Site icon DarshanTalks Podcast

Auditing, Artificial Intelligence and Augmenting Pharmacist Options: An Interview With Former FDA Auditor Patrick Stone

white robot wallpaper
https://media.blubrry.com/darshantalks/content.blubrry.com/darshantalks/ep205_auditing-artificial-intelligence-patrick-stone-interview.mp3

Narrator: This is the DarshanTalks Podcast. Before this week's podcast auditing artificial intelligence and augmenting pharmacist options in interview with former FDA auditor, Patrick Stone, Darshan will introduce this episode with the recap for the week of Thursday, July 23rd, 2020.

Darshan: This week was really interesting because there seemed to have been two separate issues that have grabbed a lot of the headlines. The first seems to relate to the FDA's announcements. The second relates to clinical trials. Let's talk a little bit about FDA announcements, the most important one, from the perspective of what a lot of people were waiting for is that the FDA just came out with a draft guidance on how to do cannabis related clinical research.

The cannabis community has consistently pointed out that the FDA seems to withhold research and discard research. So the idea of putting out guidances on how to achieve that furthers the FDA's position that it wants to encourage the research as appropriate. Second announcement that I saw is the idea that the FDA is extending enforcement discretion for certain regenerative products or regenerative medicine products.

Darshan: That should be interesting as well because, regenerative medicine has gotten a lot of FDA issues and while certain types of 361 CTP S have not received attention, some have, so the FDA is continuing to extend its enforcement discretion. The other things to look at is the fact that the European Union is still putting out some more information. And big question is, are companies ready for the clinical trial regulations that are coming out? In the context of transparency, the big one is the fact that there will be a central website that will allow for companies to talk a little bit about the type of clinical research they're doing.

That goes to the spirit of transparency. This EU clinical trial portal and database will be a single central portal or platform for sponsors to submit applications and notifications. And it's built to allow one time submission across multiple European countries. That should be interesting for a lot of companies. At the same time, there have been other interesting announcements in the context of clinical research. People are starting to question whether in this era of diversity, why are patients diverse, but clinical research primarily being done on white individuals.

Darshan: There is also some question about... well, the FDA also announced for example, that they will resume domestic inspections and foreign inspections probably will be soon thereafter. But outside that piece of clinical research, the other bit of information that did come out is that the US is accusing hackers of trying to steal coronavirus secrets for China. Recently there was another allegation that Russia was doing something similar. So coronavirus is causing companies and countries to try to save lives and why I think that that is inherently a good thing.

I'm not sure that stealing researchers is how you get to it, but maybe there needs to be a discussion about how do countries share data in the context of a pandemic. And if all these countries are trying to steal US data, the question then becomes is US really paying for research in the beginning itself? Are other countries carrying their weight in other times. The other questions that keep coming up are just ways companies should consider doing research, health tech companies are beginning to complain that, while there are concerns around privacy or government surveillance, the fact that there wasn't a coordinated nationwide pandemic response shows the innovation is problematic in the healthcare industry.

Darshan: Now inherently, that's not news because I think that everyone knows that that progress has slowed the healthcare industry, but that's because you're talking about something extremely personal for individuals. And so it's important that health tech understand where these challenges are coming from. I see a lot of the questions being raised similar to what I've seen some of the PhD students say, which is if science comes first and that's absolutely true, but why science is important. It's not the only criteria. For example, you have to consider what people's opinions on privacy might be. I just read a recent article saying that, I think it was the New York Times actually that came on and said that China is trying to catalog the DNA of all males in China to create a central depository and repository.

The question is that if you're doing that, while there may be some significant value for research and for scientists in general, does that necessarily inhibit those individual's rights? It's something to consider. Similarly, tech companies have always used red teams and having red and blue teams implies having a significant debate. And maybe there is a need of having scientists do something similar. This could be applicable in the context of healthcare and research, but also in the context of AI, I.e artificial intelligence. So something else to keep in mind.

If you think I missed anything that wasn't newsworthy, if you think that there is something really interesting happening, feel free to reach out to me. You can find me on Twitter @darshantalks, or just feel free to reach out to me at darshan@info.com. I look forward to hearing from you.

