Disrupting The Traditional Clinical Trial Model

Voiceover:
Welcome to Cowan Insights, a space that brings leading thinkers together to share insights and ideas shaping the world around us. Join us as we converse with the top minds who are influencing our global sectors.
Charles Rhyee:
Hello, my name is Charles Rhyee, Cowan's Healthcare Technology Analyst and welcome to the Cowan Future Health podcast. Today's podcast is part of our monthly series that continues Cowan's efforts to bring together thought leaders, innovators, and investors to discuss how the convergence of healthcare technology and consumerism is changing the way we look at health, healthcare, and the healthcare system. Today we'll be talking about clinical research and the landscape for clinical research is evolving with many sponsors looking to leverage technology to address inefficiencies in the traditional site-based clinical trial process and counter the rising cost of developing new therapies. The FDA has looked favorably on technology enabled and decentralized clinical trial models even prior to COVID, but the pandemic has led to broader adoption of these technologies and models clearing the way for technology companies to continue disrupting the industry at greater scale.
In our view, technology holds the potential to adjust address some long-standing issues and research, including a lack of accessibility and diversity that may pave way to a convergence of clinical care and research. To explore this topic and more, I'm joined by David Fishbach, founder and principal of Xcel Executive Business Advisors, which provides corporate strategy, operations and leadership development for technology, life sciences and healthcare. David's career has included past leadership experience with many of the leading technology companies that served the biopharma industry, including Viva Systems, Oracle Health Services, and Phase Forward. David, thanks for joining me today.
David Fishbach:
Charles, thanks very much. It's good to talk to you again, and I appreciate you being here.
Charles Rhyee:
David, I wanted to maybe start going back a little bit into your background and maybe just ask the question so how did you find yourself here in this industry, in the biopharma industry?
David Fishbach:
Yeah, thanks for asking Charles. I started my career in mechanical and aerospace engineering and actually had pretty traditional engineering roles for the first eight years or so. Right around the turn of the millennium, all of my buddies were joining software startups. So I thought, oh, this sounds like it could be interesting. I started asking around and looking for opportunities and I ended up meeting Dr. Paul Bleiser and Paul was the founding CEO and visionary of a pioneering company called Phase Forward. We were doing what now would be called cloud-based software systems for the testing of pharmaceutical products and medical device products and drug and medical device safety. I wasn't a founder, but I was pretty early and got to experience the phenomenon of compressed experience that you get in the startup world. That really set me on a different path.
I was at Phase Forward for a little more than 10 years through our IPO and through I think eight or nine acquisitions in operational and cultural integrations, some of which we did better than others. Then in 2010 we sold to Oracle. Then I was part of Oracle Health Sciences for three years until leaving Oracle to go to another life sciences SaaS company called Viva Systems, which some of our listeners may be familiar with Viva and I was brought there to build up the North American Professional services team as Viva moved into serving the pharmaceutical development side of their biopharmaceutical customer base. Again, not a founder, but another IPO. Now since 2015 I've had my own consultancy in this space. At this point, my teams and I over the years have put our hands on technology, software services and support for literally thousands of clinical trials and throughout the spectrum of pharmaceutical development and commercialization. It's a fascinating space. It's one I have no plans to leave anytime soon.
Charles Rhyee:
Yeah, that's pretty interesting. That's amazing. Maybe jump back, face forward, you kind of mentioned today we were looking at cloud-based platform. Maybe if we back up, at that time, how did the industry run? How were clinical trials and the data collected and processed relative to today?
David Fishbach:
Yeah, it's a great question Charles, and one that folks don't really spend a lot of time on in recent years. Back before we were using cloud-based systems for this, what would happen is there would be a protocol which this is still true, that describes to the health agencies, the US FDA or its sister agencies around the world, what we're going to test, how we're going to test it, what we're looking for, how we're going to protect the patient volunteers, what measurements we're going to take. Then the pharmaceutical company or the contract research organization supporting it would provide paper case report forms in literal triplicate with literal carbon. You would write height, weight, blood pressure or the patient's answers to a quality of life questionnaire, whatever information you're collecting. It would be filled out in triplicate and you would keep one and you would take the other two and stuff them in a FedEx envelope or UPS or DHL, I don't mean to play favorites.
