18 Conference Report #techonomy2018

Daniel Kraft on Exponential Medicine for Global Health


  • Daniel Kraft at Techonomy2018, Sunday, November 11, 2018. (Paul Sakuma Photography)


Daniel Kraft
Faculty Chair, Medicine & Neuroscience, Singularity University

Description: Here is technologized healthcare going? Can the U.S. tackle its healthcare mess? And what will exponential medicine mean around the world?

The following transcript has been lightly edited and condensed for ease of reading.


Speaker: Daniel Kraft, Singularity University

Introduction: David Kirkpatrick, Techonomy

(Transcription by RA Fisher Ink)

Kirkpatrick: Daniel Kraft is the head of all the medical stuff at Singularity University and he’s got a great talk about exponential medicine. So Daniel, take it away.

Kraft: Thank you, David. So we live in a pretty extraordinary time for health and medicine, particularly the opportunity to globalize health care around the planet. And as the theme is here at Techonomy, it’s how do we take these technologies and converge them to address a lot of the grand challenges we have across health care?

And, of course, many technologies are riding Moore’s Law, the power of computing getting faster and cheaper. This is my iPhone 2 from 10 years, it felt amazing 10 years ago, now it feels slow and clunky and has a low-resolution camera. And it’s these exponentials which have shifted what we can do and what we can put in our pocket. There’s also transforming many elements of health and medicine as we get mobile and connected and data-driven. And it’s these exponentials which often can be a bit surprising about their speed and pace and their potentials impact many fields, particularly the realm of medicine.

And when we think about Moore’s Law in medicine I think we should think about this convergence of other fields from AI and robotics and synthetic biology and low-cost genomics that are coming together faster, cheaper, and better and giving us the opportunity to address some of the grand challenges that we have across the health care continuum from hopefully optimizing health, wellness, prevention, keeping us healthy before we get sick, early diagnostics, picking up disease at stage zero rather than at an advanced stage, more personalized, precise therapy that’s easier to take and less toxic, how we can democratize and digitize health care around the planet and finally, how can we all play a role in discovery, about participate and crowdsource the future of medicine.

And all of those can ride exponential rails as we address challenges in health care which include rising costs, aging populations, access to care, many parts of the U.S. and the rest of the world have a shortage of doctors and nurses and specialists. So no matter what happens to Obamacare or Trumpcare, which is really no care, or Putincare, we need to think of new ways to bring health care to new folks. We’ve lots of exponential data coming at us but how do we make that useful information at the site of point of care for the clinician and patient? How do we un-fragment it, how do we help our regulators and our payers get on top of these exponential rails? So lots of challenges, including the fact that even where countries have health care, 5 million die because of poor quality health care in low and middle income countries, more than the 3.5 million who die from not having access to care. So lots of challenges and lots of opportunities.

Part of that is shifting our mindset. We’re still stuck across the globe, really in a sick-care model. And what do I mean by sick-care? We have very intermittent data that we can collect from ourselves or patients, an occasional blood pressure check, EKG, if you have diabetes or hypertension, maybe you’re faxing or PDFing the data to your doctor, whether they want to see it or not. So with intermittent and episodic data, we have a sick-care system or a reactive, we wait for the patient to show up with a heart attack or a stroke or in my field of oncology, to present with late stage disease as opposed to early. And I think the meta-potential of all these new technologies is to move us to much more continuous and proactive and personalized health and medicine in the developed and the developing world.

Now, part of the challenge is getting us out of our old buckets and mindsets whether it’s how we define medical specialties by anatomy or how we un-silo all the siloed data which is often stuck between EMRs and health care systems and pharma companies and to connect the dots and take that data and turn it into knowledge which is often doubling every 73 days. It’s hard to keep up with.

So as we look at these massive new sets of data and hopefully new knowledge, how do we translate the clinical utility for each of us as individuals, as patients, as caregivers, and as health care systems? Well, I get an interesting seat at the table, I’ve been the chair of medicine since we started Singularity University, 10 years ago. And there we have an interesting lens to look at this cross-convergence and bring new people and players together. Seven years ago, I started a program called Exponential Medicine where—and we just had our eighth program last week at the Hotel Del Coronado with the world-wide spin and several faculty who are here today. And it’s a pretty incredible thing when you bring folks from all around the world, patients, nurses, providers, payers, investors to look at what’s here today and where are things heading. It’s a touch and taste the future today to catalyze into the future.