Narrator: This is the DarshaTalks Podcast, regulatory guy, irregular podcast, with host Darshan Kulkarni. You can find the show on Twitter at Darshantalks or the show's website@darshawntalks.com.

Darshan: Hello, everyone. Welcome to another episode of Darshantalks, we have a very special guest for us. We have Patrick Stone and Patrick is an author. Patrick is ex FDA. Patrick is an auditor and a mediator who comes in and reviews facilities for a bunch of different types of facilities, but a recent note he's been reviewing auditing and remediating, drug companies, but he's obviously had experience with food companies before and devices. Is that all I could have missed anything?

Patrick: No, that's very accurate. Thank you.

Darshan: Of course. Patrick, how have you been?

Patrick: Well, I've been very busy and not traveling, which is not normal for me. In the past nine years, I've typically gone on five international trips and every other week domestic travel. So, it's been a transition. It's been a good transition, not having to travel because when you end up somewhere at midnight and you have to get to work at seven or eight, it's harder to be productive, as productive as you can be. Also saving some money on my client's side, they're reducing their costs, and realizing that we should have been doing this for quite some time.

Darshan: But now you talk about this idea of not traveling and I've done a bunch of different interviews on this, but are you seeing this rise of remote monitoring? Everyone's been talking about centralized and remote monitoring. Is COVID going to be the reason we finally really switch over to that? Or is COVID going to come and go, and we'll go back to being regular, if you will?

Patrick: Much of the audits that I had scheduled for this year and all the way through December of 2020, have all transitioned from, in-person to virtual or remote. And I do think that, this will push the monitoring side to be more remote in nature, and it will also assist with auditing being virtual now. Well, we want to take a cue on is how is the FDA going to do it? Because my clients are expecting the most value added service. And if that includes having to go to the sites or having to go to their headquarters, then that's what we'll do. Because I'm not sure that FDA is going to go virtual. I think that's what everyone's waiting on is to see, will the FDA do a virtual audit or not? And I think the answer is still pending.

Darshan: That's an interesting question. So is the FDA's position on this really as relevant as it might have been because the FDA has taken the position that they want industry to go virtual. They want industry to look at centralized monitoring and saying that, "You know what, we want you look at a risk-based monitoring plan, we want you to evaluate from a centralized perspective where the signals actually are." And if all of that is true going onsite, obviously it makes sense sometimes that there's absolutely value on that. But you're going to miss the larger signals because you're onsite looking at a single individual site.

The FDA perspective, they're really looking at problematic sites at that point, and probably already conducting a bunch of centralized review. If anything, an application review would be a type of centralized review in the end, isn't it?.

Patrick: Yes, it is.

Darshan: I guess I'm pontificating from some kind of soapbox, but what's your take, do you think that the FDA's perspective changes things? What are the upsides and downsides, if you will, on going towards this centralized or remote monitoring process? And do you also look at centralized and remote as being the same thing here, or do you see them as two separate issues?

Patrick: Well, they can be one in the same in the sense, because when you're dealing with the centralized paradigm, remote monitoring is part of that. Now, what I would say is FDA will probably take longer than industry to migrate to an all virtual or remote audit per se. I think that they will be doing more virtual audits from international perspective. We may be seeing more of this type of thing being done by FDA, but domestically, I think they're still going to try to get to the sites. And I think that sponsors will be the ones and with the help of sites, because remember the stop gap to all of this centralized monitoring, remote monitoring is access to the EMR. Allowing access to monitors, clinic sites don't want to allow even auditors access to their EMR. They want to run the EMR themselves.

Usually that takes being onsite or sharing a screen of some sort, if we're going to do this remotely, but traditionally that's been the stop gap. Now, when the sites finally realized, "Hey, we really do want to do remote because it takes a lot of time out of our schedules to get ready for a personal audit or someone coming into our clinic." And I think COVID what it's doing is because clinics and hospitals are mandating no external visitors, no extraneous work, they're saying, if you're a patient, you have to come alone, you can't come with anyone.

We're getting to this point where they really don't want you in there. And I think that is going to be a part of the norm going forward. Unfortunately, for a couple of years, we'll be keeping patients isolated until we have a good handle on preventative defenses, because we don't see a lot of preventative defense, because what we're seeing now is an explosion in cases.