Then you'd wait till you had enough of them and then you'd send them to the pharmaceutical company. Then at the pharmaceutical company or CRO, they would be what was called double data entry. And so people would take the case report form that had been filled out, two different people in two different rooms at two different times would transcribe it into software called a clinical data management system. If they both type the same thing, then the assumption was there was no transcription error and the data would move forward into cleaning. When we say data cleaning, and again, this is something that still is done as well, you're looking for data that don't make sense so that you can make sure that your data are worthy of trust and analysis before analyzing them. If for example, you had a 37 degree Fahrenheit body temperature or a patient that was listed as both pregnant and male, then it's likely that a transcription error was made.
You can do more complex checks such as seeing that a patient has a history of heart disease but no heart medication listed in their concomitant medications. This whole process of having the physician and her staff write down on the case report forms the information that they gather during the patient's visit, send it into the pharmaceutical company, do the double data entry, do the data checks. If something doesn't look right, 37 degree Fahrenheit body temperature, then you would send a question, a query it was called back to the physician's office and say, "On such and such a date for thus and such patient identifier," not name, right, because it's anonymous, "you listed a body temperature as 37 degrees Fahrenheit. Please confirm or correct." That entire process would take months, literally months.
Charles Rhyee:
A completely manual process.
David Fishbach:
An intensively manual process. Absolutely. Yeah, yeah, no doubt. And Dr. Blicher experienced this and it drove him nuts. I mean I call him Paul, but I'm not talking about him in a podcast, but he really believed along with a friend of his was more of a software technology guy that this could be improved. Even in the early days of Phase Forward and what became known as electronic data capture software, a couple of things happened. One was you had the ability to enter the information directly into the computer system at the point of care or near the point and time of care. Then we could also do some of these logic checks at the browser level before you even submit to the database. You put in 37 degrees and you click Fahrenheit and the system can go, "Wait, are you sure?" Then the person typing goes, "Of course not." And they change it to Celsius.
That really had the potential to accelerate access to the data of what was happening during the patient visits. When done correctly, that's going to provide you with a tremendous number of opportunities for operational efficiency advantage, but even things like health and safety. Imagine if you were testing a new drug and it's just intolerable for humans. The side effects are too uncomfortable or even too dangerous or even too deadly. Knowing that months earlier can really help you walk away from things that are damaging or harming or even killing patient volunteers or drugs that just aren't going to make it, that the efficacy isn't there.
Charles Rhyee:
If we think about it, then Phase Forward was at the leading edge of what became electronic data capture. I think Oracle was already in the business or was already serving biopharma. I can't remember when they bought Siebel, which I think is how they got into it. What kind of services did companies like that provide?
David Fishbach:
Yeah, great question. Even before Siebel CTMS, which is clinical trial management software, which is the operational software for the management of these processes, Oracle has a product called Oracle Clinical. The CDMS that we talked about, the clinical data management software into which you'd be doing the double data entry and in which you'd be doing the logic checks and then using that to send messages back to the research sites. The two largest of those were Oracle Clinical and a product called Clin trial. Clin trial was from a company called ClinSoft, which was one of the first acquisitions that Phase Forward made. Oracle Clinical was Oracle. Because Phase Forward bought ClinSoft and then Oracle bought Phase Forward, Oracle ended up getting Clin Trial, which means the two most broadly installed on-premise clinical data management software systems that really are still in use both belong to Oracle in the health sciences portfolio.
Now Oracle did have something called Oracle Clinical RDC, remote data capture. The idea there was to give a web-based front end directly into the Oracle clinical software in order to provide an electronic data capture solution and if you had Oracle Clinical, which roughly half the industry did, they offered RDC for free. It still didn't penetrate. It was offered, but it wasn't really designed with product management discipline and product marketing discipline of really understanding the use cases and the use patterns, workflow data flow and activity prompts. That's why Oracle went 10 years later, 12 years later, looking for making an EDC acquisition even though they still had Oracle Clinical RDC. Right around 2000, I mean I could go on Charles, right? Yeah, these were very interesting years. Stop me if you want to ask me a different question, but right around the turn of the millennium Phase Forward Medidata and a company called Data Track, which is still around, were really the three front runners. Data Track lost a few key head-to-head contracts and that really gave Phase Forward and Medidata the advantage to move ahead.
Other players came in later. Viva's a relatively late entrant that has built up a book of business. Medidata has been bought by Dassault, but when Oracle bought Phase Forward in 2010 phase forward had the largest installed base of clinical trials running EDC with a product called Inform. Medidata was number two. During the fear, uncertainty, and doubt that can come with an acquisition medidata did a very good job of growing their market share, took over the first position, and they do still have it with their Rave EDC product. I don't have exactly recent data, but Rave with Medidata Dassault and Inform which is part of Oracle Health Sciences are still the two largest. Vault EDC from Viva has been growing very quickly. You don't want to count out incumbents like Clario and Calix that have been around for a long time, although with different names.
