Case Study: e-Clinical development: From final trial to pharmacy in 30 days! - Stan Crosley, Eli Lilly & Co.
CASE STUDY: e-Clinical development: From final trial to pharmacy in 30 days! Stan Crosley J.D Legal Counsel and Advisor to e.Lilly Eli Lilly & Company
By on Feb 8, 2001CASE STUDY: e-Clinical development: From final trial to pharmacy in 30 days!
Stan Crosley J.D
Legal Counsel and Advisor to e.Lilly
Eli Lilly & Company
I am an attorney for Eli Lilly. I'sm actually Legal Counsel and Advisor to e.Lilly which is a division within Lilly now. And Doctor Dobbs's anecdote hit particularly close to home for me, I don'st know about any of you. And you can add to that the Espresso machine which, while it took me a while to figure out, I became quite addicted to after having Mr Coffee Machines in most hotel rooms in the US having an Espresso machine isn'st a really good idea. An attorney wired on caffeine is never a good thing.
You may be thinking that an attorney talking to you about how to take more risk is a little bit like the Government telling you how to save money on your taxes. I don'st necessarily disagree with that but I think I have a unique perspective in that I have been involved in a lot of the e-business and information technology dealings of Eli Lilly. In fact almost all of them since I'sm kind of the evil that must be gone through to pull off a transaction in a lot of ways. One of the things that I think people need to keep in mind when they'sre talking about in this conference and how we'sre going to squeeze out efficiencies is that while healthcare is roughly 15% of the GDP in the US and a good percentage of that is waste, a lot of that waste is really top line for somebody. And so when you's re at a conference like this and people talk about, we'sre all about getting efficiency and gaining efficiency, you have to understand that there are competing interests that exist in creating efficiencies and that some people who try to create efficiencies for you are actually not really making you more efficient. Which kind of comes back to the premise of my conversation is that, if all we'sre about is trying to get efficiencies and time to submission, then I think we'sre going to lose. And I make that comment based on the fact that if you look historically at advances in efficiency, you can look to things like the fax machine, voice mail, email, larger hard drives. I don'st know about you, but when I was in the mid 80's and working in practice, I had one of the first notebooks that came out that had a megabyte of hard drive. At the time my comment, I can remember very clearly to my colleague is, what in the world am I going to do with a megabyte of hard drive space?'s And at the time Word Perfect took about a tenth of that. Of course it wasn'st more than a year or so and with Moor's Law in full gear, that 20 meg was necessary, then 40, then 80 and now you can'st get by with less than a few gig. All of that is to say that I don'st think any of those things necessarily made me more efficient in what I do. What it did is it increased the volume. The volume of data is enormous and we'sve heard many presentations about that. And I would content that ECRF's, remote data capture, all of these things will simply increase the data and the volume that we'sre going to have. It's no mistake that a lot of the vendors now, a lot of the software vendors are coming and pitching software on a licensing model that is tied to data bit processing. Today you look at this and you cost the model out and you have efficiencies. 5 years from now, 6 years from now, Moors Law doubles every few years, the cost to run that software is going to be in order of magnitude larger than it is today, because of the amount of data we'sre going to push to that system. All of that kind of goes to say that electronifying anything isn'st really the answer. I don'st think any of you think it is. But that is a large part of the premise of what I want to talk about. The concept of saying that we'sre going to take a drug and move it from last pace and visit to launch or to the pharmacy shelf in 30
days, in essence, is not so much about squeezing the time to submission, as it is about squeezing the whole cycle, wherever that time can be found. Part of it clearly is going to be the time to submission, kind of a last patient visit to submission. You hear people talk about 30 days, 60 days, 90 days, that's clearly caught up in this concept. And what this project is about, clearly is going from 1-3 years, mostly toward the 3 year mark, down to 30 days from the last patient visit through approval of the product to the pharmacy shelf.
It's an effort in engineering processes. And even has important regulations and laws. We have to have radical improvements in speed. We have to have optimized product decision-making, interactive label, online label, all concepts you'sre very familiar with and industry-leading designer clinical plans and studies.