And so with that mindset we like to look at challenge areas and think about new ways to address them. And one area that I’ve always looked at is how do we improve how we take our medicines. First of all, we should obviously optimize prevention so we never need it but those who do take medications it’s often hard to take them. We have 4.5 billion drugs prescribed, prescriptions, every year in the United States. This year alone, that’s 15 for every man, woman, and child. And for most of us, our experience with taking our medicines is too many pills, hard to read instructions, one-size fits all dosing that we’re not even usually taking as indicated. And so that leads to several major challenges. Even when we do take our medicines, they often don’t work. For the top 10 grossing medications prescribed in the United States this year alone, they are only effective for about 1 in 5 to 1 in 24 of those of us who are taking them. And so it’s great if you’re number one but what about everybody else? And on top of that, even if your medication is working, there’s often side effects or when it doesn’t work, Aspirin is only effective in about two-thirds of individuals but everybody who takes it has the risk of GI bleeds and other elements that kill thousands every year. And it’s adverse drug reactions like these that by some indications are the number four leading cause of death around the world. Part of these adverse drug reactions is based on challenges with dosing. So I trained at Stanford and then went to Mass General to train in internal medicine and pediatrics so I’ve spent a lot of time with little people in pediatrics where in the NICU, for example, in the neonatal high intensive care unit we’d be dosing to a fraction of a mg for every kg of weight. And the next night I might have been doing a shift in the adult emergency room and taking care of a frail nursing home patient or maybe in the next 10 minutes a 200-300 pound football lineman. By most cases, we’d be dosing the elderly patient and the large football player with the same doses, sort of one-size fits all dosing. And I think we can do better in our now connected digital age to improve dosing and selecting the right drugs for the individual based on age, sex, and racial elements.

Another major challenge is folks don’t take their medicines as indicated, particularly because we’re often a poly-pharmacy. About 40% of Americans over 65 are taking 5 or more pills, often many 15 or more pills, it’s hard to keep track of them. It’s often cut them in half to get the dosing appropriate. And it’s the low adherence or low compliance rate that is extremely expensive. If we even improved adherence on statins by 10-20% we’d save millions and billions of dollars and save thousands and thousands of lives. So there’s lots of improvement. And we think we’re in this amazing, connected AI-driven internet of things age but we’re really not in an era of precision medicine yet but an era of imprecision medicine.

So that got me thinking, what if there might be a better way? How do we take these technologies to go beyond the pill cutter and the fax machine to get better adherence and more precision and personalization for health care and therapy around the planet? Well, part of this starts with a lot of the new, amazing exponential technologies that we have on our wrist. We’re only nine years into the Fitbit world and we’re like five different versions of trackable devices today, that means we’re off in the world of quantified self where you can own your own data on your Smartphone but now we’re starting to shift to quantified health where that information can flow to health care provider, into your health care system, and be used in smart ways to optimize and measure and optimize your wellness but also do early diagnosis and then to manage therapy.

So let’s take a common example of a disease, hypertension, high blood pressure, it’s the number one leading cause of early death and morbidity around the planet. About half of US Americans, adults, have hypertension, less than half have it well-controlled. And while part of the challenge is you have to take two or three medications and you get spot checks but now you can buy a clinical blood pressure cuff. Some are being shrunk into your watch. Some are going to dissolve into seamless radar-based ones that can screen your blood pressure continuously, some of the startups are working on this. So if I had hypertension, I could be streaming my live data and my doctor or my cardiologist or an algorithm could use that data to optimize my medications and optimize my flow.

Now, other elements include our mobile watches today, our Smartwatches, or Apple watch, even before Apple could do it can measure an EKG, the company that makes this device is doing over 1 million EKGs a month and crowdsourcing that data. We can collect information on EKG patches that are basically an intensive care unit on a patch that can transmit my vital signs anywhere and soon my patients or your data anywhere, not just in real time but retrospectively. So you can manage patients in an asynchronized and virtualized manner.

And there’s many other examples of what we can do today from shockables, it can change your behavior, hearables that play music but also track your steps and heart rate, ringables that can track your sleep, something so important to health and medicine. There’s a lot to do in exponential shrinking all the way to the point where we can not just check our weight but our shape and in this case scan me and determine not just my kgs but my muscle content and fat mass and use that to help hopefully optimize a diet or a therapy.