Charles Rhyee:
When we think about EDC, obviously kind of changes the industry and it seems pretty transformative. When we think about other solutions that have come into the market that are now core to how clinical trials are run, it seems like a lot of them were just digitizing paper-based sort of analogs, right? We have e-versions of patient report outcomes, versions of clinical outcomes assessments, we call it ePRO, all these E acronyms added to these, what were paper-based manual processes. EDC seems like a real innovation because we're leveraging the power technology. Were there other kind of big innovations during this period of the 2000s, 2010s that kind of came here into the market that you know find pretty interesting?
David Fishbach:
Yeah. Thanks Charles. First of all, yes. Right? If you think about the core processes of pharmaceutical development and clinical trials as a subset of pharmaceutical development in the clinical trial processes, depending on how you split it up, it's 30, 33, 35, 37 different core processes and EDC obviously doesn't enable or electronify all of them. Another technology that saw a lot of growth and had significant impact on the way clinical trials were conducted in the early part of this century was IRT, interactive response technologies. IRT itself has been around for a very long time. I mean in the 1970s you could call your bank and get a press one for savings, press two for checking, but to move that onto computers and onto the web and into the cloud really opened up a tremendous number of use cases.
Even things like electronic clinical outcomes assessments or electronic patient reported outcomes and things like that, handheld diaries, many of those were enabled by the progress that was made 15, 20 years ago in interactive response technologies in general, because EDC at its core consists of collecting clinical data, clinical meaning about humans, about the patients, but it's all clinician provided data. In any clinical trial there are multiple other data sources, whether those are labs or imaging or patients filling out pain surveys or parents making observations about their children's behavior. There are many other sources that are not collected by and provided by the clinician.
Charles Rhyee:
Today, how much do you see the dependence on what is now less more patient facing technologies? I would imagine we think about the last 5, 10 years and the sole movement in digital health and consumer kind of more directed technologies kind of connecting consumers in healthcare. I would imagine a lot of that is really applicable in the clinical trial space. What are you seeing in terms of that, if we're thinking about IRTC, how is that developed now? I'd imagine app-based kind of programs delivered to patients and they can do all that entry into it, into their phone, what are we seeing in terms of that?
David Fishbach:
Yeah, yeah, it's another great question. We undeniably have seen over the last 20 years that the number of data sources has proliferated tremendously. The just sheer amount of data that we're collecting for any given average clinical trial has multiplied tremendously. It's about sevenfold, the amount of data that we collect now compared to the amount of data that we used to collect. As you and I had a chance to talk about in a different session, oddly, it hasn't really made us seven times faster or more effective or better. There's opportunities there that we can talk about in another moment or thread. The patient facing technologies, it's been very interesting because in the early days of this, you would provision a handheld device that you would send to the patient.
Now everybody has a handheld device and they want to use their own. This is also true for clinicians, this is also true for really everybody who's involved. They want to bring their own device and use the one with which they're familiar. User interfaces, patient interfaces, workflow in the patient experience have really come to the forefront because it needs to be easy, but it also needs to be something that patients are going to be willing to do over the long run. Recruitment of suitable patients for the study of any particular clinical trial given its phase and the therapy that's being studied, recruitment is challenging, it's gotten more challenging over the decades, and retention is challenging and as patient behavior evolves more broadly, especially in the United States and Western Europe, but elsewhere as well, into consumer-like behavior than patients that are participating in clinical trials have higher expectations and lower tolerance for the curation of their experience as they try to participate in helping to study these therapies.
Charles Rhyee:
That's interesting.
David Fishbach:
Which I think is good.
Charles Rhyee:
Yeah, it's interesting you bring that up and that's kind of what I was getting to, but I think a lot of times when investors are looking at some of these companies, they focus on the core technology, but a lot of times it's what part of the market are they going after? What type of pharma companies are they targeting? The companies themselves talk about the patient experience, the user experience, and I think a lot of times investors discount that little bit. Because we all have these handled devices and we look at apps and maybe think that what is the difference between one app versus the other? From what you're saying, it seems like that user experience, patient experience, clinician experience, do you find that that ends up being a real differentiator when contracts are being decided, kind of head-to-head comparisons? Does pharma not yet pay as much attention to that?
