May 22, 2024
Why is the confluence of healthcare and life sciences
happening? What are the two biggest mistakes of technology in
healthcare? And how can research insights be embedded into every
care decision? We will find out all that and more with our guest
Dr. David Feinberg, a medical professional and healthcare industry
executive and current Chairman of Oracle Health.
--------------------------------------------------------
Episode Transcript:
00;00;00;02 - 00;00;27;22
What makes multidisciplinary collaboration the key to health care innovation? What is the effect of bundled, integrated solutions on the patient experience and how can we invest in what matters most while streamlining the entire process? We'll find all that out and more on this episode of Research in Action. Hello and welcome to Research in Action, brought to you by Oracle Life Sciences.
00;00;27;22 - 00;00;52;08
I'm Mike Stiles. And today our very special guest is Frank Bateman, a digital health data and technology executive. He's currently a senior advisor to Oakland's De Silva and Phillips and was a former chief information officer of the U.S. Department of Health and Human Services. Oracle Life Sciences has an e-book coming on the next phase of growth for the Life Sciences industry, and Frank was a really valuable resource for that.
00;00;52;08 - 00;01;22;00
He's got a lot of great thoughts on how pharma and biotech are investing in tech to support things like personalized medicine, improved clinical trials and drug safety tracking. That's why we wanted to get him on the podcast. So Frank, thanks so much for joining us. Thanks. It's great to be here, Mike. We appreciate it. Well, we got a lot of ground to cover, but I know you went into corporate strategy in the beginning of your career and through the bulk of your career, but obviously somewhere down the line you started crossing paths with government.
00;01;22;00 - 00;01;42;04
So what did that involve? How did that happen? Well, I've been lucky enough to pursue my interests wherever they took me. I hadn't expected to pursue a career in the life sciences and health care when I started out focused on nuclear arms control. But my interest in technology actually came about from my work on verification measures for a nuclear test ban.
00;01;42;21 - 00;02;09;05
Technology first took me to IBM Research and then under IBM corporate strategy, as you mentioned, when in in corporate, I oversaw the company's ten year outlook. And as a tech company, we saw high performance computing in the life sciences staring us in the face. We needed to be in it. And our chairman at the time, Lou Gerstner, accepted a recommendation that we invest 100 million to launch a business unit focused on the life sciences.
00;02;09;19 - 00;02;36;24
So I love the idea. You were actually serving in the Obama administration. White House Entrepreneur in residence. I love the idea of an entrepreneur in residence because one doesn't quickly equate government with speed, original ideas and innovation. Were you impressed by or frustrated by the speed at which you could bring things to full fruition in government? Impressed? Absolutely frustrated.
00;02;37;00 - 00;03;04;25
Yeah. Our times sometimes there are arcane processes that get in the way of novel solutions, but I always thought that had great admiration for the dedicated dedication the mission demonstrated by civil servants. Doing things differently was really a hallmark of the Obama administration. It wasn't just the Entrepreneur in Residence program you mentioned. Obama appointed the nation's first chief technology officer, the first chief information officer.
00;03;05;06 - 00;03;31;08
He launched the US Digital Service to provide agencies with a different approach to software development. He created challenge that guards as a means for agencies to seek innovations by awarding modest prizes as opposed to large government contracts. It brought new voices to light. I look at our current government a lot, like most governments, it's inherited its structure from the industrial age.
00;03;31;18 - 00;03;58;12
For the most part, it's organized by industry, by vertical. There's an Agriculture Department, energy, health, defense and so on. The congressional appropriations process is what exacerbates the problem in this information age. I really believe that Multi-disc culinary collaboration is what brings about solutions. And I don't have a background in biochemistry, but I worked with biochemists to explore therapies that made effective use in both of our disciplines.
00;03;58;25 - 00;04;23;21
If you think of Tesla for a moment, the company has innovations, it has inventions. But its real success was that of an integrator. It brought together knowhow from battery management, aerodynamics, automobile engineering, software development and legacy. Automakers had been working on these problems in building an EV for years, but their approach failed to deliver a car with mass market appeal.
00;04;24;00 - 00;04;47;06
And I think that's precisely what we need to do in the life sciences now, is bring the disciplines together and organize to solve problems. Now, I think the listeners are starting to see why you're such a fascinating person to have on the show. You've been exposed at high levels to nearly every component of health care, and through most of that you were tasked with being really a futurist and a trend spotter in it.
00;04;47;06 - 00;05;08;17
So just keep my head straight. I'm going to cover things with you in buckets now. The first being what the challenges and opportunities really are in life sciences. Fun fact for our listeners can bring up at their next dinner party. When things get dull, it takes about $2 billion and 10 to 15 years to get a drug to market.
00;05;08;17 - 00;05;30;27
Now, for most people who have gotten used to rapid advancement, getting things they want and need on demand, that sounds absolutely crazy. So can technology kind of change this equation soon? Mike I don't think that's crazy at all, and I really believe that we're on the cusp of change. One of the startups that I worked with, Empower Medicine, is a really great example.
00;05;31;11 - 00;06;04;00
What they're trying to achieve is a complex endeavor. It depends upon bringing together people from different disciplines to work across the universe of stakeholders. And going back to the Tesla example, GM and Ford built highly structured teams in engineering designed propulsion. But Tesla was a software company from the start. So I think the challenge is how do you, as a life sciences company, mimic what Tesla did to bring together the disciplines and focus on the entire process of drug development?
00;06;04;14 - 00;06;33;17
It's almost like if technology isn't the answer, what is? For instance, it's the only way really to capture the volume and sources of adverse events, right? We always look at adverse events and drug discovery thanks to that observation. Technology can do wonders, but it isn't nirvana. I it does great things, but I think it's always important to remember in health care there needs to be a human touch because health care at its core is about people.
00;06;33;28 - 00;07;02;27
Technology is already making waves in clinical trials and there's so much more to come. We're on the early stages witnessing that impact. Things like electronic patient reported outcomes and various sensors are beginning to gather data from patients during trials and during real world use. And this technology facilitates the capture of adverse events actively and passively, leading to just a wealth of data and deeper understanding of therapeutic effects.
