The Patient Equation

by Glen de Vries

  9. Jeff Margolis, “How Consumer Data (Not More Clinical Data) Will Fix Healthcare,” MedCity News, April 9, 2018, https://medcitynews.com/2018/04/consumer-data-not-clinical-data-will-fix-healthcare/.

  10. Glen Tullman, interview for The Patient Equation, interview by Glen de Vries and Jeremy Blachman, September 3, 2019.

  Conclusion

  If there's one lesson to take from this book, it's this: our health follows paths, paths that are the summation of everything from our DNA to the ways we interact with the world and the effects of the environment on us. Like a trail through the woods, we can observe where we came from. But imagine those paths meandering—not through a forest, but through a multidimensional space encompassing all of the things we might, now or in the future, have the ability to measure: our genes (and which of those genes are on or off), our blood chemistry, blood pressure, the function of our organs, the way we think, the manifestation of those thoughts in our behavior, and so much more.

  A map of a forest is two‐dimensional. A rocket in space traces a path through three dimensions. To envision the paths in the n‐dimensional space of every aspect of our biology, at every scale, is a staggeringly overwhelming problem. But the paths exist. And we have new tools by which we can trace them and predict where they will go.

  We can supplement all of our traditional medical knowledge with the digital trails we leave behind us every day. We can use the phenomenal connectivity and computational power available today to figure out what dimensions of information are relevant and predictive of our health futures. We can simplify the problem of trying to measure and comprehend it all. We can combine the known paths for patients, both in and outside of research programs, into a map—a multidimensional map, to be sure, but one with borders, with lines that delineate when and how health can be maintained, and diseases managed or cured.

  These paths, these maps, are patient equations. The digital ways we create, harvest, and combine data at scales never seen before in the history of medicine will help us discover them, and will help us reverse‐engineer them. Perhaps not perfectly right away, and not in a way that will automate every decision we make for our care—but certainly in a way that is additive to our quality of life, and to our longevity.

  All of this presents incredible opportunities for improving our health as individuals, and for unleashing a new era in the practice and the business of life sciences and medicine.

  Professor Michael Snyder, director of Stanford's Center for Genomics and Personalized Medicine, has been called “the world's most bio‐tracked man.”1 He has spent the past few years wearing devices that measure everything about him and his environment. In the course of doing so, he diagnosed his own Lyme disease from an algorithm that saw an infection hiding within his data.2 He believes that combining genetic information with our exposome—the chemicals and organisms we encounter every day in our environment—have so much more to tell us than we've let them. “We now think we can tell when you get sick before you realize it,” he says.3

  Snyder and others, according to the New York Times, envision doctors doing more than running lab tests and checking vital signs. “They will scrutinize your genome for risk factors and track tens of thousands of molecules active in your body. By doing so, the doctors of the future will identify diseases, and treat them, long before symptoms appear.”4 It's those layers in our layer cake, all making themselves known. Snyder ran a study that found that 53 of 109 subjects discovered something meaningful about their health—undiagnosed diabetes, heart disease, and more—from being tracked.

  “We were able to go back and see molecules that were clearly starting to rise months before the diagnosis and then dropped with treatment,” the lead author of the study, Sophia Miryam Schüssler‐Fiorenza Rose, told the Times.5 “We think these might be very valuable early markers of disease,” she said. One person in the study “learned they had early‐stage lymphoma before they showed symptoms…[from] a test of the person's immunome, which measures levels of immune chemicals in the blood.”6

  Cue Health is developing a “miniature medical lab for the home,”7 which would collect saliva, blood, or a nasal swab, test for illness, and make actionable recommendations. Alphabet and Apple are developing new health capabilities in their watches with every release. There are, by one report, 17 diseases with biomarkers to be found in human breath.8 Thirty‐eight of the companies listed in the Fortune 50 have some kind of digital health care offering.9 The question is: can yours beat them?

  There are pitfalls as we move forward, of course. There is a naiveté right now in the business world that big data and machine learning produce magic,10 but it's not magic at all. These are merely new technologies that happen to be so much more advanced than the ways we used to think about these problems. They can seem like magic, sometimes. But digital technology, data science, and artificial intelligence are not immune to the real‐world requirements of proving therapies to be safe, effective, and reliable.

  Whether the therapy in question is a molecule or medical device, or if it is digital, taking advantage of a patient equation‐based view of the world allows us to prove that safety, efficacy, and value in more precise ways than ever before.

  Companies that already play in the life sciences space are the most well‐equipped to lead this revolution. They have the expertise in evidence generation, with techniques that aren't necessarily new. What's new is the relevant scale of evidence generation, from populations down to individuals down to cells—which means that new entrants to the health care ecosystem may be required to bring the massive connectivity and processing power needed to the patients, providers, regulators, and payers involved. But, make no mistake, the life sciences industry has to lead the way.

  Privacy and Transparency

  I've spent almost no space in this book talking about an issue that trips a lot of us up when we attempt to go to market with data‐driven products and services in medicine: privacy. And while it's true that there may be a learning curve for patients (and doctors) to trust algorithms with their health data, at the end of the day, I've largely ignored the issue because I think it's one that's worth putting to the side.

