Summary

In this conversation, Martin Ramirez interviews David Hammond, an expert in FDA-regulated fields and the academic director of the BRAMS program at the University of Washington. They discuss the University of Washington's master's program in regulatory affairs and the career paths of its graduates. They also explore the role of AI in medical devices and the challenges it presents in terms of regulation. The conversation highlights the ever-changing regulatory landscape and the need for the FDA to adapt to technological advancements.

Takeaways

The University of Washington offers a master's program in regulatory affairs that attracts students from various backgrounds, including those already working in the industry and those looking to change careers.

Career progression in regulatory affairs often starts with entry-level positions such as regulatory associate or assistant and can lead to management or director roles.

AI in medical devices presents challenges for regulation, particularly in terms of ensuring the integrity of the outputs and addressing the changing nature of AI algorithms.

The regulatory landscape is constantly evolving, driven by factors such as innovation, industry needs, and changes in leadership.

The FDA strives to keep up with the pace of innovation but often relies on industry precedents and guidance documents to navigate new technologies and applications.

Transcript

Martín Ramírez (00:00.991)

Hi everyone, this is Martin Ramirez, co -founder and CEO of Signify. Today I'm here with a good colleague and friend of mine, David Hammond, who has over 25 years of experience. I need to do that again. The teleprompter is going way too fast. One sec. Welcome to the magic. Speed being too... Okay, the beginning. Hi everyone, this is Martin Ramirez, co -founder and CEO of Signify. I'm here with David Hammond, who has over 25 years of experience in FDA regulated fields.

He began his career as a clinical research coordinator before transitioning to the medical device industry, specializing in class three medical device trials. After a series of roles in medical device companies, he became a consultant for biotechnology firms in 2008. Currently, David is a regulatory and clinical research consultant in private practice and an associate teaching professor and academic director of the BRAMS program at the University of Washington. Very lucky to have you here, Dave. How are you doing?

Dave Hammond (00:59.819)

doing well, how are

Martín Ramírez (01:01.341)

I'm doing well. Very exciting times in all things regulatory. As you know, we are newer entrants, just IT and AI nerds who see an opportunity to hopefully be of help. But today is all about your perspective. I know you have a very unique point of view in the sense that you're helping train the future regulatory affair experts of our industry.

I will let you take it from there and tell us a little bit more about the academic lens and what's happening in the University of Washington in all things regulatory affairs.

Dave Hammond (01:34.221)

So we have a master's program at the University of Washington. It started as a couple of certificates 25, 26 years ago. Really the issue was that local companies, primarily Phillips at the time, was having trouble finding regulatory professionals and they were having to basically import them from California, from the East Coast, from Minneapolis region, which has some pretty strong regulatory. And they worked with some of the people at the university to set

an opportunity to do post -graduate training. And so a certificate started, like I said, way back in 1998, I think was our first class. was Dr. Tom Hazlett and Martha Feldman were the first instructors in this program. That one spun out a second certificate in clinical research that turned into a master's degree in 2009. I've been teaching in the program since 2005, just bits and pieces of

So this is, I can hear, number 19. I teach a number of classes in clinical research, some in reg affairs, and then I work with what we call our practicum, which is our big capstone project the students have to complete to graduate. We changed directors a few years ago when our original director retired. Our interim director, Andy Sturgatius, was with us for a couple of years, and then when he went back to his regular job, which was doing 17 other things at the University of Washington,

I got promoted to the academic director and so, which really just turns out to be doing a whole bunch of paperwork and budgeting stuff that's so much fun to do. But I stuck with this program. I didn't really think that I would get into education. I think if you ask my parents of my two sisters and I, who is the most likely one to end up as an academic, it was not me. I was not a gifted student or well, motivated student, I guess you would say in my undergraduate years.