00;07;03;19 - 00;07;31;23
This could uncover unexpected drug interactions or shed light and personalize or genomic attributes. Sometimes, though, adverse events are not obvious. And that's that's really another role that technology can play because of its ability to capture so much data, it may find unexpected things to match what's going on in the market. Actually, Oracle just merged its health care and Life sciences organization late last year.
00;07;31;23 - 00;07;55;24
Why do you think those two things are coming together? I know you talk about bringing things together and that's just like one example of it. Yeah, I think that's a really great example. I like to think of health as being all encompassing. The life sciences exist to support health. The same could be said for payors, providers, physicians, health systems, pharmacies, patients, Cros, even employers.
00;07;56;09 - 00;08;24;11
Each has their role to play. The vast majority of companies across the health sector have a mission or model that says something like Patients are the reason we're in business. Well, I'm not questioning it. In fact, I'm pretty confident people are involved, they're sincere. But if serving patients is your mission, I'd ask, when was the last time you took a look at your organization to see if it is optimally designed to address the needs of patients in this information age?
00;08;24;28 - 00;08;54;23
We know that siloed organizations underperform multiple disciplines and experiences are not considered. Information isn't shared in much. The way I spoke about HHS is being a reflection of the health sector by having a research component, by having a regulatory component, by having a provider component. I think that those companies that integrate health disciplines need to step out of their comfort zone in the same way that Oracle combined those pieces.
00;08;55;07 - 00;09;24;18
Now put I want to put that futurist hat on and tell us which innovations you think are going to have the most profound impact. On average, Mike's like me and say the next decade, What do you see coming? So I think it's important to have a framework to think about this. And and I've begun to craft a mind map to identify emerging use cases for AI because it's their adoption that makes real change possible downstream.
00;09;25;01 - 00;09;52;06
The framework that I propose is first, think about what are the emerging use cases where good enough, where is today? Suffices seconds Think about the next hurdle that generative AI crosses. What does that hurdle enable? And third, when you look at the first use cases of health, what are the second order needs that become possible? Things that haven't been able to be addressed.
00;09;52;20 - 00;10;19;05
The good enough example concept deserves an example. There's a startup by the name of Hai Labs that makes use of artificial intelligence, and for disclosure, I'm on the company's board. Hi Labs motto is We clean dirty data to unlock its potential for health care. Heaven knows if you've been around health care, you know about Dirty data. Hai Labs has mastered the capability that it is good enough for health plans.
00;10;19;05 - 00;10;49;18
Who can address incomplete claims, claims data, flawed provider directories, even incomplete clinical data plans. Love the product because it solves the problem they have today. Tomorrow, it might be good enough for clinical studies. It isn't today. And that's the framework I think we ought to be exploring when we think about what is generative. AI's impact on health care, what's possible today, what's good enough, and what's that going to train the large language models to do tomorrow.
00;10;50;05 - 00;11;24;20
Another example I find rather inspiring is a nonprofit by the name of Every Cure, launched by David Feigenbaum. Based on his own experience as a med student, he was diagnosed with Castleman Disease, a cell disorder of the lymph nodes and he nearly died after discovering that a 25 year old drug would block Castleman his pathway. He started every cure which is making use of AI to sort through well-documented commercial therapeutics to discover what might be repurposed.
00;11;25;02 - 00;11;47;27
You just don't know where AI is going to take you. And I think you need to look at the indicators in the marketplace to say, Oh, that's happening now. What possibilities does that create for the future? So the next bucket is personalized medicine. We've also become a culture that's really used to getting catered to from grocery stores, knowing what we usually buy to Netflix, knowing what movies will probably like.
00;11;47;27 - 00;12;12;26
We really gotten used to that. Health conditions are seen by patients as a very personal thing. So what are the remaining roadblocks that we're hitting and delivering? Truly personalized and customized medicine? So I have every confidence in personalized medicine. I have worked around it for years now, and there are things to know about individuals that are cheap and easy to collect.
00;12;12;26 - 00;12;41;08
But there are also things that are really difficult and costly to capture. And for each category, I think we need to be asking ourselves the question, What can I do with this knowledge? If I know something about this individual, can I do something? And personalization powered by digitization. I think a good example for patients with type two diabetes, It's moved quite swiftly because that knowledge is easily captured and it can be turned into coaching and medicines.
00;12;41;19 - 00;13;16;16
But there are many other diseases where personalized option doesn't yet offer a therapeutic advantage. How do you protect health information while also making it widely available and shareable to everyone who needs it? Isn't that another barrier? It is. Ultimately, I think patients need to be in control of their own health records. It's the only viable solution if patients are always wondering whether their data is under someone else's control or someone else is profiting from it or using it in ways they don't agree with, then they're not going to share their data.
00;13;17;01 - 00;13;39;15
So we need to find a mechanism to empower patients to control their data, their health data granularly. We've talked a lot on this show about real world data and real world evidence. Should we be am I overhyping what our would and RW we can lead to? Well, I think electronic health records are full of errors. We all know that.
00;13;39;24 - 00;14;07;29
But the question we should be asking is what's good enough and for what purpose? As more medical doctors are born, digital people coming out of med school in their twenties now have only done medical digital like the tech industry, collaborates on standards and competes on performance. Real world data will get better and generative A.I. will have an effect as well.
00;14;08;11 - 00;14;35;23
So I think we need to look at again, it's an evolution. What's good enough and understand that we're heading in that direction because all of our stakeholders are increasingly doing their their jobs only digitally. So the next bucket would be clinical trials. What can we do from a data collection angle to make clinical research move better, more efficient and faster to work better for the patient?
00;14;36;07 - 00;15;09;00
I was with a startup by the name of Empower Medicine and Mark Lee, the CEO of Empower, has a set of PowerPoint slides that I think do a great job of illustrating. The problem is analog to clinical trial data is a greenhouse. It's purpose built for one study. It's costly and the investment cannot be repurposed. When the study is completed, the well-manicured greenhouse is the most that isn't economically sustainable, nor does it capture evidence that might inform science.
00;15;09;16 - 00;15;36;28
So I'm on a separate note. I think we're missing an opportunity to capture data from populations that are representative of the disease being researched. It's obviously a bit more effort and takes some creative thought. So while there's pressure to enroll patients in studies, the lack of diversity impairs our understanding of the disease. And to your earlier question, it slows down the adoption of personalized medicine.