  Not because it isn't incredibly important. There are laws and ethical considerations that should be and must be adhered to and thoroughly considered. But if the end product of creating a patient equation‐powered future will benefit people and society as much as I believe it will, mechanisms can and will inevitably be created to protect our data and the information generated by it in order to realize the full societal benefit.

  Notably, issues of consent must be respected. The standard ethics of clinical research applies, and we shouldn't be leveraging data about someone without their permission—the same way we can't experiment on someone unless they agree to be experimented on. But the technology, to be quite honest, cannot be contained. And the benefits to individuals will doubtlessly motivate consent, as well as incentivize governments, regulators, and corporations to come up with solutions to privacy issues that are reliable, transparent, and auditable.

  The reality is that while many of us may not feel comfortable signing a piece of paper saying that our insurance company can see our data—we don't want them to have the power to raise our premiums because they find out we eat horribly and could be blamed for, say, our recent cardiac incident—we probably post enough on our social media pages for them to figure it out anyway. The data we give off each day, those digital trails, are too numerous to control, and attempts to wall them off are destined to fail, just like the containment of the dinosaurs in Jurassic Park. Unless someone is willing to go completely off the grid, not use a cell phone, a computer, a credit card, an electronic bus ticket, or ride through a city armed with cameras on stoplights, their data is out there and will find ways to escape.

  The even larger truth is that we shouldn't care, because in a patient equation‐powered world, the more data that is out there, the more we are helped. The more doctors and researchers le
arn about us, the better they can find the right treatments at the right time for whatever condition we are faced with. In the movie Gattaca, people are placed in a caste in life based on their DNA—so there's an incentive to pass off someone else's genetic code as your own. But if this book has shown anything, it's that genes are only a small piece of the puzzle. Genotype is dwarfed by phenotype. Predictions from genetics alone are just a tiny, tiny piece of the overall picture.

  Incentives really are ultimately aligned—government and industry do better when citizens and customers are healthy. Like it or not, even your employer and insurer are on the side of your health. You are a more profitable subscriber and a higher‐revenue‐generating employee if you are well. It's not just that I believe everyone in every business has the right intentions—although in life sciences and health care, I mostly do (there are easier industries to act badly in). But, if you'll allow me to leave questions of overpopulation and the environment in the same pile as privacy for the moment—in the pile of things that are incredibly important, but not relevant to realizing the value of patient equations—it's economically sensible for them to keep everyone alive and healthy for as long as possible. And that's especially true in a world of value‐based care, where the physician, the physical therapist, the pharma company, and everyone in the health care value chain is compensated based on positive outcomes and effective prevention.

  So What's Next?

  Regardless of your particular role in health care, perhaps entirely confined to wanting your own health to be cared for, I hope this book has been a call to action. The power of predicting the health of an individual through generating consistent maps of population health is spectacular. Plotting the equations in the multidimensional space discussed will require industrial investment, academic contributions, and the consent of individuals. Everyone has an important part to play—contributing data, funding investments, and producing and practicing medicine, all centered around patient equations.

  This is not meant to be a definitive work on the subject. The interviews, anecdotes, and ideas presented are like the digital trail of one person's health—interesting, hopefully useful, but limited to who I am and what I've experienced. My hope is that the extraordinary seat that I've had for over 25 years—at the center of the digital revolution in clinical research and its connected ecosystem of health care—and the insights of those interviewed for the book provide perspective that can be turned into action.

  Today, we are curing diseases recently thought incurable, and turning previously fatal diseases into manageable, chronic conditions. We are targeting molecules once thought impossible to interact with through medicine, creating medical devices that seem imported directly from 1960s science fiction movies like Fantastic Voyage, and accessing computational power and the kinds of data that would be unimaginable when I was doing my first laboratory research a quarter of a century ago.

  Of course, an easy argument can be made that health care and life sciences aren't “well.” Providing everyone with access to the highest‐quality care is a problem we have barely chipped away at in most countries, let alone on a global scale. The costs for developing new drugs and medical devices, and the incentive systems for effectively deploying them to patients, are headwinds—strong ones—blocking forward motion in medical innovation. How much more could we be advancing the state of the art in health care tools if we were able to turn them into tailwinds, powering us toward progress, supported by the digital power and biological innovations sitting in data centers and laboratories around the world?

  We must work together to make that turnaround happen. Hopefully this book helps us all think about new ways to harness these revolutionary techniques and technologies along a path to some of the lofty future possibilities described in these pages.

  This book comes at an interesting time in my professional life. Shortly before completing the manuscript, Medidata was purchased by Dassault Systèmes, a company whose software has literally transformed the way planes, automobiles, and probably many—if not most—of the products and services that are involved in the “stuff” around you are designed and manufactured (just take a look around whatever room you are in, and think about all of the design and manufacturing that has gone into creating it).