And I fell into this field entirely by mistake. started teaching as a fill in for Martha Feldman. She asked me to come in and do a couple of guest lectures. And when she just got to be too busy to continue to do this, she asked me if I would be willing to take on her half a quarter and then a half a quarter became a quarter and a half. And now it's five or six classes and all of these other things. And it's turned out that I really do enjoy teaching. I enjoy the students. enjoy the environment.

Martín Ramírez (03:45.414)

Mm.

Dave Hammond (04:02.097)

And so it's turned into something that I would not have thought of myself as for a career, but you know, here we are. Exactly.

Martín Ramírez (04:03.647)

Mm.

Martín Ramírez (04:08.12)

That's how it goes. Yeah, that's fascinating. And in terms of the background and demographics, this being a graduate program, where are most of your students coming from? Which discipline serve as the base for getting a master's in regulatory?

Dave Hammond (04:25.377)

They sort of tend to fall into a couple of categories. We have students who are already in the industry in some way. Maybe they're working in QC, maybe they're working in the lab on manufacturing. They're just starting their career in clinical research for reg affairs and they want formal training. A lot of us started with on the job training. This is how I learned to do the entire job. And on the job is great in that you're getting paid while you do it, which is, you know, it's a nice thing.

The problem is that you're kind of trained in a very specific siloed area. I started out in antibiotics and then I worked in drugs for rheumatoid and osteoarthritis. So that was what I knew. I knew drugs. I knew those two particular fields. I knew very particular stages of drug development. And I learned it a particular way. I learned it my trainer's way of doing things. Not necessarily the right way, but their way of doing it.

Martín Ramírez (04:56.447)

Mm -hmm.

Dave Hammond (05:19.423)

And so we have people that are starting in these areas and they want to expand their, their area. They maybe want to move out of the lab. They want to move out of QC. They want to get promoted. And that's sort of that other group is that in a lot of companies, you don't need a master's degree to get into regulatory affairs. There's a lot of people out there that just do it with a bachelor's degree. And that's how I started. But what we're seeing is that more and more frequently to get promoted.

Martín Ramírez (05:37.45)

Mm -hmm

Dave Hammond (05:45.559)

to move up the ladder, to move to management, to the director of the VP level, especially in larger companies, there's a requirement for an advanced degree. I mean, it may be an MBA, it may be an MHA, but why not get a master's in regulatory affairs since that's your field? So we have those sorts of people, and that's where some of our students come from. But the other half are people who did something else and they want to change. We have individuals who may be trained as a dentist or trained as a pharmacist or trained

Martín Ramírez (05:55.039)

I see.

Martín Ramírez (06:08.255)

We have individual students who are being

Dave Hammond (06:14.709)

as a clinician and nurse, they want to do something different with their career. And we see this a lot with students from outside the US, especially students who have licensure in their country that does not transfer into the US. And they don't want to lose or waste all of this medical education and this years of experience that they have, but they don't want to have to go through the process of getting relicensed in the US. And they look around and they think that working for the companies that develop the products might be a good use of their experience.

Martín Ramírez (06:19.097)

Mm

Martín Ramírez (06:27.067)

I see.

Dave Hammond (06:42.529)

We get that kind of mix of brand new to the field and people who are just looking for a broader education and to

Martín Ramírez (06:45.139)

Mm -hmm.

Martín Ramírez (06:48.587)

But is it correct to say that across the body of your students, you have a foundation in the sciences? These are individuals who come from STEM programs and they're now contextualizing that for regulatory affairs as a discipline in the industry. Is that fair or am I missing something there?

Dave Hammond (07:10.463)

No, no, I would say that most of our students do come from a STEM background. My undergraduate degree was in life sciences. Many of ours were in healthcare or other things. I do want to clarify because I always have to say this because the individual who's in charge of our practicum and one of my earliest students in the master's degree always likes to point out that her undergraduate degree was in English and she's absolutely one of the sharpest regulatory people I've ever worked with. And certainly she writes better than a lot of them. But yes, most of them come from medicine. They come from tech. They come from

Martín Ramírez (07:32.671)

That's great.