The strategy. It's really kind of a 5 part that you could explode up to 20 or down to one, depending on how you want to look at this and I'sve chosen the 5 part method. And with that you have an internal and external urgency. A regulatory understanding and influence that we have to get our hands around. Data standards that we either have to monitor or influence or be involved in. Internal process innovation, again the time to submission. Monitoring and influencing other dependencies that you can'st gauge right now. They'sre too unknown. Getting right into it, when I talk about conveying urgency, I think most pharma companies and from the presentations you'sve heard this week, it's pretty evident that most pharma companies believe there is a sense of urgency, that we will not be able to maintain current business imperpetuity. What isn'st so clear is that the other constituents have that same feeling, the regulators, the regulatory bodies, our vendors, believe it or not in that there is a lot of money being made currently in the way business is currently done. There is an enormous effort to not necessarily make that more efficient. As we go about talking about this urgency then, it's important to understand that I'sm not just preaching to the choir, I'sm just not talking to pharma and all of pharma are shaking their heads and saying, yes, we agree,'s because any one of the pharma companies can talk about conveying urgency. Who I am talking though too, is the regulatory agencies and some of the vendors that are here as well and that this isn'st a sustainable model. If you look at the FD review times, they'sre actually down. Total development times are increasing, but FD approval time is down. And yet we'sre confronted with rapidly increasing costs and you can see the pace of escalating cost is not sustainable. Estimated cost of developing new drugs, fairly wide range, fortunately the margins in the election in Florida weren'st quite this wide, because otherwise we never would figure out who was going to be President, but 231 million based on the compound started in 's70 to 's82 and adjusted to current 1998 dollars or you can believe the Lehman Brothers estimate which is $608 million. Either way the cost of developing a new drug is staggering. The drivers of the increased cost, the number of clinical trials run per NDA actually submitted is, if you look at from 1984 to 1995 roughly a 2.5 fold increase in just 10-11 years. Again, other drivers, the number of patients in clinical trials doubles. Clinical lab tests double. Case report forms are going to grow and I would suggest that when you get to full ECRF uptake that will increase by potentially an order of magnitude. And the reason for this is the comparative studies that are necessary now, the different evidence that the FDA and other regulatory bodies are asking for. All this means, is that there is an incredibly low probability of technical success. And if you look at the therapeutic areas, Printed Amassi; CNS, for instance, in 's85, 6% success rate on entities. And if you look that as a percent by where in the clinical trial process these new chemical entities were abandoned, you can see that you'sre still losing 51% in phase 2 and 3 which is enormous cost.
Global diversity hasn'st gotten probably as much attention as it deserves and that is that the country and regulatory requirements make it incredibly complicated to get any efficiency on a global basis. Regionally it can be achieved, but globally what you do in one area will not necessarily translate. A good indication of this is digital signatures. Digital signatures are now starting to roll across the world, but clearly you can'st have digital signatures universally yet and even Europe is there, the US is there and some parts of Asia Pacific, but Latin America is lagging behind there. So, can you count on digital signatures in a large way? The cultural differences and the treatment and diagnosis are rampant. And then it comes back to why we'sre in this business, why we'sre not making peanut butter. In the last 40 minutes, 10 new victims have schizophrenia, 132 deaths due to cancer, 228 fungal infections, 20 hip fractures, 1300 cases of anxiety, 1500 surgical procedures requiring pain treatment and 72 deaths due to cardiovascular disease. And make no mistake about it, pharma is who everyone is looking to, to make a difference. Universities and hospitals and researchers are making extraordinary efforts, but the people who are bringing products to market and making these products available, it is pharma.
The historical last patient visit to submission and you can just kind of look at this as a Lilly perspective of, and the idea that it's not really changing. If you look at the plan to cross our product mix, again, it ranges from 70 to almost 600 days. So, we have actually achieved efficiencies in our system yet? I don'st think so. Is there a lot of room to squeeze and do an L30 SA last patient visit to submission in 30 days or an L60S or an L90S, absolutely. But that's not the full game that's afoot right now. Part of this urgency is built upon the inevitability that e-business transformation has begun, quite obviously, that's why one's here today. You believe that there is going to be some transformation that's occurring. You can look to other industries, Noosh.com is an interesting site if you want to go out there and take a look, and NowDocs.com. Both of them talk about electronic document management in a way that's a little more unique. More importantly for what we'sre doing, physicians are changing. Uptake is beginning to occur. 89% of physicians now are using internet at least 6 hours per week. Now if you look over to what they'sre doing on it, 61% are still looking for stock tips or buying a car or going onto E-Bay. And you get down to exactly what they'sre doing, practice administration, some clinical info and some specific clinical info is where we really want to try and understand on their use. But the uptake is occurring. The number of practice web sites is growing, largely I believe that is due to pitches made by vendors, allowing doctors to go online without any cost to the Doc. What's missing there is how rich that environment is or robust and if I could pause just a minute and talk about one of the things that I'sve noted from consultants and others who work in the area traditionally, and no offence meant to anyone, is that there is a lexicon that pops up whenever people slip into e's speak. And one of my hobbies is to kind of track terminology in the area. Robustness now has given way to instantiate, by the way. And so if you have a vendor talking to you about instantiation, that's something you just stop him and ask him what exactly they mean. I had to get that clarity around robustness, so I feel guilty for actually using that word right now, since I'sve called on many people on the carpet for using it. E-mediate, next economy, new economy, all that kind of speak, it makes your simpler if you ask them to tell you what you mean. So, what I mean by robustness here is, how many patients are actually on the physician web site? How many web sites are just dot.com with no-one out there and no use of this system? I don'st think the surveys have shown that but it's the next step. The bottom line is they'sre starting to use e-medicine, they'sre starting to adjust to it. Clearly legal issues remain and regulatory issues but the changes are
happening. Consumers are way ahead of Docs as far as the general uptake in use for health reasons. It was mentioned earlier that health is the number one search topic on the internet now. Clearly it over took pornography about 2 years ago and kept running. I had to use pornography at least once in my conversation, just to get attention. Jan Leschley, we need to throw out the old healthcare delivery system and accept that the company, the patient is king, and we better service the king. And that's really what we'sre trying to talk about as far as making the patient centric.