00;15;37;14 - 00;16;09;00
You know, in all honesty, none of my guests have ever exactly rave about the state of electronic health records. How do you think those issues have to get solved in order to improve clinical trials? Well, Mike, I'm not raving, but ours have come a long way over the past 15 years. Your question is interesting, though, because it focuses on clinical trials and for the most part, providers at the point of care are not focused on clinical trials.
00;16;09;16 - 00;16;44;03
That's pharma's interest. Our challenge ought to be to make electronic health records better for everyone. If we take seriously the opportunity to reimagine clinical trials, why should the data from point of care be separate from the trial data? You could argue it's a historic anomaly akin to our discussion of siloed verticals. I'm not saying there should not be a separate clinical trial system that might manage the trial or produce analytics about the trial, but the data about patients should be captured in the EMR and not through a redundant data entry.
00;16;44;03 - 00;17;04;22
Let me give you an example. I used to forget my wallet or my keys every time I left the house. Now my phone has all of those responsibilities and more. It's become more valuable and I rarely forget it. So I guess the question I have is how do we make our more valuable to all stakeholders? And I think that's something Oracle is really leaning into.
00;17;04;22 - 00;17;37;10
With that acquisition of Cerner. It finds itself with the largest components of that equation, so it can then proceed with solutions that do connect clinical trials to points of care. Do you think an undertaking like that is just an example of common sense? I do, and I suspect that many tech vendors are racing to make this happen. It'll be a while before the evidence is sufficient to enroll patients, but generative AI is ready, suggesting patients for studies based upon our data.
00;17;37;19 - 00;18;05;23
So in some sense, where it's good enough for some purposes now and we can only imagine what it might be around the corner, you know, I think of about how clinical trials could be fundamentally changed. I think about reduction of chaos really by using standards and automation. That's accepted pretty much throughout the industry, which means more digitalization. Am I an idiot thinking that's possible?
00;18;06;23 - 00;18;34;27
I'm not going to say that, Mike, thanks. But I do think your question is a certainty and I'm betting on it. Meaningful digitalization requires a rethink. However, of what we're trying to achieve and what the necessary steps are along the way. So doing unneeded steps faster won't have much of an effect. Amazon didn't just give you a shopping cart for your goods.
00;18;35;12 - 00;19;02;18
They changed the shopping experience by providing suggestions for accessories, storing your payment information, delivery preferences, and giving you reviews of those products. We need to be thoughtful about how do we change the process rather than speeding up the unnecessary stage gates along the way. It's all about simplification with a focus on the patient. I don't mean that as a platitude.
00;19;02;18 - 00;19;27;13
Every drug company, as I said, talks about its work in terms of the patient, but it's about understanding the patient's preferences and prioritizing them. I love that. Well, when you said, you know, doing unnecessary things, unnecessary steps faster doesn't get us anywhere, that's very smart. You touched on it, but AI and drug development specifically is kind of its own bucket.
00;19;27;13 - 00;20;04;07
How is pharmaceutical research and development about to be transformed because of a I mean, what roles does it play in getting these drugs to market faster so they can help people sooner? So the mind map that I mentioned I think is informing second order outcomes. And using this framework, I've begun to focus on a few areas. First is clinical research asking the question how does clinical research change when generative AI solutions become good enough to enable patients to provide raw, real world data from digital health devices?
00;20;04;18 - 00;20;32;02
Will that make it easier to recruit patients? And then there's another question what responsibilities the sponsors have when those devices deliver worrying evidence. The second area that I've been thinking about, the second order outcomes is the patient experience. It's never fun to be a patient, but in the current environment you need to be a bookkeeper, an administrator, a note taker, a risk manager, a data interpreter and an advocate.
00;20;32;12 - 00;20;58;27
There are impressive A.I. solutions to each of these challenges that I've seen in development now. So the question we ought to pose is what happens to the patient experience when these solutions are bundled and integrated with one another? And does that amount to a virtual concierge? Since it weaves data across providers, labs, pharmacies, payers and tech stacks, the patient wins.
00;20;59;10 - 00;21;24;22
But I've come to wonder which health sector is when and which lose. Are there any ethics or security concerns that's unique to applying AI to health care? Certainly we've heard the criticisms about, you know, well, AI scrapes the web and turns out not everything on the Internet is true. So, you know, is there any kind of danger of bad data being pulled in and applied by A.I.?
00;21;25;08 - 00;21;44;26
There are tons of concerns and there are think tanks out there publishing reports on these. But the truth is, the genie can't be put back into the bottle. A number of companies have put forward thoughtful ethics guidelines, particularly from the tech sector. But we can't allow the rules to vary from company to company, and we can't depend upon self-policing.
00;21;44;26 - 00;22;09;22
The stakes are just too high. Instead, we need Congress to act in established guardrails that allow the AI industry to grow without causing harm to individuals. Congress largely ignored privacy over the past couple of decades, while the rest of the world moved ahead on that front. We shouldn't allow this to happen again because A.I. arguably poses a much greater risk.
00;22;10;07 - 00;22;35;03
When states are forced to act, we end up with a patchwork of rules that are easy to circumvent. Yeah, you brought up a really good point that, you know, while our focus is on medicine and pharma and clinical research and patients, government and business does enter the picture, how are the pharma companies responding to things like the U.S. Inflation Reduction Act and the price pressures that they're facing?
00;22;35;15 - 00;23;03;05
Well, I can't speak for the pharma companies. I do observe their attempt to prevent it from going into effect, the price pressures, the controls. But I think ultimately we need to get to a point where there is meaningful digitization to allow a rethink of what we're trying to achieve so we can streamline processes. You mentioned about how other countries jumped on the regulation of AI so much sooner than we did.
00;23;03;26 - 00;23;34;11
What are the drug costs and medical procedure cost disparities between the United States and seemingly the rest of the world? I mean, it seems like our costs are always so infinitely higher. They are. And as an American, I've got to say, I can't explain it and I am frustrated by it. And I'm frustrated when seniors or people who don't have resources can't get the medicines that they need because they're being gouged.