David Fishbach:
That is a great question and it's got a lot of parts to it. Yes, pharma is looking at patient experience, clinician experience, ease of use, repeatability, reproducibility as they need to, but it's happening right now by proxy in two ways. One is that often, certainly not always, but often, even if the biopharmaceutical sponsor has an enterprise contract with a major EDC provider, that doesn't necessarily mean that they're going to be picking and choosing every supporting technology package. Your platform players like Oracle or Viva that offer a lot of these things don't necessarily offer everything. Even if they offer something like IRT, you may have a specific use case or need where your contract research organization is going to be advocating to say we need something really focused or really excellent in this particular space. The CROs often are leading the vetting and selection, especially of the smaller more point solution players.
When I say by proxy in two ways, one is that the CROs are doing this for the biopharmaceutical sponsors in many, certainly not all, but many situations. The other thing is it's pretty hard for a senior vice president of clinical operations to figure out whether every particular app that's going to go on every particular device is something that people are going to be happy to use. The proxy there is really customer success. Even though I believe very strongly in the philosophies of customer success 22, 23 years ago, customer success wasn't a thing except maybe sometimes a new name for the help desk. And so when I say customer success, what I'm talking about is demonstrated measurable post-delivery operational success with the use of the technology asset. Even if they don't know exactly why, the sponsors and the CROs undeniably are putting effort into track record, reputation, and posture of customer success. Is this a trusted bear of hands that is going to make me successful?
Now, if you're a technology and service provider, one of the ways that you do that is by making your user interface easy to learn and use. The customers in their buying journey are not playing with the user interface as much as they're talking with their colleagues or other patients or reading industry reviews and articles about whether or not you are a trusted pair of hands. Part of that, I mean every market is price sensitive. I'm not saying it's not a price sensitive market, but if you consider the roughly $3 billion that is Center Watch tufts as their current estimate for about how much money you'll spend bringing a new drug to market and selling it all the way through the end of its patent life. The technology spend on that 3 billion is big if you're a software company, but it's small if you're dividing it by 3 billion. You're not going to try to pinch your pennies on your technology and service providers as much as you're going to try to make sure that you're working with a trusted pair of hands that is going to make you successful.
Charles Rhyee:
It's interesting, and I know we're going to talk about DCT in a second. Maybe we can kind of jump there a little bit. We just put out a survey this morning for those listening. It's early in early February that, and one of the things that we asked about was expectations for DCT usage, et cetera. By far, everyone expected the number of trials to be conducted through DCT to nearly double in the next three to five years. The amount of spend committed to DCT to grow by roughly 50% in the same span of time.
The interesting thing that jumped out was that two thirds of respondents said they were looking to traditional CROs to help implement that versus going to dedicated TCT players in the market. It kind of jives with what you're saying that they can't vet all these opportunities or platforms themselves, so it seems like they're relying on their CRO partners to do that work for them. Does that make sense or is it that you think that a lot of CROs are developing these capabilities? I know PPD talks a lot about it, so does IQVIA. Where do you see them in that process of having that capability versus what they've traditionally done. I guess?
David Fishbach:
Yeah, it's a great question. So top of market in biopharmaceutical companies, top 20, maybe even the top 50, often they're going to negotiate enterprise level contracts with their technology providers and then they're going to tell their contract research organization partners, this is our technology package. The CROs are going to say, "Okay." Because that's the business that they're in. There's a tremendous portion of the market where the biopharmaceutical sponsor enters the process relatively technology agnostic, and they just want the CROs to make it happen and make it happen effectively. They'll invite or allow the CROs to select the technology package. This is the rationale for some CROs trying to bring some of these technologies to market in a classic horizontal integration play. Historically over the last 25 years, that has had flashes of success, but mostly has been disastrous investments where money and jobs get lost because providing technology in this market is very different than providing service in this market.
The same goes on the flip side. I mean there are software and service companies that have tried to provide clinical services. Again, some flashes of success, but mostly each side of that has discovered how easy their partners make it look and that it's not as transferable as the other business that you're in. The other thing that's happening is that there are so many different technologies that are available, and so you can't really pick a technology package without having an execution plan for the clinical trial, and you can't really have an execution plan for the clinical trial without doing some detailed planning. It's natural for the CROs to be involved in at least influencing the selection of the technology packages. The other thing that happens, Charles, is that even with these enterprise level contracts, the length of those contracts tends not to line up exactly with the duration of any given clinical trial.