Martín Ramírez (07:39.692)

And I do have a few questions around the career path that you mentioned, but before getting to that, what are the core tenets of the curriculum? What are they learning and how is the program making them a better regulatory affairs once they go back to industry?

Dave Hammond (07:56.279)

So regulatory affairs, it's sort of a, there's three kind of major areas in regulatory. One is what we consider to be sort of straight regulatory. So these are things like regulatory submissions, dealing with agencies, dealing with labeling and promotion practices. There's quality and quality is quality systems, quality manufacturing, putting correct procedures in place. And then there's clinical research. And,

Our program, we sort of jokingly refer to as a three by three approach. So not only do we teach you those three basic tenants of regulatory. So when you leave the program, you understand how to practice, again, what we would call traditional regulatory, quality systems, quality manufacturing, good manufacturing practices, and you have experience in clinical research. We taught you the rules and the ethics and how to design the studies and this sort of thing. The other thing is that we want to make sure that they have

broad exposure in terms of product types because drugs and devices, biologics, and really even beyond this into things like dietary supplements and cosmetics, they all have their own sets of regulations. while some experience crosses over fairly well, a lot of it doesn't. So we teach you the three big product types, drugs, biologics, and medical devices, and we teach you the three major areas, which are regulatory quality and clinical research. a student that comes

doesn't just focus on one thing. They don't just learn to do clinical research. They get a broad spectrum and especially because the companies in the greater Seattle area are small, most of your regulatory people don't get to focus on just one activity. You know, they're also called on to deal with the FDA, to deal with quality, maybe to help out with clinical research. And so having a broader educational exposure is really useful to

Martín Ramírez (09:25.929)

Mm -hmm.

Martín Ramírez (09:30.483)

Right, right.

Martín Ramírez (09:39.859)

And once they go back to industry and starting as a junior RA professional, what is the pattern that you're seeing in that career progression from junior entry level to where does it go from there? What are the opportunities for people who are considering pursuing education in this field?

Dave Hammond (10:00.365)

So, you know, if you go into sort of strike regulatory, a lot of people who don't have any experience start out as a regulatory associate or regulatory assistant. The terminology kind of varies from company to company. title of the job doesn't necessarily change. I've seen regulatory associates with 20 years of experience in the field. They just, they keep the title. They're just, their portfolio of what they do grows. And of course their salary goes along with that. You tend, once you get into this to specialize either by

You find a thing that's interesting. For example, maybe your interest is in promotional materials. And so you're handling the advertising, the communications, that sort of thing that a company does. Or maybe your interest is really in regulatory submissions, the US or the rest of the world. Again, sometimes that's a personal choice and sometimes you just kind of end up there. You know, your job has you doing this and you get good at it you just keep going. Some of them move into management. I've had a number of students, especially ones that graduated a few years ago, that have moved into manager roles.

Martín Ramírez (10:46.023)

Hmm. It happens, yeah.

Dave Hammond (10:58.955)

or director roles, it sort of depends. Not everybody wants to be that leader. Some of them want to be an individual contributor. Some of them love to be a leader. They'd like to manage people. They'd like to train. And so it kind of depends, but we tend to see that people that come out with really almost no experience before they came into the program, they are starting in that regulatory assistant or that junior regulatory associate position within two or three years. They tend to have moved up to a kind of a full regulatory associate.

Martín Ramírez (10:59.817)

Mm -hmm.

Martín Ramírez (11:06.697)

Mm -hmm. Mm -hmm.

Dave Hammond (11:27.777)

sometimes into the world of project management. then up from there, like I said, students that have been out of the program five, seven, eight years, they're often into leadership positions. Clinical research flows pretty much the same way. Again, it depends whether they want to stay with the research facilities, the Fred Hutch's, the Swedish, the University of Washington, and they want to move to industry because the jobs are different, but it's that same progression. A lot of what the field gives you is they want to see experience. They want to see work experience doing these things.