Darshan: You talk about this perspective of, how sites really want to talk about sharing the EMR, if you will. And the question I land up with is, have you heard of information blocking and the Cures Act? Is that come across your table yet?

Patrick: No. No, I haven't. That's something new to me.

Darshan: Yeah. Essentially what it is is, under the Cures Act that came out saying that you can't prevent systems from sharing information. It was really set up because these large EMRs were preventing other systems from sharing information. The question becomes... and the Cures Act basically says, you can block information from crossing these boundaries because you want to control your EMR and control patient data. The question that comes out of that is if you cannot stop information from crossing like that, can you therefore prevent the FDA from saying, "I want access to your EMR. You're required to give me access under the Cures Act under this prevention of information blocking."

Therefore the FDA gets a back channel into the data and therefore starts looking at these types of systems, and starts looking at remote monitoring at a site level. Do you think that that's a direction that the FDA would pursue or quite honestly, it's just not worth, that's just not how the FDA works internally?

Patrick: I think that it'll take time for them to get to that point, but what they will do is give guidance to sponsors or CRO is to say that, based on this legislation, you have to allow if the plan calls for centralized remote monitoring, you have to allow for it and you have to work within those parameters. And if they use the law to their advantage, I think that we will see an explosion of virtual reviews of monitors not going to sites every month or every other week like they've done traditionally, and you'll see them getting online and doing their monitoring remotely, which will be very efficient.

I think that it will cut costs for our drugs. I think that was the original intention of all of this going electronic, not just for storage archiving and backup system, but also to cut costs because the costs are enormous. And where do you cut? You obviously don't want to cut corners to market, but you want to cut corners like this, where you're spending money to travel people around, that in other cases wouldn't have to. Now we know that you don't have to do that. What COVID is doing is not changing the paradigm. It's showing us that the paradigm works, that it can be effective, and it could be cost saving.

Darshan: It's interesting you talk about... I have this perspective that industries that are finally moving towards patient centricity, and patient centricity to me means four basic components that comes down to privacy. It comes down to transparency. It comes down to intellectual property. So the company itself to protect its own intellectual property. And then on the other hand access, and to me, this type of information blocking, this type of, "Oh, I can't share this information with you." Those are the transparency angle of this. Have you heard of the blue button program?

Patrick: No, I haven't. But what I have seen is Europe does a lot better job at securing patient information and keeping it secure, and they do a better job at sharing things in a non changeable manner. Because it's one thing to grant access. And it's another to ensure that in data integrity, that nothing will change or nothing can be changed while the review is happening.

Darshan: 100%. I think you're right. A lot of what I'm talking about really started in Europe and is now being considered in some way in the US. This blue button program I was talking about earlier, the way this came out is, it's a department of defense thing, that HHS has adopted. And essentially they're saying that we want patients to be able to access their own data. So if you're creating a new EMR, they're basically allowing for like a download button where you can download your own data.

Question is if patients participate in studies because they want to help the study itself, will they allow for the download of data and therefore the transmission of data, and therefore allow for someone like you to be able to remotely access this data and help other patients, which is the reason why they joined in the first place. Based on your interactions. Do you see that being a feasible thing and, or useful thing?

Patrick: It's definitely feasible, as far as usefulness, I think that there should be a time period where, if you're in a blinded trial and if you migrate into an open label phase part portion, I think that's where you can allow for this downloading of data, but during certain aspects of a trial, depending on the way the trial is structured, I don't think that's 100% feasible from your first informed consent to your last. I think there will be issues where we have to keep things in a blinded manner, and allowing for some of this would inherently allow for unblinding.

Whether it's a data on blind or a treatment on blind, we don't want either one. As long as it's done carefully and thoughtfully, it will be valuable not just for the patients records that they can bring these records to them to the next time. Because I think what you see a lot is patients don't remember what they did or what they were treated with, or if something is longterm five years, six years, it's hard to recall that. So having a jacket, having your file and the ability to provide that file to your next primary care person or your principal investigator, is valuable to them because they get the record straight and they don't have to be looking for your records in the ether.

They have to go searching for, oh, this doctor and that doctor and this hospital system. I think the value is more in the portability of your file, and the ability to not have those inclusion, exclusion violations that we see too often or these waivers.