  As a product life cycle management platform, Dassault Systèmes—just like Medidata aimed to do before the acquisition—wanted to bring the incredible power of their platform to the world of designing and delivering drugs and medical devices. Now, post‐acquisition, Medidata and my own personal ambitions in that regard haven't changed. However, our plans to bring to life a platform that spans dimensions from molecular to individual to population‐level modeling and research for life sciences certainly got a lot more credible. And now, with a larger sandbox and more computational toys around me, if I'm to take my own call to action and act on it, I'm realizing clearly what prompted Dassault Systèmes to start poking around Medidata in the first place: that in addition to thinking about platforms, data, and the artificial intelligence on top of it, simulations will play an incredible part in the future of research, medicine, and patient equations.

  Bernard Charlès, Dassault Systèmes' CEO and vice chairman, speaks about the importance of the “virtual twin,” not just in health care, but based on his experience with simulated versions of what is being created in reality across other industries. In life sciences, we're used to a tiny fraction of drugs—as few as 1 in 10—that enter clinical trials making it all the way through the process to become drugs on the market. Imagine that ratio when it comes to passenger aircraft (the same order of magnitude, and actually more expensive to create than a drug development program). Would the industry function if they only flew 10% of the time once manufactured? Of course not. They fly 100% of the time.

  Is the analogy a perfect one? Far from it. But are there lessons to be learned and ways to think about how industries like aerospace can develop incredibly complex products at a staggeringly more reliable rate of success than we do in life sciences? I think that answer is an obvious yes. So as we—the combined Medidata and Dassault Systèmes—pursue some of the ideas in this book, the continuation of the past 20 years of work at Medidata‐standalone, you can expect the appearance of virtual twins at every scale. Drugs. Cells. Organs. Patients.

  I am convinced that they will be another dimension by which we can multiply the evidence we are able to factor out of every individual unit of patient data gathered. My patient equation work will continue—hand in hand with patient simulations—and I hope and expect it will intersect with everyone's efforts as discussed in, and happening around, this book.

  Notes

  1. Dana G. Smith, “Meet the World's Most Bio‐Tracked Man,” Medium (OneZero), May 8, 2019, https://onezero.medium.com/meet-the-worlds-most-bio-tracked-man-2077758cf5a2.

  2. Veronique Greenwood, “The Next Big Thing in Health Is Your Exposome,” Medium (Elemental), November 5, 2018, https://medium.com/s/thenewnew/the-exposome-is-the-new-frontier-e5bb8b1360da.

  3. Ibid.

  4. Carl Zimmer, “In This Doctor's Office, a Physical Exam Like No Other,” New York Times, May 8, 2019, https://www.nytimes.com/2019/05/08/science/precision-medicine-overtreatment.html.

  5. Ibid.

  6. Dana G. Smith, “Meet the World's Most Bio‐Tracked Man.”.

  7. “Cue Is a Miniature Medical Lab for the Home,” Cloud9Smart, May 20, 2014, https://www.cloud9smart.com/cue_health_tracker.

  8. Amanda Hoh, “How a Breath Test Could Reveal What Disease You Have,” ABC News, July 31, 2017, http://www.abc.net.au/news/2017-07-31/detecting-disease-in-breath-with-world-smallest-breathalyser/8759050?pfmredir=sm.

  9. Don Jones, “Conference Talk at Medidata NEXT Event” (October 2016).

  10. Wikipedia Contributors, “Clarke's Three Laws,” Wikipedia, February 1, 2019, https://en.wikipedia.org/wiki/Clarke%27s_three_laws.

  Acknowledgments

  Writing this book has been an incredible process, providing an exc
use to reach out to people across health care—in some cases people I've had the pleasure of working with, and in other cases people whose work I've admired from afar. Huge thanks to everyone who gave their time to talk through the issues and ideas in this book, and allowed me to share their perspectives: Dr. Don Berry, Anthony Costello, Kara Dennis, Dr. David Fajgenbaum, Robin Farmanfarmaian, Dr. Graham Hatfull, Jamie Heywood, Dr. Julian Jenkins, Dr. Stan Kachnowski, Pascal Koenig, Dr. Jerry Lee, Dr. Veena Misra, T. J. Sharpe, Alicia Staley, Glen Tullman, and Dr. Daniel Yadegar.

  I will never be able to repay my literary debt to Jeremy Blachman, my extraordinary co‐author. Without his perception, partnership, and persistence this book would not exist.

  I want to thank my Medidata colleagues, past, present, and future as part of Dassault Systèmes. Some of them appear in the book by name, but I am equally thankful to all of them. The work we've done together is quite literally this book's foundation. A particular thanks to Nicole Pariser, who was instrumental in turning the idea into a plan—and in giving it a title! Special thanks as well to the tireless Dana Suchow, who made sure all of the conversations, approvals, and endless logistics necessary for the book came together, and to Jenni Li, who turned slides, sketches, and pictures of whiteboards into the graphics in the book. And of course to my friends and co‐founders Tarek Sherif and Dr. Edward Ikeguchi. This book scratches at the surface of a journey we began over 20 years ago, an incredible experience, the enormity of which I find it hard to even put into words.

  To the faculty and my friends at Carnegie Mellon University, New York University, and Columbia University: the privilege you have extended to me, allowing me to teach and to collaborate far beyond my academic credentials, is one of my greatest pleasures, and has forged much of the content in this book.