Martín Ramírez (11:33.331)

Mm -hmm.

Martín Ramírez (11:51.261)

same progression. A lot of what the field gives you is they want to see experience and once you've worked experience doing these things, but to get them put the door

Dave Hammond (11:57.313)

But to get that foot in the door and to have that knowledge to get you in the door in the first place, that's where the education is

Martín Ramírez (12:02.461)

Gotcha. Is there any portion of your graduates who go to the regulatory side of the equation as in working for an FDA, for example, or EU, whatever applicable notify body, or are they mostly going to the maker of the products and services that are regulated?

Dave Hammond (12:24.557)

So most of the students go to the makers, they go to the product developers, mostly because that's where a lot of the jobs are. But we've had students every year who've gone to work not only for the FDA, but for Health Canada, to regulatory agencies in other countries, such as in Taiwan or in Japan. I believe we've even sent them to Korea, Australia, and the Middle East. like I said, that's a far less common sort of activity to get into.

Martín Ramírez (12:30.836)

Gotcha.

Martín Ramírez (12:43.977)

That great.

Dave Hammond (12:51.469)

We found that to be especially valuable for students who had prior clinical training. So people that were pharmacists, were dentists, were physicians, especially outside the US, and veterinarians for some reason. The agencies love to hire veterinarians, but those are ones that we often find end up working for agencies.

Martín Ramírez (13:07.431)

Right on. And Dave, thinking about a different topic that is everywhere we look now, it's AI or artificial intelligence. And this is something certainly close to hard to us here at Signify. We are an AI company. part of the Allen AI Institute, the incubator side as well. And when I personally think about AI as an IT person, I'm thinking about optimizations that we can perform against information systems, processing of documentation.

and the like, when we think about AI in medical device, that's a whole different field. The implications of the value and the implications of the risks are orders of magnitude higher than if I miscalculated which line to say next in a document. So what can you tell me about AI in medical device? What are you seeing from your perspective? We'd love to hear your take on that.

Dave Hammond (14:02.145)

You know, in the timing of this is for two of us, just yesterday submitted in conjunction with one of my graduate students and two people from Microsoft Research, an article to the Wraps Journal on AI and medical devices, or what we refer to as MLMDs or machine learning medical devices. I knew I was going to get that abbreviation wrong. AI has been used for a number of years, not so much in the device, but in the development

Martín Ramírez (14:14.163)

That's awesome.

Martín Ramírez (14:19.177)

for watching.

Dave Hammond (14:28.629)

A lot of these devices use complex algorithms to make determinations, to make decisions, to provide guidance or instruction. And AI is used to mine the data, to help fine tune that algorithm so that it's smarter and better than it was before. But that's that sandbox use. And really the future of AI is, you as you and I have talked about, I think as people are recognizing, is that put the AI into a device as it leaves the market, as it leaves the company and goes to market so

Martín Ramírez (14:40.689)

with some smarter better than it was before. But that's that sandbox use and really the future of

Martín Ramírez (14:50.055)

is that put the AI into a device as it leaves the market, it leaves the company that goes to the market so that it's available and the device can sort of learn as it goes. This is a really

Dave Hammond (14:57.037)

you know, it's available and that device can sort of learn as it goes. This is a really tricky area with the FDA because one of the things that you learn, I think, in day one in quality systems is this idea that you want consistent product. Product one should be exactly the same as product 1 ,000, which is the same as product 1 million. That's the idea of a quality system. It's consistent process, it's consistent product. And with AI, of course, these things are going to learn. And as they learn, they're going to change.

Martín Ramírez (15:25.509)

Let me change here.