The transformation of healthcare, we believe is going to come in probably 3 phases. Operating efficiency, enhancing existing services and then redefining the offer. With Phase 1 you can get this list and you can each create your own. Most of our companies now are probably well in the middle of all this. I think we'sre touching at least on each of these areas, if not fully engaged dabbling. You look at Phase 2 and you get to web site practice, market research, online patient scheduling and refills for a Doc. Online patient scheduling, that will be a dream come true. There are some sites that are trying and to my knowledge haven'st had a tremendous amount of success in actually scheduling the visits. When that happens I think you actually will have patients visiting doctors sites religiously. Interactivity with patients. You'sve heard that doctors don'st like email with patients. They have a life that they want to lead at home. The last thing they want to do is go home and answer 80 emails. So, how do we get over that hurdle? A shift from a reactive to a pro-active medicine and this is where it really starts to hit home when we can do something like that. The off hours chat rooms. Again, you'sre going to have to pay the doc for off hours chat rooms and wisely so.
Phase 3 is where you start to get to, Doctor Dobbs mentioned the wireless body sensors and there's the ample instance of tele-medicine that in essence is going on right now. But these are the phases were, if you get down to the bottom of this list, you start to get into the areas where the hair starts to stand up on the back of my neck and issues like privacy pop up with gene chips.
The network changes in drug development are probably the most important and that's where you talk about a CRO and a need for a traditional CRO will decrease perhaps as pharma's have tools to more easily manage trials themselves. However, opportunities of biotech's may improve and you can read each of these labels. When you see CRO you could also read full service pharma. When you see pharma you could read CRO. It's crossing over that's hooked up with a sales organization and an R&D shop. So, that's where the labels are misleading here. Because of the coordination and collaboration tools that are available, pharma can clearly conduct trials at lower cost but this can also lead to disinter mediation. The very tools that are going to allow us to do things, as big pharma, will allow a CR to step into our shoes and do much the same thing or the biotech to step in and do it. Regulatory is worth looking at. Will regulatory actually evolve to an advisory role? I think that's something that's going to be necessary in order for us to achieve anything other than shrinking and having cost efficiency rung out of the pre lead submission time.
Regulatory understanding and influence. Each of these issues, rolling submissions; pilot program with the regulatory agency, the Pedufa 3. Harmonization, electronic submission or review process; the ERES issues; acceptance of foreign clinical data pricing. Each of these needs to be addressed in an aggressive and up front manner with the regulatory agencies, not only in the US but worldwide. And I can tell you that each of
these is incredibly delicate issues to engage in with regulatory bodies that rolling submissions alone, when you talk about it, has all kinds of connotations that are raised. And as an attorney I can tell you that I'sm not comfortable and non of my colleagues with the concept necessarily of a rolling submission. And you get beyond rolling submission, you talk about real time access by regulatory bodies to data and you have to figure out exactly how that conversation is going to occur, but I'sm telling you that if we don'st have those conversations, if we don'st engage on each of these levels, we'sll never get beyond just shrinking lead time to submission. We'sll never get beyond just trying to create efficiency and actually get to a transforming model because, again, you make things efficient but you can'st actually go to transformation until you engage regulatory bodies in the conversation and it's not going to occur.