00;23;35;08 - 00;24;01;10
Pharmaceutical companies who are charging two and a half to three times what they charge in Western developed nations in Europe. I really do think there needs to be a rethink of the way pharma does its business to streamline it and take unnecessary steps out of the process that could reduce the costs of drug development. Yeah, and a lot of that cost in our system isn't even directly healing patients.
00;24;01;10 - 00;24;30;20
It's administrative costs. It's inefficiencies in everything from staffing to supplies and other verticals and other businesses. Those are areas where tech is really being aggressively applied to get to those efficiencies. And you're saying maybe health care is playing catch up? I think it is. You know, there are two sectors that are laggards in adopting technology globally and it isn't just in the U.S. it's government and it's health care.
00;24;31;02 - 00;24;56;19
Health care has gotten on the bandwagon, particularly in certain sectors like pharma. Every sector in health care needs to do this, though, because the economics of health care are not sustainable, as in other industries. Health care writ large needs to ask what's best for the patient and determine what's the most efficient way of getting there. Delivering that those who employ the greatest creativity will serve both patients and shareholders interests.
00;24;57;02 - 00;25;25;11
So, you know, as I think about what a pharmaceutical company looks like today, or I think about what a payer looks like today, I think the question I have is, is there something outside of your sector that you could do that would deliver value to patients and better outcomes? If there is, why are you doing it? Are you happy with the degree to which research data is being shared?
00;25;25;20 - 00;25;54;15
Currently? Let me suggest that we ask the question just a little differently. Could we improve the sharing of research data? And without a doubt, the answer is yes. What if we think out of the box here and we empower patients, as I said earlier, to make the decision, perhaps all informed consent going forward could include a question where the patient consents to release anonymized data not only for the sponsor, but for all of science, for all researchers.
00;25;54;28 - 00;26;21;16
Putting on my privacy hat, I think it's fair to say that we all expect to have control over our personal health records, and we need to empower patients to make these decisions. And I suspect there are enough examples of this now. I suspect that when patients are asked, will you make your personal information, your health records available to science for future generations, the answer is almost always going to be yes.
00;26;21;28 - 00;26;47;16
Yeah, I agree with you. Turns out not everyone's a nice guy like Frank here. Cybercrime is real. Health care organizations, particularly have been in the news lately for all the wrong reasons. Oracle's Larry Ellison and Seema Verma just wrote about it and the Wall Street Journal. Is that a winnable fight? It feels like we're getting to a place where everyone's just accepting that there is no security and we're just going to have to live with it.
00;26;48;03 - 00;27;09;19
I think it comes down to how you define winnable. I hate to tease that out, but there will be cyber attacks and there will be breaches. You can't stop them entirely, but you can sure cut down on your risk profile. Companies who are diligent can dramatically reduce the risk of appearing on the front page of the Wall Street Journal as opposed to the Opinion Page.
00;27;10;00 - 00;27;50;28
There's no silver bullet, though, and it's unlikely that proprietary technologies can beat attackers, especially when nation states are involved in the attack. When I was in government, I got a close up look at the industry, the health care industry and cybersecurity. We were in the early days of creating industry specific communities. In particular, we launched the health ISAC, which means information sharing and Analysis Center in 2010, and it immediately provided a view into breaches, a view that enabled others across the health sector to shut down the vulnerabilities that were successfully used to attack someone else.
00;27;51;20 - 00;28;15;20
In many instances, it wasn't the technology that failed us. It was social engineering that led to the breach. So expound on that. The difference between, well, obviously technology can do what it can do and that it has its shortcomings. But what do you mean? It was social engineering that failed us. Usually attackers will find a vulnerability. It could be a helpdesk.
00;28;15;29 - 00;28;44;15
It could be someone in an accounting office that has access to the system. They'll call and they'll sound serious. They may even have gotten some personal information from someone else to pretend that they're that person and doing that, they will change a password. They will gain access to a system. So it isn't the technology that failed. It's that there were other access points to the technology that someone socially engineered.
00;28;44;26 - 00;29;04;27
So humans are fool able. Oh yeah, we are not you and I, of course. But you know, other humans are. I hold in my hand the last bucket, which is if I were in charge of everything. If you were in charge of everything in the many components of health care, they would listen to you and follow your recommendations.
00;29;04;27 - 00;29;30;29
What would those recommendations be? As we sit here today, in 2024, I can dream, can't I? Make sure you can. I'm putting I'm making you head of HHS now. I guess my suggestion is what I call threading the needle. By that I mean laying out a business process that begins with life sciences research and ends with providing life saving therapies to patients.
00;29;31;14 - 00;29;59;13
And then ask yourself, how can we invest in what matters while streamlining the entire process? Because there are just too many stakeholders, too many people taking a profit, too many unnecessary steps in a process that, as I said, was designed during the industrial age and isn't needed anymore. Technology can play a crucial role, but so too will company culture, expertise and perhaps most importantly, stakeholder engagement.
00;29;59;29 - 00;30;32;18
Everyone has to be on board for changes, these kind of structural changes to succeed. Does this mean bringing back some aspects of clinical research into pharma away from crows? I don't know. Maybe. Does it involve making use of hours for real world data? I think certainly perhaps it involves personalized medicine and genomic testing would make it unaffordable. But in a world of value based care, is there a way to use the outcomes to pay for the entire therapy?
00;30;32;28 - 00;31;11;22
I think it's quite likely that generative AI is going to change the health sector, making it more efficient, less bureaucratic, better integrated around delivering value. So I think those companies that don't act could very well find themselves with a consequential decision down the road. However, companies that pursue a strategy that really rethinks with the patients in the center and delivering therapies and the science behind doing so, I think will see their benefits to not only their bottom line, but they'll provide the best care that they say they want to provide by focusing on the patient.
00;31;12;17 - 00;31;33;20
It's really great advice that should probably be heeded. Frank It's been great. Again, thanks so much for being with us. I'm sure our listeners may want to follow you or find out more. What's the best way for them to do that? Well, I'm currently on a social media hiatus, and for you, I do avoid it. But certainly anyone can follow me or connect to me on LinkedIn.