Dave Hammond (15:27.027)

And that's the reality of this. And so if they're all using centralized, know, centralized server, a centralized brain, I guess is the wrong word to use, that sort of thing, then they're all changing in the same way, but they're still changing. If they're not connected, then they're all changing in different directions. So now a hundred thousand devices based on the data that they're getting in the use scenarios are going in a hundred thousand different directions. And this is really tricky because how do you catch when a device is starting to

off the rails when it's making, it's gotten bad data, it's making bad decisions. And this is where there's kind of this interesting crossover. And again, this is something that we talk about in this article, which hopefully will be out, I think in the fall edition of the journal is this idea that you, when we look at medical device, we do a risk analysis, we make decisions about a device for a particular use case for a particular

Martín Ramírez (16:08.808)

Mm -hmm. Mm -hmm. Mm -hmm.

Dave Hammond (16:22.551)

type of patient when used in a certain way in compliance with its operating instructions. And we think about risk in this way. One of the things I learned in writing this work in with the Microsoft team was that AI as a whole looks at risk much more broadly because AI is a generic, it's not a generic tool, but it's a generalizable tool. One that you can apply to lots of things. And so you're thinking about not only how would it work in a medical device, but how would it work in a biologics facility? How would it work in, I don't know, in a weapons mania manager? How would it

Martín Ramírez (16:49.736)

So start thinking about the ethics and how these things can change. There are two sets of risk analyses that don't overlap.

Dave Hammond (16:51.869)

anywhere. And so you start thinking about the, you know, the ethics and how these things can sort of change. And there are two sets of risk analyses that don't overlap very well. And so we have to sort of break it down and say, okay, how do we apply this, you know, what they're thinking about with AI, this much broader umbrella to into a medical device. And what we're finding is that the FDA, and to give them credit, the FDA is always a little

Martín Ramírez (17:04.893)

Okay, how do we apply this, was there thinking about with AI this much broader compromise to...

Right.

Dave Hammond (17:20.065)

They're slow, they're cautious when they make decisions, when they issue guidances. And they'll issue a lot of sort of forward thinking statements and documents, but they don't make decisions quickly because they kind of want to see where things are going before they try to draw lines in the sand. And you're just starting to see coming out of the FDA discussions about machine learning, medical devices and controls that need to be put into place. ISO standards around medical devices are just starting to implement AI.

Martín Ramírez (17:44.558)

So,

Dave Hammond (17:49.109)

And so I think what we're going to see over the next couple of years is that there's going to be a lot of gray. The FDA is going to struggle with what to do with these things. There's going to be a lot of new pathways going through. And until firm rules are written, we're going to end up doing a lot of what we call sort of precedent hunting. We're going to look for predicates. We're going to look for how the FDA has handled other products. But I think that it's coming. There's nothing that can be done about this. AI is going into medical devices and the FDA has to figure out a way to regulate this.

Martín Ramírez (18:05.85)

Right. Right.

Martín Ramírez (18:17.921)

Right. Yeah.

Dave Hammond (18:18.699)

That's gonna be a big, it's gonna be a big step for them. I wish I had a simple answer, but nobody does.

Martín Ramírez (18:23.423)

It is fascinating because, and I'm speaking again, not by any means from the perspective of a regulatory expert, but thinking about the technology, right? You can have under the definition of quality you presented, right? Like device 1000 can architecturally be the same as the first one. The data and the quality of the data and the relevancy of the data for the patients or whatever the device has gathered will end up producing different outputs, right?

the consistency might not even be the architecture and the technology in itself, but rather the outputs that it generates, which is an interesting catch 22, because you will argue now you have a more personalized medical device, but where is the boundary, right? Like where it goes beyond personalization and is doing more harm or adding more risk than delivering bad value, it's a very difficult conversation to sort out for certain. Go ahead.

Dave Hammond (19:20.013)

It is, and one of the things that we run into is, you know, it depends on the sort of use of the device. I clinical decision support software, which is something people point to a lot, is there's always what we call the learned intermediary in between. The software may give the clinician, the nurse, the PA, whomever it happens to be, advice about, you know, the patient, may want to do this, or it looks like the patient's blood levels are changing this way, you know, it might be time to consider this sort of test.