Data standard is kind of a sleeper in a lot of ways and that is because it's one of the biggest risks. You'sve got Cdisk and HL which probably the majority of you are familiar with, with the FDA; conversations that are going on; trying to get industry standards around data and analyses; try and harmonize. But in order for us to actually achieve a process where we are going to shrink this to a 30 day launch, we have to have data harmonization because everyone has to be looking at data in the same way and analyzing the data in the same way. And that's not only among farmers, that's also each of the research centres, the CRO's, the laboratories. The other sleeper here is the electronic patient record and the reason I say is that in the race to have data standards, is it going to be that the regulators determine what the data standards need to be? What they'sre comfortable with. Is it that all of big pharma and the CRO's get together and say, no, this is what we believe data standards are.'s Or, is it going to be Microsoft winning the doctors offices and electronic medical records and saying, you know what? You guys have been over there working, but all of your clinical investigators, here's the data model they'sre working on because we have the doctor offices tied up in EMR's and if you bring in a new data set, they'sre going to have to have another interface. And so which of those bets is going to play out and which one is going to win? I'sm not sure but there is a race for the office that's going on, there's no question about it. And there are a number of players who are running for it.
The internal process innovation gets back to us doing things more efficiently internally in order to get the documents ready. The rolling regulatory reviews is just the foundation. There's the clinical data management on top of that and then if you can get to the planning and standards and process improvement initiatives, and these are simultaneous, these are not, even though they are dependencies, they'sre dependencies that have to be initiated at the same time and that's where the rub comes in and that's where the risk comes in. If you take any one step too far before the other starts to get laid, you have more risk that what you'sve done is going to have to be redone.
Our internal process improvement initiatives involve each of these issues. The knowledge management is a major initiative in the company. Trying to understand the data we have in-house now and how we use that data and how we can use it, what are the consents that we'sve gotten that allow us to use data in a certain manner? Are we being patient friendly? Increased planning and project management. I'sm not so sure that increase is the right word here. I think there's a lot of planning that goes on. I'sm not sure that it's the knowledge management type of planning and so that's how we'sd change that caveat. Reduce the volume of the submission and that's is what is in direct contrast to the electronification, the ECRF and the e-data capture and the tendency will
be to blow up the volume of data that we submit, rather than try and get exactly the information that we need to prove and the case we'sre trying to prove, rather than just flooding the FDA or the other regulatory bodies complete the majority of their work before the last patient visit. Again, back to the online interactive label. You get data quicker and cleaner. The advantage of an ECRF and the e-data capture is that this does allow you to get cleaner and quicker data. The robust document management as you design the document in anticipation of the changes that may be coming as far as the ECRF having a more robust document that you can, and is robust again. Having a more specific document that you can alter on the fly which would lead to improved submission publishing.
Clinical improvement capabilities, really labels, driver behaviours, again optimized protocol development, connectivities between suppliers and patients and investigators, each of these steps is a heroic effort. Connectivity alone, for those of you who have been working in trying to network, even our own organization or your own organization, connecting patients and trying to figure out how they fit in with, talk about electronic patient recruitment, clinical trial recruitment, keeping them engaged in a real connected manner. The continuous analysis that's necessary. Enterprise wide portfolio decision making, I really have to stretch to imagine that we can actually do enterprise wide portfolio decision making. Portfolio management is a challenge at any pharma I'sm sure and Lilly is no exception.
The simplest vision of near term steps that have to happen if you get to near term step. That has to happen before we get to the 30 days to product approval from last patient visit, is really a data management platform and is getting industry and all of it's players engaged in a data management platform discussion that has a web interface, that has the patient/physician leads, the direct patient input and has actually access and can be read by regulators. And how that whole interface works and how you'sre able to actually webify, given the regulations and the laws that we have, is going to be the key. Because if we simply increase the volume that is going to the FDA without having the FDA involved or other regulatory bodies involved, up front early on in the assessment of that data, we will simply end up flooding the office because they are asking for more data now, we are willing to give more data even beyond what they ask for to prove the case and the end result is going to be you have a finite number of regulators who are going to not be able to assess all the data that's in front of them and we could very well do the very thing that we don'st want to do and that is slow the process down so they consider all of the data.