00;31;33;28 - 00;32;05;25
Okay, great. And for our listeners, if you want this level of smart all the time, go ahead and subscribe to the show right now. And if you want to learn more about how Oracle can accelerate your own life sciences research, just go to Oracle dot com slash life dash sciences and we'll see you next time.
What makes multidisciplinary collaboration the key to health care innovation? What is the effect of bundled, integrated solutions on the patient experience and how can we invest in what matters most while streamlining the entire process? We'll find all that out and more on this episode of Research in Action. Hello and welcome to Research in Action, brought to you by Oracle Life Sciences.
00;00;27;22 - 00;00;52;08
I'm Mike Stiles. And today our very special guest is Frank Bateman, a digital health data and technology executive. He's currently a senior advisor to Oakland's De Silva and Phillips and was a former chief information officer of the U.S. Department of Health and Human Services. Oracle Life Sciences has an e-book coming on the next phase of growth for the Life Sciences industry, and Frank was a really valuable resource for that.
00;00;52;08 - 00;01;22;00
He's got a lot of great thoughts on how pharma and biotech are investing in tech to support things like personalized medicine, improved clinical trials and drug safety tracking. That's why we wanted to get him on the podcast. So Frank, thanks so much for joining us. Thanks. It's great to be here, Mike. We appreciate it. Well, we got a lot of ground to cover, but I know you went into corporate strategy in the beginning of your career and through the bulk of your career, but obviously somewhere down the line you started crossing paths with government.
00;01;22;00 - 00;01;42;04
So what did that involve? How did that happen? Well, I've been lucky enough to pursue my interests wherever they took me. I hadn't expected to pursue a career in the life sciences and health care when I started out focused on nuclear arms control. But my interest in technology actually came about from my work on verification measures for a nuclear test ban.
00;01;42;21 - 00;02;09;05
Technology first took me to IBM Research and then under IBM corporate strategy, as you mentioned, when in in corporate, I oversaw the company's ten year outlook. And as a tech company, we saw high performance computing in the life sciences staring us in the face. We needed to be in it. And our chairman at the time, Lou Gerstner, accepted a recommendation that we invest 100 million to launch a business unit focused on the life sciences.
00;02;09;19 - 00;02;36;24
So I love the idea. You were actually serving in the Obama administration. White House Entrepreneur in residence. I love the idea of an entrepreneur in residence because one doesn't quickly equate government with speed, original ideas and innovation. Were you impressed by or frustrated by the speed at which you could bring things to full fruition in government? Impressed? Absolutely frustrated.
00;02;37;00 - 00;03;04;25
Yeah. Our times sometimes there are arcane processes that get in the way of novel solutions, but I always thought that had great admiration for the dedicated dedication the mission demonstrated by civil servants. Doing things differently was really a hallmark of the Obama administration. It wasn't just the Entrepreneur in Residence program you mentioned. Obama appointed the nation's first chief technology officer, the first chief information officer.
00;03;05;06 - 00;03;31;08
He launched the US Digital Service to provide agencies with a different approach to software development. He created challenge that guards as a means for agencies to seek innovations by awarding modest prizes as opposed to large government contracts. It brought new voices to light. I look at our current government a lot, like most governments, it's inherited its structure from the industrial age.
00;03;31;18 - 00;03;58;12
For the most part, it's organized by industry, by vertical. There's an Agriculture Department, energy, health, defense and so on. The congressional appropriations process is what exacerbates the problem in this information age. I really believe that Multi-disc culinary collaboration is what brings about solutions. And I don't have a background in biochemistry, but I worked with biochemists to explore therapies that made effective use in both of our disciplines.
00;03;58;25 - 00;04;23;21
If you think of Tesla for a moment, the company has innovations, it has inventions. But its real success was that of an integrator. It brought together knowhow from battery management, aerodynamics, automobile engineering, software development and legacy. Automakers had been working on these problems in building an EV for years, but their approach failed to deliver a car with mass market appeal.
00;04;24;00 - 00;04;47;06
And I think that's precisely what we need to do in the life sciences now, is bring the disciplines together and organize to solve problems. Now, I think the listeners are starting to see why you're such a fascinating person to have on the show. You've been exposed at high levels to nearly every component of health care, and through most of that you were tasked with being really a futurist and a trend spotter in it.
00;04;47;06 - 00;05;08;17
So just keep my head straight. I'm going to cover things with you in buckets now. The first being what the challenges and opportunities really are in life sciences. Fun fact for our listeners can bring up at their next dinner party. When things get dull, it takes about $2 billion and 10 to 15 years to get a drug to market.
00;05;08;17 - 00;05;30;27
Now, for most people who have gotten used to rapid advancement, getting things they want and need on demand, that sounds absolutely crazy. So can technology kind of change this equation soon? Mike I don't think that's crazy at all, and I really believe that we're on the cusp of change. One of the startups that I worked with, Empower Medicine, is a really great example.
00;05;31;11 - 00;06;04;00
What they're trying to achieve is a complex endeavor. It depends upon bringing together people from different disciplines to work across the universe of stakeholders. And going back to the Tesla example, GM and Ford built highly structured teams in engineering designed propulsion. But Tesla was a software company from the start. So I think the challenge is how do you, as a life sciences company, mimic what Tesla did to bring together the disciplines and focus on the entire process of drug development?
00;06;04;14 - 00;06;33;17
It's almost like if technology isn't the answer, what is? For instance, it's the only way really to capture the volume and sources of adverse events, right? We always look at adverse events and drug discovery thanks to that observation. Technology can do wonders, but it isn't nirvana. I it does great things, but I think it's always important to remember in health care there needs to be a human touch because health care at its core is about people.
00;06;33;28 - 00;07;02;27
Technology is already making waves in clinical trials and there's so much more to come. We're on the early stages witnessing that impact. Things like electronic patient reported outcomes and various sensors are beginning to gather data from patients during trials and during real world use. And this technology facilitates the capture of adverse events actively and passively, leading to just a wealth of data and deeper understanding of therapeutic effects.
00;07;03;19 - 00;07;31;23
This could uncover unexpected drug interactions or shed light and personalize or genomic attributes. Sometimes, though, adverse events are not obvious. And that's that's really another role that technology can play because of its ability to capture so much data, it may find unexpected things to match what's going on in the market. Actually, Oracle just merged its health care and Life sciences organization late last year.