Martín Ramírez (19:29.884)

Mm

Dave Hammond (19:49.537)

But the doctor has to make the decision. They're that learned intermediary. And the same thing we see in the world of pharmacy. The software may say these two drugs may be incompatible or the patient shouldn't take these two drugs together because they're going to start to compound. But the pharmacist is the expert. They went to pharmacy school. And so there's this ability to override. Where it really gets dangerous is when you're dealing with untrained individuals. So home use devices, things that a person is going to interact with without.

Martín Ramírez (20:14.751)

Mm Mm -hmm.

Dave Hammond (20:18.763)

that learned an intermediary in there so that, when the answers start to change, the physician can say, whoa, whoa, no, no, no, that's not the appropriate thing to do. It's missing a piece of information. It's misinterpreting something. The individual at home, we get trusting. I mean, I don't know how many times I have taken a wrong turn because my smartphone has told me, turn right here, get on the freeway, and now I'm stuck in a traffic jam. We don't think, we just listen to what it has to say.

Martín Ramírez (20:19.985)

Right. Right. Right.

Dave Hammond (20:44.171)

I think that's one of the things that it's a bigger concern is, that the stuff is to move into places where people are interacting. And we see this in students. We see this in the news. People use AI and they don't think there's that lovely, the legal case where the guy basically submitted something to a judge that had how many completely made up precedents that were, you and he, then he claimed that he used AI to check whether those precedents were actually real or not,

It seems like sort of a bad idea to use AI to check AI. But the point being that they didn't bother to do the research. And the number of things that I hear people say to me, well, I searched this in the AI part of the search engine, whatever it happens to be, whether it's Google or Microsoft or whoever, told me this because it's tempted to summarize to make it easier. And for some things, it does a perfectly good job. But sometimes the answers are just wrong. And so it, like I said,

Martín Ramírez (21:32.094)

Hmm.

Martín Ramírez (21:37.684)

Yeah. Right. Yeah. I'm remaining extremely optimistic about the value that it can generate. And I think not only from the computational advancements that we have,

Dave Hammond (21:40.801)

When you don't know any better, you just sort of take what it says.

Martín Ramírez (21:54.523)

the integrity of the outputs, getting to a conclusion sooner, having that intelligence deployed as close to the action where those decisions needs to be made. I think it's a great evolution of computing in general. do believe again, particularly once we're speaking about a medical device or other higher risk implementations, there's a lot to learn. And if I can borrow from the FDA lingo, we might not have a predicate today because it's literally an innovation.

We're seeing how he goes.

Dave Hammond (22:27.213)

And that's the reality is that it is in the long run going to be beneficial. There's bumps in the short run and this is what we run into. Like I said, this case that keeps coming up when we talk about AI, I think it was more of an outlier than it was a sort of standard example of how this stuff is used. But we will get there. And the truth is that software can do more. mean, the AI is going be able to do more than we can. I laugh, I just retook a biostats course a year ago.

Martín Ramírez (22:32.965)

Mm -hmm. That's right.

Dave Hammond (22:56.385)

And when I learned it in college a very long time ago, we did all the stuff manually because the software, I mean, you had to do all this crazy hard coding into like Excel to make it calculate. And now they don't even bother to teach you. like, yeah, this is the formula, but don't bother. Just put it into mini tap or whatever the software is because the computer's faster and better at it than you are. it's, and then we're going to see the same thing with AI. just, you know, we need to get

Martín Ramírez (23:05.939)

Mm -hmm.

Martín Ramírez (23:15.249)

Yeah.