Although I have seen clinical trials moved from one technology package to another for a reason, it's extremely rare. What's much more common is if you're a GlaxoSmithKline or a Pfizer or a Merck, you're starting 300 to 350 new clinical trials every year. Your new starts are going to be in whatever your preferred provider platform is now, but you're going to be continuing to run ones that you've started at other times that may be on other vendors technologies or on older versions of your current vendor's technology. There tends to be an attitude in pharma that if it's working, we're not going to upgrade it, we're not going to migrate it, we're not going to change it, we're going to run it out where it is. And so you get these proliferation of vendors, technologies, and versions really in everybody's shop. That complexity is part of what the CRO is helping you to manage.
Charles Rhyee:
Then if we think about new entrants in the market, and maybe in this aspect, I'm thinking companies like Science 37, Medable, Thread and others who are trying to enter this, how do you demonstrate customer success? I mean obviously they've now started to show repeated bookings with existing clients, and I guess that would be a good sign, but if you're coming new to this market with a technology package plus service, what is the best way to convince a large pharma company to take that risk?
David Fishbach:
Yeah, it's a great question and it's fun for me because I do sometimes have an opportunity to advise, I'll say point solution or niche players on how to sell against the platforms, but also how to advise the platforms, how to sell against the point solution players. If you're a Science 37, if you're a Medable, Caster is another good example. You have good software, you have good technology package, you've gone out and hired a bunch of industry veterans whom we all know who have companies and their resumes just like I have on mine. You're going to go in there and you're going to try to demonstrate that you have the people who have the perspective and the knowledge to really understand what you as a biopharmaceutical sponsor of a clinical trial program are trying to do. One of the reasons why you hire veterans in this market is because this stuff is complicated and veterans know what they're doing, but also because then those relationships accrue and they come with them and they go, "Oh, well I don't know much about Medable, but wait, I know that woman and that woman and that guy."
Maybe they're onto something here. The next thing you have to do is really go for that tip of the spear maneuver. If you're selling mid-market or if you're selling a little bit down market, you know may want to be starting with the CROs and try to get partnerships in place with the CROs and really have a very clear message as to what differentiates you from one of the incumbent platform players like in Oracle or a Viva or a Medidata who do have a lot of these same capabilities to conduct some assessments and gathering of data outside of the physical office.
We said DCT, we kind of jumped right into it. We didn't really define it. Digital clinical trials, distributed clinical trials, remote clinical trials. There are all sorts of different terms all around centered around the idea that maybe there's even more that we can do outside of a hospital or clinical setting than we've been doing historically. From one point of view, that's always been true. People have been taking their pain surveys or writing in their medication logs from their houses, but the pandemic, as you indicated in the introduction, has inspired people to look at other places where this technology could be applied.
That's why we're seeing a lot of growth in things happening outside of the clinic walls, and that's the opportunity that Medable and Caster are trying to seize. You have to talk to a sponsor and say, look, you're running rave. You've been running RAVE for years. Rave is great at what Rave does, but the world is changing. Your patients don't want to come into the clinic, pick one trial or pick one clinical program. I want to show you what I've built and I want to show you what the people whom I've hired, many of whom you already know how to do with our next generation modern technology.
A lot of times in this moment in a sales cycle, I'll go, Hey, there's a reason why the internet is faster in Bucharest than in London because it was built later. Let me show you what I've built. Let me show you what I have here to offer you. The other advantage to that, Charles, is that the proliferation of biopharmaceutical companies being responsible for the book of business of clinical research worldwide is just staggering. I forget the exact numbers, but maybe 6 to 10 times the number of companies now are responsible for the top 50% or 80% numbers of clinical trials being run worldwide. There's less consolidation in the bulk of business of clinical research, which means that smaller biopharmaceutical sponsors, again, tend to have more flexibility for experimenting with different technologies or taking their CROs technology recommendations.
Charles Rhyee:
How much then has the FDA's guidance, particularly around diversity changed how sponsors are thinking about this. Right? Because you look at the data only maybe was it 2% of eligible people participate in trials, they're predominantly affluent and Caucasian. FDA is saying, "Hey, look, we need your trials to be more representative of the patient population you intend to treat." That kind of paves the way for more decentralizing this process and moving away from this site-based model. It seems like that's why sponsors are thinking about that. It seems like a driving force. Is that what you're hearing in discussions when you talk to companies about how they're thinking about things?