00;07;31;23 - 00;07;55;24
Why do you think those two things are coming together? I know you talk about bringing things together and that's just like one example of it. Yeah, I think that's a really great example. I like to think of health as being all encompassing. The life sciences exist to support health. The same could be said for payors, providers, physicians, health systems, pharmacies, patients, Cros, even employers.
00;07;56;09 - 00;08;24;11
Each has their role to play. The vast majority of companies across the health sector have a mission or model that says something like Patients are the reason we're in business. Well, I'm not questioning it. In fact, I'm pretty confident people are involved, they're sincere. But if serving patients is your mission, I'd ask, when was the last time you took a look at your organization to see if it is optimally designed to address the needs of patients in this information age?
00;08;24;28 - 00;08;54;23
We know that siloed organizations underperform multiple disciplines and experiences are not considered. Information isn't shared in much. The way I spoke about HHS is being a reflection of the health sector by having a research component, by having a regulatory component, by having a provider component. I think that those companies that integrate health disciplines need to step out of their comfort zone in the same way that Oracle combined those pieces.
00;08;55;07 - 00;09;24;18
Now put I want to put that futurist hat on and tell us which innovations you think are going to have the most profound impact. On average, Mike's like me and say the next decade, What do you see coming? So I think it's important to have a framework to think about this. And and I've begun to craft a mind map to identify emerging use cases for AI because it's their adoption that makes real change possible downstream.
00;09;25;01 - 00;09;52;06
The framework that I propose is first, think about what are the emerging use cases where good enough, where is today? Suffices seconds Think about the next hurdle that generative AI crosses. What does that hurdle enable? And third, when you look at the first use cases of health, what are the second order needs that become possible? Things that haven't been able to be addressed.
00;09;52;20 - 00;10;19;05
The good enough example concept deserves an example. There's a startup by the name of Hai Labs that makes use of artificial intelligence, and for disclosure, I'm on the company's board. Hi Labs motto is We clean dirty data to unlock its potential for health care. Heaven knows if you've been around health care, you know about Dirty data. Hai Labs has mastered the capability that it is good enough for health plans.
00;10;19;05 - 00;10;49;18
Who can address incomplete claims, claims data, flawed provider directories, even incomplete clinical data plans. Love the product because it solves the problem they have today. Tomorrow, it might be good enough for clinical studies. It isn't today. And that's the framework I think we ought to be exploring when we think about what is generative. AI's impact on health care, what's possible today, what's good enough, and what's that going to train the large language models to do tomorrow.
00;10;50;05 - 00;11;24;20
Another example I find rather inspiring is a nonprofit by the name of Every Cure, launched by David Feigenbaum. Based on his own experience as a med student, he was diagnosed with Castleman Disease, a cell disorder of the lymph nodes and he nearly died after discovering that a 25 year old drug would block Castleman his pathway. He started every cure which is making use of AI to sort through well-documented commercial therapeutics to discover what might be repurposed.
00;11;25;02 - 00;11;47;27
You just don't know where AI is going to take you. And I think you need to look at the indicators in the marketplace to say, Oh, that's happening now. What possibilities does that create for the future? So the next bucket is personalized medicine. We've also become a culture that's really used to getting catered to from grocery stores, knowing what we usually buy to Netflix, knowing what movies will probably like.
00;11;47;27 - 00;12;12;26
We really gotten used to that. Health conditions are seen by patients as a very personal thing. So what are the remaining roadblocks that we're hitting and delivering? Truly personalized and customized medicine? So I have every confidence in personalized medicine. I have worked around it for years now, and there are things to know about individuals that are cheap and easy to collect.
00;12;12;26 - 00;12;41;08
But there are also things that are really difficult and costly to capture. And for each category, I think we need to be asking ourselves the question, What can I do with this knowledge? If I know something about this individual, can I do something? And personalization powered by digitization. I think a good example for patients with type two diabetes, It's moved quite swiftly because that knowledge is easily captured and it can be turned into coaching and medicines.
00;12;41;19 - 00;13;16;16
But there are many other diseases where personalized option doesn't yet offer a therapeutic advantage. How do you protect health information while also making it widely available and shareable to everyone who needs it? Isn't that another barrier? It is. Ultimately, I think patients need to be in control of their own health records. It's the only viable solution if patients are always wondering whether their data is under someone else's control or someone else is profiting from it or using it in ways they don't agree with, then they're not going to share their data.
00;13;17;01 - 00;13;39;15
So we need to find a mechanism to empower patients to control their data, their health data granularly. We've talked a lot on this show about real world data and real world evidence. Should we be am I overhyping what our would and RW we can lead to? Well, I think electronic health records are full of errors. We all know that.
00;13;39;24 - 00;14;07;29
But the question we should be asking is what's good enough and for what purpose? As more medical doctors are born, digital people coming out of med school in their twenties now have only done medical digital like the tech industry, collaborates on standards and competes on performance. Real world data will get better and generative A.I. will have an effect as well.
00;14;08;11 - 00;14;35;23
So I think we need to look at again, it's an evolution. What's good enough and understand that we're heading in that direction because all of our stakeholders are increasingly doing their their jobs only digitally. So the next bucket would be clinical trials. What can we do from a data collection angle to make clinical research move better, more efficient and faster to work better for the patient?
00;14;36;07 - 00;15;09;00
I was with a startup by the name of Empower Medicine and Mark Lee, the CEO of Empower, has a set of PowerPoint slides that I think do a great job of illustrating. The problem is analog to clinical trial data is a greenhouse. It's purpose built for one study. It's costly and the investment cannot be repurposed. When the study is completed, the well-manicured greenhouse is the most that isn't economically sustainable, nor does it capture evidence that might inform science.
00;15;09;16 - 00;15;36;28
So I'm on a separate note. I think we're missing an opportunity to capture data from populations that are representative of the disease being researched. It's obviously a bit more effort and takes some creative thought. So while there's pressure to enroll patients in studies, the lack of diversity impairs our understanding of the disease. And to your earlier question, it slows down the adoption of personalized medicine.