Darshan: Talk to me a little bit further about patients centricity about remote monitoring. In your experience now having... I think of you as an expert in this area, and I've always thought about the fact that you've been in trenches, where you have battled investigators who won't show you documents, or individuals who would but they didn't know what they were doing or just were incompetent or whatever it was. And then obviously the vast majority are good at what they do and are hoping to work as effectively as they can with the FDA, with sponsors. What are the upsides and downsides of switching over to remote monitoring?

Patrick: Well, the upside would be monitors have a tough job and they have to wear many hats. And some of those hats will be diminished with virtual only. So they'll have to be a hybrid of onsite and virtual. And the plans would have to be a risk-based approach in that manner. But I would say the upside would be that, monitors have to look at almost 100% of primary and secondary data or data validation. They have to review those and show that, through data integrity that the source matches the CRF or the data listings that are going to be sent.

That upside is that they can do that more efficiently and truly be 100% without worrying about, "Okay, what time does my plane leave? What time does the taxi get here?" All of that stuff goes away. Now, some of the hats that they wear are getting the sites in compliance, like steering them back onto the protocol and showing them that, "Hey, look, this is why you're having deviations or so many deviations." And some of that training may not be as effective, or they may not catch who exactly is doing the bulk of these deviations, where are they coming from?

Some of that being at the site will be diminished. If you're not at the site, there are certain cues or non-verbals, or just understanding how their clinic works, so that, I think would be some of the downside because remember, monitors have to not just look for compliance from the PIs perspective, but they have to look at enrollment, and they have to look at milestones. And there's a lot of other aspects of monitoring that are not necessarily in a plan. They're part of the monitors, regular daily duties of, "Okay, I'm not just looking at the subjects that I haven't reviewed since last time I'm looking other areas." Those other areas will be a little more diminished if we just do remote.

Darshan: One of the things I didn't mention, but it strikes me as interesting right now is where people are losing jobs. Do you think that this type of changeover will result in monitors having to be retrained and, or losing jobs? Or do you think this was already part of what you guys are already did and therefore it's not so much needing to be retrained as much as just more work coming up your way? And it does this result in more employment opportunities therefore?

Patrick: Yeah. I think that we're not going to see less monitors. No, unfortunately we see too many unqualified monitors, monitors that maybe have non-scientific degrees or, they took a few classes and now they think they're ready to monitor. I think we see too much of that. What we need to do is, let sponsors understand the quality and the qualifications of a monitor and help sponsors understand how pivotal monitors truly are to their project. And FDA focuses a lot on how much monitoring is done. Even the guidance tells you for QA oversight and monitoring. The more you have the last your review cycle will be.

Obviously that comes with the more quality oversight and QC monitoring, you have will reduce your review cycle. It's stated in the guidance itself. So monitors shouldn't necessarily worry about job security because there's so many pipelines coming out that there will always be a need for monitors. And even if they become as efficient as they possibly can conducting remote reviews, there will still be a need for an army of monitors. And this is the way it's going to be.

But again we need sponsors to understand what qualifications are truly the most important and necessary for a monitor. And what's the disease state, and what is the product itself? Is it understandable? There's a lot of areas there that sometimes are missed on, we just need a body to go monitor. Hopefully that's not the thought process anymore.

Darshan: That's interesting. What does the future 'good monitor' look like in your opinion? Let's say, you're now a sponsor. You are now looking at a bunch of different CROs and you have figure out which CRO you're going to choose based on the types of monitors that they have? who do you look for?

Patrick: I would be the toughest QA person reviewing CVs. I really would, because I have seen the worst of monitors and I've actually had to assist with closing down monitors in Texas because they were hiring librarians, they were hiring basket weaving degrees and things of this nature. And it's like, "Wait a minute, do you have any scientific qualifications for yourself or your monitors? Do you have any scientific background?" So I think one of my biggest parts would be, do they have at least a biology degree, or some degree in science, chemistry, microbiology, whatever? Some scientific discipline that you'll be dealing with.

Then obviously there'll be the, have you ever worked with this class of drug or with this disease state? I would be wanting to see monitors that have three to five years experience in this class of drug or type of disease. For my taste, I would be looking for those physician assistants or nurses that either left for their own reasons, or they wanted to transition into the clinical trial place.