And for those of us who don’t want to wear anything, MIT engineers have modified wi-fi to pick up the vital signs of up to 10 people in the same room at the same time. So we have tremendous new ways to collect this data, this digital exhaust, sometimes called the digitone. And we can even go beyond that, picking up our labs, a lab on a chip can be done in a rural village on our Smartphone, the results delivered by the internet or by drone. And we’re in an era of low cost genomics, we’re at $1,000, soon to be a $100-dollar genome where I can now obtain my pharmacogenomics, how do my genes impact what drug dose I might want for a blood pressure med to a statin. Do I need normal dose, high dose, low dose or a different medication all together? Or, using crowdsource for genomic information, to understand that diabetes is at least three distinct subtypes. Or, do fun things like share my genome print socks that are really made based on my genome, that could be engaging right? But it enables us all to sort of donate our data. You’ve heard of the Framingham trial, now there’s a Framingham trial to an all of us trial being initiated by NIH this year, you can all donate your genomics and other information.

So now, hopefully we can start to crowdsource, be smarter and use AI or I like to call it IA, intelligence augmentation to pick the right prevention or therapy drugs that might be best for you. But, still, our cutting-edge way of tracking our medications still looks like this, right? We still have some smarter ways, connected pill bottles, some will squawk at you apps, so smart packaging like PillPack that was just bought by Amazon for $1 billion dollars. But we’re still taking a pile of pills.

So this got me thinking, what if we could make better use of the polypill? Polypills come in cold and flu medications, lots of medications combined or a medication which combine several medications. And these have been done in a global health study to look at prevention. Folks with a high-risk of heart disease or stroke who had not yet had an event took polypills with a half dose, ACE inhibitors, statins, aspirin, other blood pressure meds and dramatically reduced their risk for heart attack or stroke. So I thought what if we could personalize these to be very individual for preventional therapy, what if you could essentially print your own polypill, it would be built for you, based on you and could adapt to you even day to day based on your digital exhaust and other data.

So we’ve got personalization, right? You can preprint everything from your braces to orthopedic devices to hearing aids off your cell phone. What if you could then print essentially your own personalized polypill, I call it the intelli-meds, this idea that instead of taking a pile of six generic pills, you could combine these into a single medication. And this is sort of how it works. Instead of, again, multiple tablets we have what we call micro-meds, each one or two mgs that you could sort out of cartridges, like ink, but made out of generic drugs that could sort and match the doses that you need. And of course, we start with generic drugs, you could eventually influence the life, long or short acting based on codings on each micro-med. And then you could have a printer in your corner pharmacy or world village that would print your medications based on you and what you needed. And these would also already be meds that you are allowed to take together and well-proven and start with generics which are 80% of the medications done today.

So once you print your intelli-med you could do some fun elements like your print your name on it, a QR code, and essentially here is our prototype printer which we just finished, the software enables you to dial in 16 different drugs, the printer essentially works like this, there’s silos with different intelli-meds, we adjust the dose with a robotic arm and as you adjust the dose it releases and they fall down the silo and essentially print your personalized intelli-med right there at the point of care or eventually on your bathroom counter.

So that’s the big picture idea. And the idea here is that you could print a pill potentially every morning and based on your digital exhaust, if you needed more blood thinner, less Lasix, different doses, you’d be able to improve your adherence and ease of taking medications. Now, all of this would potentially have global impact, again, for prevention and adherence. As more and more folks are connected to the internet we can provide data as information. Community health workers can use a whole new set of digital tools to collect information on their patients. Digital doctors or community health workers bag, all these things are becoming digitized and democratized for doing diagnostics and for doing therapy through virtualized care, through tele-health, the virtualized visit with an individual clinician but increasingly driven by AI and chatbots, again, democratizing access across the planet. And now chatbots and platforms like Good Doctor in China, over 200 million individuals had almost no access to care, now have that opportunity.

So as we pull all these new forms together, I think we can all start to accelerate discovery, you can download clinical trials. Of course, we’ve now changed our driving habits in the last 12 years, we used to use paper maps 12 years ago, now we couldn’t imagine driving out without our Google maps and our Waze. What if we had a Waze for health care where we were crowdsourcing our health journeys, it was intelligent, it would guide us to our health destination. It could course correct in real time, it would be gamified, and our information could be shared globally. We’d give information and we’d get information back.

So I think part of this future of health care is thinking of ourselves not just as organ donors and blood donors but as data donors as we accelerate our learning. So with this mindset, whether it’s taking a medicine, prevention, global health, I hope we can shift from our episodic, reactive, imprecise, siloed world of sick care and move to one of continuous, proactive, personalized, and participative health care that globalizes around the planet. Thanks, very much.


Leave a Reply

Your email address will not be published. Required fields are marked *