David Fishbach:
Yeah, I mean the short answer is yes. What's very exciting about this is that diversity, equity, and inclusion broadly in the business world is something that people intellectually have known or thought about or talked about as being probably a good thing to which we're not paying enough attention for decades. We now have a generation moving into leadership positions that genuinely believes that, that genuinely believes that it is a disservice and it is inequitable and it is even immoral to have such differentiation of access to good healthcare, good clinical research about healthcare. When you look at the history of redlining to be blunt and systematic disinvestment in certain neighborhoods and the populations that live in them and investment in other neighborhoods and the populations that live in them, and you look at the research that's been done and is being done about the social determinants of health, most major research hospitals are not in those disinvested neighborhoods.
This model in which we expect and ask that the patients will come to us, meaning they'll see our advertisements, they'll trust us, they'll choose to come to us, they have the resources or the flexibility of job or the habits of travel and mobility to be able to come to us not through anything perhaps nefarious or deliberate, but as a systematic result of these other inequities, you're going to get mostly white affluent patients and in many cases mostly male patients. There's a tremendous number of medicines that are prescribed for women that were never tested on women because they were just tested on men.
Now, that's not happening anymore, but it happened for years. With a generation that understands that and believes that, with access to technology, that means that there's more that we can do outside of the clinic and with the war for having enough patients that have the right profile to be acceptable subjects to participate in our testing. These are all factors that are going to move us toward a more conscious and conscientious and hopefully effective diversification of clinical trial participants.
The exception to this is phase one. Depending on how much our listeners know about the phases of clinical trials, a phase one clinical trial for the testing of a new drug is really just designed to determine whether it is safe and tolerable. Has nothing to do with whether it has therapeutic benefits. Because we think it's probably going to be okay, but because there could be some really uncomfortable or even dangerous or deadly side effects, phase one participants usually are paid, and guess what? Most phase one clinics are in those disinvested neighborhoods that we talked about. If you start opening your eyes and looking for it, the inequity is everywhere and it's devastatingly tragic and for me, infuriating. It's not something that we can fix overnight, but it is something that these technologies but also an awareness of these problems. It's very hard to hit a target you're not aiming for. I think that, Charles, your question is very prescient and I think it's really interesting that things are moving in that direction.
Charles Rhyee:
Maybe then talk about if we think about the technologies that exist today, what are some of the limitations if we think about where we want to go in terms of how to improve access for clinical trials or even beyond that, just to more efficiently run clinical trials, to be more effective, to bring that 3 billion cost down to some degree. Because to me, it sounds like, I mean, remember what that number was a billion that wasn't that much long ago.
David Fishbach:
It wasn't that long ago.
Charles Rhyee:
Exactly, and I forget the exact number for every day. It's millions of dollars if you have delays. Obviously pharma looks at that and they're trying to manage that. What technologies are not available today that you think is necessary or what kind of technologies that what is missing that we still are sitting at 3 billion?
David Fishbach:
It's a great question, Charles. There are some technology ideas that I do want to talk about. Before I do, let me address that 1 billion ballooning to 3 billion situation. A couple of things, 20 years ago, we are trying to sell this and we'd go, "Listen for a blockbuster drug every day of patent protected sales is a million dollars a day, and you have to get your drug to market sooner." Most pharmaceutical development has moved toward drugs for smaller patient populations, so that generally isn't true anymore. Even if it were, that's actually not the big bet, in my opinion, on the 3 billion because we still have 90, 92, 93, 95 drugs that come out of the lab and start being tested in humans that never make it across the finish line for the three or five or seven that do. Failing faster is a tremendous-
Failing faster is not proper English, but it's what we always say, failing more quickly or faster failure I guess would be pairing adjectives and nouns and verbs and adverbs correctly, but that is one of the big bets that we have in order to bring down that 3 billion is if every one of our investigational new drugs that is not going to make it across the finish line were abandoned a month earlier or two months earlier or three months earlier, then in addition to all those patients that weren't given something that was potentially harmful, that was not adequately therapeutically beneficial, were also going to save a ton of money, and that 3 billion includes the loaded costs of the things we try that don't work. Now, this is science. If everything we try doesn't work, we're not trying enough things, but if we can fail more quickly, we can begin to change the economics of that.
To me, the second opportunity to change the economics of it is we've all gotten very excited about all of the new capabilities that we have. As we've talked about before, we're now collecting seven times as much data as we once did from innumerable additional data sources like wearables that we didn't used to have, but it hasn't really made us that much faster at getting new medicines approved, and it hasn't made us that much faster at walking away from the ones that need to be canceled or abandoned. Using more data science discipline to stop collecting what we're not really using and using more data science discipline to perform meaningful, rigorous, rapid adaptive analysis of the information that we are collecting, we need to not necessarily collect less data, but make sure that the data we are collecting is being put to good use quickly. If it's not, then asking ourselves and each other why we're burdening the system with collecting it.