I would be looking for those type of credentials because let's face it, monitors are not just connecting the dots, and they never have just connected the dots. They're not just scouring your data to see where you deviated. That's a part of their purview, but a lot of it is, patient safety, like, are you enrolling subjects with the correct histories with documented data that shows they have this disease state and that they meet every inclusion and don't mean exclusion?

They're really special case employees. They're not QA, because obviously QA, you want to hire understanding, but at the same time, they have to understand details that aren't written in a protocol or investigator brochure.

Darshan: I'm going to challenge you a little bit, because I want to understand your thought process here. The guidance, if I remember correctly also says having requisite experience, but you place a higher value in the degree. Tell me what the experience suggests and why that distinction matters to you?

Patrick: Well, three to five years in the program area is still minimal experience. But if you set the bar too high, you'd never find the right fit or the monitor that you need. I place a high value on degree or background because, in many cases it's hard to translate a literature degree with understanding a gene therapy product, a plasmid or some sort of new chemistry. And in order for them to understand the protocol itself, they have to do homework. And I would rather not have to have my monitors do homework to understand their job well. And experience counts for a lot.

If you have someone with a literature degree that has 10 years experience, hopefully they've gotten to a level where they do understand things and maybe that could be a consideration for the monitoring, but at the same time, I really want that background because the way I look at it as the way FDA hires individuals, because I was a part of the recruitment of FDA minorities.

Darshan: Interesting.

Patrick: And so, because FDA puts us a high background on science-based degrees. And Europe does too. If you look at the way EU hires their inspectors, they won't touch you unless you have 10 years experience in the field. And a lot of them are PhDs, they're MDs, they're pharm Ds. If you look at the quality of inspector from the EU, the caliber is much higher. It's a lot harder to get in. But going back to the monitor part and the CVs and the expectations of background, I think it's important for monitors to have some scientific background. It seems like a big stretch to start off your career as an English major and transition into biotechnology, drug development.

Darshan: To me this is fascinating. I'm going to take a slight detour because you said something that fascinates me, in that you said that in Europe, the monitors tend to be pharmacists. I've been pharmacists. And my experience in the US is that number one, pharmacists aren't as involved in clinical research at all. As a pharmacist I know I get a bunch of people asking me, "Well, how do I get into clinical research?" And I go, "What would you do?" Most of them end up gravitating towards being what's called medical affairs or being in sometimes advertising, sometimes MSLs. But most don't go into clinical research. Why do you think your... and if your experience indicates the US, either doesn't have enough or has more, depending on which place you're looking at pharmacists, and why does it matter?

Patrick: I was speaking more to the EMA or EU auditors, like the regulatory auditors, but I have seen when I go to Europe and Italy, France, I have seen MDs as monitors. I'm like, "Wait a minute. Why are you monitoring? This is weird to me. Do they pay you enough?" Monitors tend to be on the middle scale or lower scale because they're the army, they're the ones where the rubber meets the road. So, you wouldn't necessarily expect to get MD pay for monitor, but I do see it. And I think that when I've asked these monitors, like, "Wow, how did you make this transition?" And they were like, "Well, I just..." Even pharmacists, per se, they said, "I didn't want to be behind the counter filling scripts anymore. I wanted to be in the innovation part and this is where they've started me out."

As obviously they move up the ladder fairly quickly, they become managers of monitor teams, or they become cleanups, or like you said, regulatory. Regulatory is kind of hard, because if you switched from pharmacy to regulatory, you're talking about legal aspects and submitting items, and it's a lot more complicated, but for the cleanups and monitoring part, we do see this in Europe, actually I have one of my best friends is a PhD in neuroanatomy and he's a monitor. I don't know, he tried to get an FDA, they wanted to put him in imports. And he was like, "Wait a minute. I have a PhD, I am not going to do imports. I want to be in domestics. I want to do the drug work and innovation."