Martín Ramírez (23:21.149)

Yeah, I hear you. And innovation is one of the catalysts to regulatory changes. There is also better data evidence, things that occur post -market. One of the things that I quickly learned when I jumped into these new journey that we're going through is how much change

Even though from the outside it might be easy to be a little bit of a cynic and say these regulations never change, but even if the corpus of the regulation remains the interpretation, the guidance, how it's been implemented in industry, there's a lot of entropy when it comes to the regulatory landscape. What are you seeing from your perspective? What are some of the key considerations when thinking about that ever -changing regulatory landscape?

Dave Hammond (24:05.997)

So, one of first things we teach the students is that if you look at a new regulation that the FDA has issued, a new section of the CFR, it's always in response to something. The FDA is a very responsive agency. Something bad happens, the FDA changes a rule. Something bad happens, the FDA changes a rule. And that was sort of in the old days before these things had sort of expanded out. The reality is now, the FDA is having to respond to lots of things. So it's not

to bad things that occur with patients, things that you hear about in the news. But again, all of these rules around machine learning and AI are coming because this is where companies want to go. This is what makes the tech better. The FDA has to come up with the rules for it. But you deal with financial realities. Drugs get to be too expensive. The therapies can't be manufactured fast enough. We see this today with some of these weight loss medications

Martín Ramírez (24:44.861)

Mm -hmm.

Martín Ramírez (24:52.383)

the therapies can't be made in fact, you're fast enough. We've seen this today with some of these weight loss medications that can't be made in fact, you're fast enough. The FDA has a custom reading of these things. We're dealing with two of the big ones that just popped up recently

Dave Hammond (24:59.711)

that you can't be manufactured fast enough and the FDA has to rethink these things. We're dealing with two of the big ones that have just popped up recently that are, you know, we're seeing, I'm seeing with some of my companies that I work with, is the FDA has just put out some early guidance on how to, how a company can respond to social media. So when somebody has posted to social media about, don't like this product, this product is completely safe, or whatever they want to say, things that are incorrect.

Martín Ramírez (25:13.265)

is the FDA has just put out some early guidance on how to, how a company can

Dave Hammond (25:28.651)

You know, the company was always in a tough place, the manufacturer, which has a responsibility to say only things that are correct and cleared by the FDA. How do you respond to this? The FDA has put out guidance about how to respond, when you should respond, what you should say, how you should deal with this. And again, this is because the market has gone in a different direction. We're now starting to see it's not just ask your physician for advice, as we used to see in all the drug commercials. It's

Martín Ramírez (25:52.85)

Right

Dave Hammond (25:55.019)

Let's ask this TikToker for advice about whether I should take this particular medication. So the FDA is having to respond to this. And then of course, you know, there's the reality that as the, government changes, as leadership changes, you know, no matter what happens coming in November, there will be changes to how the FDA behaves, to how things are reacted, the, the, where the, you know, the leadership wants to focus effort and money. And that's going to change how the government, you know, spends their time. What do they spend their time worrying about? What do they spend their time reacting to? And

Martín Ramírez (26:05.62)

changes as leadership changes, no matter what happens coming in November, there will be changes to how things are reacted, where the leadership

Martín Ramírez (26:21.688)

Right.

Dave Hammond (26:25.133)

We laugh that sometimes the regulations don't seem like they change, but it's all the stuff underneath the surface, which is what are they leaning on? What are they implementing? What are they spending their time enforcing? Where do they practice what we lovingly refer to as enforcement discretion, which is it's illegal, but we don't have the energy and time to deal with this. So just, we're not going to deal with that right now. And that's the stuff that can change quite quickly depending on changes in regulations, changes in situations, changes in technology or changes in leadership.

Martín Ramírez (26:54.919)

Yeah, is it a correct assertion to think of and not to keep hammering on the FDA, but is one of the biggest regulatory bodies that we have in the United States? It seems to me that the velocity of change provoked by innovation might always be faster than the reaction time from the regulatory organization.

how do they keep up? It sounds like a never -ending chase of finding what's ground truth from the perspective of the regulatory body.