00;15;37;14 - 00;16;09;00
You know, in all honesty, none of my guests have ever exactly rave about the state of electronic health records. How do you think those issues have to get solved in order to improve clinical trials? Well, Mike, I'm not raving, but ours have come a long way over the past 15 years. Your question is interesting, though, because it focuses on clinical trials and for the most part, providers at the point of care are not focused on clinical trials.
00;16;09;16 - 00;16;44;03
That's pharma's interest. Our challenge ought to be to make electronic health records better for everyone. If we take seriously the opportunity to reimagine clinical trials, why should the data from point of care be separate from the trial data? You could argue it's a historic anomaly akin to our discussion of siloed verticals. I'm not saying there should not be a separate clinical trial system that might manage the trial or produce analytics about the trial, but the data about patients should be captured in the EMR and not through a redundant data entry.
00;16;44;03 - 00;17;04;22
Let me give you an example. I used to forget my wallet or my keys every time I left the house. Now my phone has all of those responsibilities and more. It's become more valuable and I rarely forget it. So I guess the question I have is how do we make our more valuable to all stakeholders? And I think that's something Oracle is really leaning into.
00;17;04;22 - 00;17;37;10
With that acquisition of Cerner. It finds itself with the largest components of that equation, so it can then proceed with solutions that do connect clinical trials to points of care. Do you think an undertaking like that is just an example of common sense? I do, and I suspect that many tech vendors are racing to make this happen. It'll be a while before the evidence is sufficient to enroll patients, but generative AI is ready, suggesting patients for studies based upon our data.
00;17;37;19 - 00;18;05;23
So in some sense, where it's good enough for some purposes now and we can only imagine what it might be around the corner, you know, I think of about how clinical trials could be fundamentally changed. I think about reduction of chaos really by using standards and automation. That's accepted pretty much throughout the industry, which means more digitalization. Am I an idiot thinking that's possible?
00;18;06;23 - 00;18;34;27
I'm not going to say that, Mike, thanks. But I do think your question is a certainty and I'm betting on it. Meaningful digitalization requires a rethink. However, of what we're trying to achieve and what the necessary steps are along the way. So doing unneeded steps faster won't have much of an effect. Amazon didn't just give you a shopping cart for your goods.
00;18;35;12 - 00;19;02;18
They changed the shopping experience by providing suggestions for accessories, storing your payment information, delivery preferences, and giving you reviews of those products. We need to be thoughtful about how do we change the process rather than speeding up the unnecessary stage gates along the way. It's all about simplification with a focus on the patient. I don't mean that as a platitude.
00;19;02;18 - 00;19;27;13
Every drug company, as I said, talks about its work in terms of the patient, but it's about understanding the patient's preferences and prioritizing them. I love that. Well, when you said, you know, doing unnecessary things, unnecessary steps faster doesn't get us anywhere, that's very smart. You touched on it, but AI and drug development specifically is kind of its own bucket.
00;19;27;13 - 00;20;04;07
How is pharmaceutical research and development about to be transformed because of a I mean, what roles does it play in getting these drugs to market faster so they can help people sooner? So the mind map that I mentioned I think is informing second order outcomes. And using this framework, I've begun to focus on a few areas. First is clinical research asking the question how does clinical research change when generative AI solutions become good enough to enable patients to provide raw, real world data from digital health devices?
00;20;04;18 - 00;20;32;02
Will that make it easier to recruit patients? And then there's another question what responsibilities the sponsors have when those devices deliver worrying evidence. The second area that I've been thinking about, the second order outcomes is the patient experience. It's never fun to be a patient, but in the current environment you need to be a bookkeeper, an administrator, a note taker, a risk manager, a data interpreter and an advocate.
00;20;32;12 - 00;20;58;27
There are impressive A.I. solutions to each of these challenges that I've seen in development now. So the question we ought to pose is what happens to the patient experience when these solutions are bundled and integrated with one another? And does that amount to a virtual concierge? Since it weaves data across providers, labs, pharmacies, payers and tech stacks, the patient wins.
00;20;59;10 - 00;21;24;22
But I've come to wonder which health sector is when and which lose. Are there any ethics or security concerns that's unique to applying AI to health care? Certainly we've heard the criticisms about, you know, well, AI scrapes the web and turns out not everything on the Internet is true. So, you know, is there any kind of danger of bad data being pulled in and applied by A.I.?
00;21;25;08 - 00;21;44;26
There are tons of concerns and there are think tanks out there publishing reports on these. But the truth is, the genie can't be put back into the bottle. A number of companies have put forward thoughtful ethics guidelines, particularly from the tech sector. But we can't allow the rules to vary from company to company, and we can't depend upon self-policing.
00;21;44;26 - 00;22;09;22
The stakes are just too high. Instead, we need Congress to act in established guardrails that allow the AI industry to grow without causing harm to individuals. Congress largely ignored privacy over the past couple of decades, while the rest of the world moved ahead on that front. We shouldn't allow this to happen again because A.I. arguably poses a much greater risk.
00;22;10;07 - 00;22;35;03
When states are forced to act, we end up with a patchwork of rules that are easy to circumvent. Yeah, you brought up a really good point that, you know, while our focus is on medicine and pharma and clinical research and patients, government and business does enter the picture, how are the pharma companies responding to things like the U.S. Inflation Reduction Act and the price pressures that they're facing?
00;22;35;15 - 00;23;03;05
Well, I can't speak for the pharma companies. I do observe their attempt to prevent it from going into effect, the price pressures, the controls. But I think ultimately we need to get to a point where there is meaningful digitization to allow a rethink of what we're trying to achieve so we can streamline processes. You mentioned about how other countries jumped on the regulation of AI so much sooner than we did.
00;23;03;26 - 00;23;34;11
What are the drug costs and medical procedure cost disparities between the United States and seemingly the rest of the world? I mean, it seems like our costs are always so infinitely higher. They are. And as an American, I've got to say, I can't explain it and I am frustrated by it. And I'm frustrated when seniors or people who don't have resources can't get the medicines that they need because they're being gouged.
00;23;35;08 - 00;24;01;10
Pharmaceutical companies who are charging two and a half to three times what they charge in Western developed nations in Europe. I really do think there needs to be a rethink of the way pharma does its business to streamline it and take unnecessary steps out of the process that could reduce the costs of drug development. Yeah, and a lot of that cost in our system isn't even directly healing patients.
