between the individual product and the environment, both human and technological, only makes sense in the early phase of development and assessment. Ultimately, the benefits and risks of the use of a product depend on whether and how it is used, and the context in which it is used.
Digital health technologies or DHTs provide opportunities to foster more efficient conduct of clinical investigations by making the operating characteristics of specific digital measurements in their context of use. For instance, they can be used to remotely record and analyze data directly from participants throughout their day. And the use of DHTs in a clinical investigation can help improve patient access to, and participation in, clinical investigations by potentially reducing the burden of required visits to a research site.
As many of you know, there are multiple projects underway to generate greater volumes of high quality, relevant digital health data, including the facilitation of efficient, streamlined randomized controlled trials and registries by extending the uses of existing digital health data. For example, one focus of our Orphan Products Grants is to support the development of virtual or digital health platforms and models that can integrate data and serve as management or monitoring tools and inform good clinical practice.
I should also point out that while the impact of these data is significant for the development of medical products and treatments, it has important applications for, and supports our regulatory decisions in food safety, tobacco products, and all of the products we regulate.
Let me identify some of the large questions I hope you will shed light on today:
Disparities: We are experiencing a dramatic decline in life expectancy, only partially due to Covid, but also including gun violence, suicide and opioid overdose. We’re talking less about the ongoing rise in chronic diseases and their sequalae like stroke and heart attack. This alarming situation is not uniform in the U.S.—differences as a function of race, ethnicity, sex, education, wealth, and rural living status are stark.
I have been unable to get a quote from Ed Yong in the Atlantic out of my head: “Technological solutions drift into society’s penthouses. Diseases seep into its cracks.” I urge you to think beyond the hypothetical or artificial laboratory-based assessment of how new drugs, devices and foods will be used across our society.
Representativeness: It’s obvious that we now have the ability to interact with almost the entire spectrum of people with a disease or at risk. While I believe that patient focus groups, advocates and representatives are phenomenal, we will eventually need to address the diversity of values, preferences, beliefs, living circumstances and economic factors by directly measuring them in broader populations. As evidenced by the ubiquity of cell phones and companies with billions of users, our challenges and limitations are no longer technological—they are embedded in our human interactions. CDRH has done some fabulous work showing that the balance of benefits and risks can look quite different for a product depending upon preferences that vary across a population.
Algorithmic bias: Algorithms, including the new large language models, evolve after they are put into practice. It is unlikely that an accurate algorithm will stay accurate when deployed in real life over time. It needs adjustment and measurement of its operating characteristics continuously throughout the life cycle. How do we get that done?
Large language models: If we’re not nimble in the use and regulation of large language models, we’ll be swept up quickly by something that we hardly understand. The great things are really great: imagine a world in which your questions were answered immediately in language appropriate for your literacy and numeracy; also your clinician can actually talk with you rather than spending all their time cutting, pasting and writing clinic notes; I could go on and on, but I see the regulation of large language models as critical to our future.
Misinformation: Many Americans die or experience serious illness every year due to bad choices driven be false or misleading information. This trend has accelerated as social networks have driven communication based on common social identities that are ripe for exploitation. This new digital era can connect us, but technologies like large language models give almost everyone the potential to produce false narratives or even so-called deep fakes—fabricated images and voices. A key part of a successful transition to digital health is an effective regulatory scheme to guide digital technologies to improved human outcomes and interaction.
Across industry, digitization and insertion of machine learning and other types of mathematical algorithms into everyday life is making a profound difference, but government agencies lag behind private industry. Quite simply, we need to assemble the resources to put in place these policies and tools and adaptively align our digital health efforts to support public health and regulatory innovation in a world that is changing rapidly.
The potential certainly exists, and we have good reason to hope that we will fulfill the enormous potential of digital health technologies and see more innovative solutions that can result in improved outcomes, but we’d better also preempt the risks.
Thank you, and I look forward to your questions.