They didn't hire him. And they were like, "Well, we have an opening for imports. You can transition to domestic later." But even in US because we have so many degreed individuals, which is another reason why I would stick to Scientifics because we have a glut of biology majors out there that are washing cars, or they're painting, or they're not doing anything in their field. So they're there. They're available. They're out there. It's just a matter of finding those individuals that gave up on their biology background. But EU is the same, there's a high demand for jobs and qualified individuals. And I think that the Europeans don't really see it as a downgrade. If they start as a monitor or if they're a pharmacist and they transition to monitoring, they don't look at it the same way we would as, "Oh, that's a step down." Well, no, in their mind, it's not.

Darshan: I've never thought of being a monitor as being a downgrade. I've always thought of it as a lifestyle choice. I find that pharmacist often like being, if they have a choice that pays equivalent, they want to stay close to home. But I know people, sounds like you are one as well. You enjoy being on the road, you are-

Patrick: I do. I do. It's hard to say that I have a family, but being on the road is just... because FDA had me always jumping through hoops and going here, there, and everywhere. It's just part of who I am now. I'm like a great white shark. If I stop moving, I think I'm going to die. But I've found in these past four months, that's not true that I can be a nurse shark if I have to. And I can just sit there on the bottom and breathe my oxygen. But yeah, I am a travel junkie. I like to experience the cultures, the foods. It's not a vacation, but you still get to see things that no one else gets to see. And be places no one else gets to go.

Darshan: Well, let me ask you one last question. We usually keep this to be like a tight 15, 20 minutes, but you're fascinating to talk to as always. So we're well past that right now, but I have one last question before we call for this interview, but what do you think about artificial intelligence? If one of the points that you're looking at is, do the numbers match? Is what you put in the CRF exactly what you had elsewhere? Like the matching. AI can do that quicker, faster. What does the value... do you think AI will become competition or do you think AI will augment what a monitor does? What do you think is going to look like? The future's going to look like?

Patrick: I hope AI does not compete with us, because there's a logic based aspect to AI that is rigid, whereas a human can extrapolate extraneous variables. I would hope that we use AI to augment data integrity. And that even with AI, I think FDA would possibly expect like a 10% review or an escalation review. So if you look at 10%, you see that AI didn't catch things, that you would bump it up to 50% or figure out. But I'm not sure that we're at a point where AI can extrapolate some of the variables that we're needing to review, but it can definitely augment and add value to just the drudgery of connecting A to B.

Darshan: Which is interesting to me. I think the idea of having a first pass through AI, then having a first transition with orders, and then it can bring in senior auditors like yourself to connect the dots, which no AI, at least not anytime soon is going to be able to see or understand. But what you're getting is value added services that an AI simply cannot provide. And sometimes your early stage monitors are still learning. To me that's where a Patrick Stone, he's already got the information, he's already looked at what the AI has caught, when it got bumped from a 10% to 50%, and now he's going this one I'm worried about, he goes, look at, to me that's risk-based monitoring at that point. And that's truly bringing in the experts to look at what does this actually mean? But do you think we're five years away, 15 years away? How far away are we from that? Or is pharma just-

Patrick: What's already happening in certain cases where AI is augmenting the source data reviews-

Darshan: Okay. I don't know.

Patrick: Yeah, it's happening in certain areas. I think, if we look to the leaders in AI, we have to look at Japan and we have to look at Europe. I would hope US would be in that category. But it seems like we're focusing our AI on like chat bots or other things that are not clinical in nature or not pharmacy or medical in nature. But just like service type. But if we look to Japan, I think their AI is robust and it's developing rapidly. And they're leaders in that. And we'll see a lot more of that usage in their area because, with ICH GCP, we're all kind of one now.

Darshan: Yeah. Well, this was as always just a super conversation. Thank you again, Patrick, for being with us. Would you be open to coming out again in the future because I expect there'll be loads of questions I want to ask more?

Patrick: Yes. Now that I'm not traveling and Dan asked me the same thing, "Can we do this again because it's been too long?" I'm like, "Look now that I can breathe. And then I'm stuck to my desk. I have a lot more time for these type of outreach and information sessions."

Darshan: Excellent. Let's find to have that soon and we'll be in touch. Thanks again.

Patrick: Appreciate you.

Darshan: [inaudible 00:42:25]

Patrick: Thank you.

Narrator: This is the DarshanTalks Podcast, regulatory guy, irregular podcast with host Darshan Kulkarni. You can find the show on Twitter @darshantalks or the show's website@darshawntalks.com.

Share this