Dave Hammond (27:31.533)

I heard somebody say this in a way that I thought, you if you're talking about your children, it would be one thing, but talking about the FDA, which is they do the very best they can. So the reality is that the FDA doesn't like to get too far ahead of things. There have been incidents in the past, if you go back and look at it, it's not the right word, but situations in the past where the FDA has tried to lay down regulations, seeing things that are coming down the

But by the time they go through all the things that they have to go through in terms of public comment periods and making changes and listening to, you know, all the sort of environment to people that are engaged, the information's out of date. The rules are out of date. This happened with 21 CFR part 11, which is about electronic signatures. Every time the FDA would release a new draft version of this, by the time it got, you know, close to being ready for implementation, it was all out of date. And so the FDA has, this is not exactly a new thing for them.

A lot of times what they do is they let industry lead and they try to work within the rules that they have and then they come back and sort of build a structure behind this. So, you know, as with everything in the justice system, in the legal system, a lot of times before there are rules, there is precedent. And so we wait for the FDA. We always used to joke that the worst thing as a regulatory person that you want to be is to be cutting edge.

Martín Ramírez (28:43.485)

Mm -hmm.

Martín Ramírez (28:52.083)

Hmm.

Dave Hammond (28:52.193)

Because when you're cutting edge, means there's no path for you to follow. We teach our students, of course, you know, as a university, you can't plagiarize, you can't steal, you can't use somebody else's work. But the reality is the world of regulatory, we're going to follow the last guy in the same pathway that they went. When they, if this was how they got through the FDA, we just, we steal that same route and we go the same way that they went. But when you're the first one, when you're the trailblazer, it's a lot of conversation with the FDA. You think about mobile apps and the first implementation

Martín Ramírez (29:15.326)

Mm -hmm. -hmm. All right.

Dave Hammond (29:20.927)

medical devices or software as a medical device on mobile phones, they were nightmares because the FDA did not like the unlocked environment of me just using my own personal smartphone to run some app and do medical things with it. And so the original phones were locked. And of course, that's an unrealistic way to handle this. It took the FDA kind of some time to thought of this idea and some precedent. And, one of the nice things about when we talk about predicate and that you let them do it. So we want to do it to process.

Martín Ramírez (29:26.079)

Hmm.

Dave Hammond (29:50.301)

And that's kind of the way that we get going is that we use that precedent, that pre -existing pathway that somebody else got to go down. And then when it happens frequently enough, then the FDA says, okay, now we're going to write a guidance document. They write a guidance document about that. And then if it becomes significant enough and it kind of outside the regulations, up the new sections of the CFR to deal with this. it's just, you know, it's hard to leave when you don't know where anybody is

Martín Ramírez (30:01.985)

Mm.

Martín Ramírez (30:11.113)

Yeah.

Martín Ramírez (30:16.542)

It's hard to believe we've gone over the past year. And with that, Dave, thank you so much for your time and insights today. I don't know if you have anything else that you want to share with the audience. Where can they find more information about the graduate degree at UW or when your article is coming out with wraps? Anything else you want to share? You got a couple minutes.

Dave Hammond (30:34.867)

I will just promote a couple of things I always promote. The UW program is known as Brahms BR AMS. It can be found at broms .udub .edu or you can simply look up MS in regulatory affairs, University of Washington. Our applications for the new year close in August this year. And we're always interested in students who want to move forward in their regulatory education. like I said, I don't have a definite date for publication, but we...

applied to join the fall release for the regulatory focus journal for WRAPS. So hopefully you'll be seeing an article coming out about risk analysis in machine learning medical devices, co -written by a graduate student and I from the University of Washington, and a couple of wonderful colleagues from Microsoft

Martín Ramírez (31:21.717)

Well, David, thank you so much. I'll see you

Dave Hammond (31:24.223)

Absolutely. Thank