The other dependencies. Wiring of big pharma. Wiring of the MD offices and hospitals and privacy regulations. The wiring of big pharma and the e-notebooks, the ERPs, the portals, the data architecture, the electronic data management, Windows roll-outs, all of these have the potential to push pharma in separate directions. Anyone who is going through a merger right now knows that the ERP systems that are disparate are no fun to try and reconcile. So, as we move down paths that are dissimilar and as even within the same pharma company you adopt technology that later you would have to change or abandon in order to embrace the standard that's been created in the industry. You are talking about inertia that may be almost impossible to overcome. And so you'sve heard a lot of talk about componetized architectures. And as an attorney who's involved in these transactions, one of the first things I always ask our IT folks is, are you sure that this is component based? Is this plug n's play enough, that if we have to abandon this we can
abandon it and it will be able to plug into something else?'s And in my laymen's vernacular that makes sense to me and I let the IT people walk away thinking that I know a whole more than I do about IT and let them try to figure out exactly how that has to occur. But each of those items under that first bullet scare me quite a bit because any one of them taken any pharma and run through and fully adopted can create enough inertia that you would not be able to abandon that to embrace some other rapidly evolving business process that is occurring.
I mentioned the wiring of the MD offices and EMRs. We'sve been talking about electronic medical records for decades. But are we getting close? And I think we'sre going to get close with the advent of voice recognition software and that's personally where I put the bet. Sometime within the next 10 years, maybe, there will be EMRs in the doctor's offices. I think it is probably going to be one of two things. It is going to be payer driven. The payer's are going to say, it's time we want electronic medical records because we want to be able to hold our formularies more tightly and we want to be able to get more efficiency out of this system,'s or it's going to be legal, where legal comes in and says, you know what? A standard of care has been set now that electronic medical records are more searchable and office A over here has electronic medical records and because they had that they were able to avoid these drug interactions or these practice errors. Office B didn'st do that where for $10 000 they could have adopted this software system that would allow them to catch that error, therefore you'sre liable.'s That may be what drives hand held prescription writing. There is already some case law that is starting to bubble up around it. And it may in fact drive EMR uptake. If that happens, what system is going to win ubiquity? Is it going to be an office based system? Is it going to be a PDA? Is it going to be a Journado? HP Journado product that does less than a full document. Don'st know. Windows ES, don'st know what's going to win, but when EMR gets taken up, if it gets wide spread use it will be in the industry's best interest to make sure that whatever they'sre doing in the clinical side and clinical trial matches up to what is happening in the doctor's office because if a doctor has to have 2 separate inputs, patient A comes to my practice to get well and patient B comes to my practice to get well but can go into this clinical trial but I have to fill out this other information, they'sre less likely to enroll the patient. And the more seamless you can make that the better you'sre going to have because this bolus of new drugs that we have in the pipeline are going to have clinicians involved and one of the ways that this can happen is if we have a sync up with EMRs in the doctor's office and what we ask them to do from a clinical trial set up. I also believe that the automated reimbursement adjudication I mentioned is going to be a big driver. You'sve got a number of players in there, Healtheon, RealMed and others that are trying to drive uptake on that. Other dependencies, privacy regulations and public sentiment. And why put those as synonymous is that at the same time that we have all this uptake of electronic records, signatures, data, internet based transactions, increasing users on the internet, the public continues in survey after survey to say that privacy is our number one concern when they submit this information and yet if you get them in a one on one setting, they will respond that they are comfortable and that the data that they are submitting in that instance, because they feel that they'sre going to get some value from it. So, the value proposition that we have to have when we start doing clinical trials online, patient recruiting, data capture, cross marketing, has to be strong enough that patients will not raise up and start talking about privacy as being a big issue and why they don'st want to participate. Regulations may force that hand that public can drive law suits that can be far more onerous than any regulation that pops up. But clearly we'sre going to have to reconcile the EU data directive and even now in
clinical trial settings all major pharma and CROs are dealing with how to handle the EU data directive. Hippa promised out sometime between the election and the end of the year which is kind of a shrinking time frame. So, the good thing about Florida is that if it keeps shrinking too much, actually Hippa may be launched next year. But clearly the Clinton administration has commented that they believe they are going to release the Hippa Privacy Regs, yet this year or prior to him leaving office. This is kind of his swan song. IRB's as part of that is also going to be tough to know.
In summary, a radical see change in drug development paradigm is what we'sre talking about trying to pull off. It's enabled by technology but clearly it is more about relationships and working with pharma, regulators in pharma, pharma and clinicians and pharma and development partners, and require establishment of common processes and standards in order to pull it off. It's not really a 3 year scenario or 5 year scenario, we really believe it is a viable future state. And begin by focusing on those events which are within our control and could yield internal improvements that reduce time to submission but also set the stage now for talking to regulators and setting up a new paradigm. Thank you.