Charles Rhyee:
When you're looking at trials, the data requirements in the protocol, how much of the data that you looked at is being collected, do you think on average is not really useful, it's just being collected because it's always been collected?
David Fishbach:
That's a great question. I am probably not as well qualified as many other people that I know to answer it, but if I had to guess it's at least 20% and it may be as much as 50% because just as with full-blown monitoring versus risk-based monitoring, monitoring is when you check to make sure that the information that's been reported to the pharmaceutical company matches the information in the patient's chart, that the people aren't either making mistakes or deliberately lying to the pharmaceutical company. When I talk about monitoring, that's what I'm talking about. For decades, we've checked everything by hand onsite and we started applying some statistics to it and saying, "Maybe we don't need to check everything. Maybe we just need to check the most important things." Well, some of the things that we're collecting aren't important enough to check. How important are they really?
I think it's a really interesting question and I think it bears some closer look. This is an area, so I promise I'd get back to technology. This is an area where machine learning mean artificial intelligence is a very, very broad term, and so I try to tend to be more specific, but using machine learning engines along with knowledgeable and capable data science professionals to really assess the usefulness and the meaning and the actionability of the data that we are collecting or the extraordinary quantities of data that we already have collected. There are insights in the data that we've already collected. This is what people talk about when they talk about virtual clinical trials. It's not talking about having people do everything from home. It's talking about, wait a minute, we already have a lot of data. Probably people talk about digital twins and virtual control groups.
We have a lot of data and there are more things that we could do with the data that we already have that could accelerate the process, lower cost, reduce the burden that we place on patient volunteers. Patients who participate in clinical trials are putting themselves in harm's way for science. The great Steve Rosenberg, one of my great career mentors, he talks about this from the podium, but not enough people do. These are people who are putting themselves in harm's way for science, and I think that we could do a better job of respecting that and of utilizing it responsibly with deliberate use of the risk that they choose to accept.
Charles Rhyee:
Just before we get to, you said there's some new technology ideas that you had, you kind of briefly mentioned it, we wanted to create, let's say, synthetic control arms using mining clinical data out of EHRs for patients who have pH therapy. How much is that actually being done today? We hear a lot of companies talk about it. Certainly the companies that sit on all this clinical data talk about how they're going to monetize that, sell that to pharma. At this point, how often is that being used? If not much, why is it not being used more? Because it seems a very obvious use case.
David Fishbach:
Yeah, well, we've said for decades that if a hospital is using electronic health record and we've got a computer screen on a laptop that's sitting next to physically the laptop in which the health record is being entered, maybe there's something other than what we call swivel chair integration where you type into one keyboard and then you turn your chair and you type into the other keyboard. We've been saying that for a long time. The challenges are in the complexity, the variability of that complexity, and then both the real and perceived fears and risks of regulatory compliance. All those challenges are surmountable, but it means that it's not easy to just snap your fingers and do. To answer your question more specifically, are we seeing digital twins in virtual control arms? Yes, absolutely. We are we seeing them in droves? Not yet. Some markets that have centralized health insurance, like Britain for example, or if you look at the Department of Defense, you're starting to see that these are organizations that have healthcare data and also are conducting clinical trials.
They own both sides of that transaction. That's where we're seeing this experimented with. First you have to follow HIPAA and depersonalize the data, but there are some really interesting and exciting things that are happening. I don't want to praise any one particular company or technology more than another, but I got to tell you, IQVIA has the undiagnosed patient finder. Have you seen this? Have you had them demonstrate it? This is machine learning. They look at the diagnostic journey of patients that eventually are diagnosed with a relatively uncommon disease, a rare disease, and given medication, and they see therapeutic benefit.
Then they look for other patients that appear to not yet be diagnosed, but are on a similar enough diagnostic journey where they're reporting the same kinds of symptoms. They're trying the same kinds of medications, they're getting the same kinds of tests and procedures. Then they can send a message that says, "Hey, it's all depersonalized. We don't really know who it is, but we suspect that you may have a patient in your care who is 53 years of age, Caucasian male whom we suggest you test for this rare disease." How cool is that?
The idea of machine learning is to be trained on patterns and to be trained on an extraordinary number of samples actually, so that the computer systems themselves determine the patterns and then attempt to apply those patterns to new data. That's the concept of machine learning. Speaking of machine learning, I don't know how long this podcast will sort of survive for the future, but the chat bot that everybody's playing with, chat bot ai, if you ask it about the future of clinical trials, do you actually get a nice little summary about what's happening with DCT. It's actually pretty impressive.