00;24;01;10 - 00;24;30;20
It's administrative costs. It's inefficiencies in everything from staffing to supplies and other verticals and other businesses. Those are areas where tech is really being aggressively applied to get to those efficiencies. And you're saying maybe health care is playing catch up? I think it is. You know, there are two sectors that are laggards in adopting technology globally and it isn't just in the U.S. it's government and it's health care.
00;24;31;02 - 00;24;56;19
Health care has gotten on the bandwagon, particularly in certain sectors like pharma. Every sector in health care needs to do this, though, because the economics of health care are not sustainable, as in other industries. Health care writ large needs to ask what's best for the patient and determine what's the most efficient way of getting there. Delivering that those who employ the greatest creativity will serve both patients and shareholders interests.
00;24;57;02 - 00;25;25;11
So, you know, as I think about what a pharmaceutical company looks like today, or I think about what a payer looks like today, I think the question I have is, is there something outside of your sector that you could do that would deliver value to patients and better outcomes? If there is, why are you doing it? Are you happy with the degree to which research data is being shared?
00;25;25;20 - 00;25;54;15
Currently? Let me suggest that we ask the question just a little differently. Could we improve the sharing of research data? And without a doubt, the answer is yes. What if we think out of the box here and we empower patients, as I said earlier, to make the decision, perhaps all informed consent going forward could include a question where the patient consents to release anonymized data not only for the sponsor, but for all of science, for all researchers.
00;25;54;28 - 00;26;21;16
Putting on my privacy hat, I think it's fair to say that we all expect to have control over our personal health records, and we need to empower patients to make these decisions. And I suspect there are enough examples of this now. I suspect that when patients are asked, will you make your personal information, your health records available to science for future generations, the answer is almost always going to be yes.
00;26;21;28 - 00;26;47;16
Yeah, I agree with you. Turns out not everyone's a nice guy like Frank here. Cybercrime is real. Health care organizations, particularly have been in the news lately for all the wrong reasons. Oracle's Larry Ellison and Seema Verma just wrote about it and the Wall Street Journal. Is that a winnable fight? It feels like we're getting to a place where everyone's just accepting that there is no security and we're just going to have to live with it.
00;26;48;03 - 00;27;09;19
I think it comes down to how you define winnable. I hate to tease that out, but there will be cyber attacks and there will be breaches. You can't stop them entirely, but you can sure cut down on your risk profile. Companies who are diligent can dramatically reduce the risk of appearing on the front page of the Wall Street Journal as opposed to the Opinion Page.
00;27;10;00 - 00;27;50;28
There's no silver bullet, though, and it's unlikely that proprietary technologies can beat attackers, especially when nation states are involved in the attack. When I was in government, I got a close up look at the industry, the health care industry and cybersecurity. We were in the early days of creating industry specific communities. In particular, we launched the health ISAC, which means information sharing and Analysis Center in 2010, and it immediately provided a view into breaches, a view that enabled others across the health sector to shut down the vulnerabilities that were successfully used to attack someone else.
00;27;51;20 - 00;28;15;20
In many instances, it wasn't the technology that failed us. It was social engineering that led to the breach. So expound on that. The difference between, well, obviously technology can do what it can do and that it has its shortcomings. But what do you mean? It was social engineering that failed us. Usually attackers will find a vulnerability. It could be a helpdesk.
00;28;15;29 - 00;28;44;15
It could be someone in an accounting office that has access to the system. They'll call and they'll sound serious. They may even have gotten some personal information from someone else to pretend that they're that person and doing that, they will change a password. They will gain access to a system. So it isn't the technology that failed. It's that there were other access points to the technology that someone socially engineered.
00;28;44;26 - 00;29;04;27
So humans are fool able. Oh yeah, we are not you and I, of course. But you know, other humans are. I hold in my hand the last bucket, which is if I were in charge of everything. If you were in charge of everything in the many components of health care, they would listen to you and follow your recommendations.
00;29;04;27 - 00;29;30;29
What would those recommendations be? As we sit here today, in 2024, I can dream, can't I? Make sure you can. I'm putting I'm making you head of HHS now. I guess my suggestion is what I call threading the needle. By that I mean laying out a business process that begins with life sciences research and ends with providing life saving therapies to patients.
00;29;31;14 - 00;29;59;13
And then ask yourself, how can we invest in what matters while streamlining the entire process? Because there are just too many stakeholders, too many people taking a profit, too many unnecessary steps in a process that, as I said, was designed during the industrial age and isn't needed anymore. Technology can play a crucial role, but so too will company culture, expertise and perhaps most importantly, stakeholder engagement.
00;29;59;29 - 00;30;32;18
Everyone has to be on board for changes, these kind of structural changes to succeed. Does this mean bringing back some aspects of clinical research into pharma away from crows? I don't know. Maybe. Does it involve making use of hours for real world data? I think certainly perhaps it involves personalized medicine and genomic testing would make it unaffordable. But in a world of value based care, is there a way to use the outcomes to pay for the entire therapy?
00;30;32;28 - 00;31;11;22
I think it's quite likely that generative AI is going to change the health sector, making it more efficient, less bureaucratic, better integrated around delivering value. So I think those companies that don't act could very well find themselves with a consequential decision down the road. However, companies that pursue a strategy that really rethinks with the patients in the center and delivering therapies and the science behind doing so, I think will see their benefits to not only their bottom line, but they'll provide the best care that they say they want to provide by focusing on the patient.
00;31;12;17 - 00;31;33;20
It's really great advice that should probably be heeded. Frank It's been great. Again, thanks so much for being with us. I'm sure our listeners may want to follow you or find out more. What's the best way for them to do that? Well, I'm currently on a social media hiatus, and for you, I do avoid it. But certainly anyone can follow me or connect to me on LinkedIn.
00;31;33;28 - 00;32;05;25
Okay, great. And for our listeners, if you want this level of smart all the time, go ahead and subscribe to the show right now. And if you want to learn more about how Oracle can accelerate your own life sciences research, just go to Oracle dot com slash life dash sciences and we'll see you next time.