Charles Rhyee:
ChatGPT, is that what we're talking about?
David Fishbach:
Yeah, ChatGPT, that's what it's called. I couldn't think of it. Yeah, ask ChatGPT about what we're going to see in clinical trials in the next 5 to 10 years.
Charles Rhyee:
Interesting.
David Fishbach:
Yeah, but don't stop booking with me just because I put you onto that.
Charles Rhyee:
Well, so maybe just to then wrap up, I guess here, maybe what are some of the technology ideas that you would like to see come into the clinical research area?
David Fishbach:
Yeah, I would like to see more attention paid to the patient experience and the patient journey in order to keep patients even more informed and feeling connected to the science that they're trying to advance and to know what now next, what to watch out for, why they're being asked to answer these questions over and over and over again, and really just curate that experience. I would like to see technology and machine learning applied to identifying an appropriate population of diverse patients that fit the various inclusion and exclusion criteria to participate in clinical research. This is an area where I think diversity and patient participation can really be amped up by the appropriate application of technology, and I would like to see technology used to sift through all the data that we're collecting and figure out whether we really need to collect all of it and make more meaningful action oriented use of the data that we already have collected.
Charles Rhyee:
I guess just to close out then, when you talk about these ideas, is this technology that needs to still be created or because it sounds like these are technologies that exist today, they just have not yet been really applied in clinical research.
David Fishbach:
I think it's more the latter and that it's not directly what you asked, but it reminds me of the hashtag no going back that's really, really popular in pharma and in the CRO world right now. What that means is we had access to a tremendous number of these technologies for a very long time, and they weren't really adopted with accelerated enthusiasm. Biopharma tends to be a technology lagging industry because of its conservatism, and that's okay. We're playing with people's lives. A certain amount of caution is appropriate. We have tremendous portion of the workforce that are digital natives. The idea that we did some of these things these ways because of COVID, and as we move out of pandemic status into endemic status, we can make all the patients come back into the clinics now. There's very little appetite for that in most of the mid-level management and individual contributor leadership and participation throughout the biopharmaceutical industry. You'll see these hashtags no going back, meaning just because the clinics are open again, doesn't mean that you have to come in to fill out your pain survey. That doesn't make any sense.
Charles Rhyee:
That's actually probably a good sign then for technology being deployed in greater frequency and then more interesting types of trials going forward then it seems.
David Fishbach:
Yeah, Charles, it's really interesting because it was a big thing in the early two thousands that a tremendous number of people had more advanced technology at their homes than in their offices and in their pockets than in their offices. That was kind of tolerated for a while. Then the impatience with it and the unwillingness to continue to tolerate it really separated out by industry vertical. But farmers coming around, I mean, the people aren't going to accept clunky old technology that may have been developed before they were even born as the next generations enter the workforce. I think that's a good thing. But any tool, it's a tool. What we do with it is what matters.
Charles Rhyee:
Right. Well, David, we can just keep talking about this forever, but wanted to close out here and a lot of great ideas and thoughts here. Really appreciate all your insight here and would love to have you back in the future and see how the world progresses. Just wanted to thank you for being on the podcast today.
David Fishbach:
Thank you, Charles. I always enjoy it. You always come ready and have a lot of great questions for me, and it's always a great conversation. I appreciate your time.
Charles Rhyee:
I appreciate that as well. Want to thank everyone for tuning into this podcast and look forward to having you join us on future ones.
Voiceover:
Thanks for joining us. Stay tuned for the next episode of Cowan Insights.

This podcast should not be copied, distributed, published or reproduced, in whole or in part. The information contained in this recording was obtained from publicly available sources, has not been independently verified by TD Securities, may not be current, and TD Securities has no obligation to provide any updates or changes. All price references and market forecasts are as of the date of recording. The views and opinions expressed in this podcast are not necessarily those of TD Securities and may differ from the views and opinions of other departments or divisions of TD Securities and its affiliates. TD Securities is not providing any financial, economic, legal, accounting, or tax advice or recommendations in this podcast. The information contained in this podcast does not constitute investment advice or an offer to buy or sell securities or any other product and should not be relied upon to evaluate any potential transaction. Neither TD Securities nor any of its affiliates makes any representation or warranty, express or implied, as to the accuracy or completeness of the statements or any information contained in this podcast and any liability therefore (including in respect of direct, indirect or consequential loss or damage) is expressly disclaimed.