Being an Engineer
Being an Engineer
S5E29 Steve Gompertz | Medical Device Quality Control & Regulatory Requirements
Steve Gompertz discusses the complexities of navigating the regulatory landscape in the medical device industry. He shares insights on the evolution of global regulations, the balance between innovation and quality control, and strategies for developing effective quality management systems for companies of all sizes.
Main Topics:
- Regulatory compliance in medical devices
- Tailoring quality management approaches for small vs. large companies
- The core components of a quality system and its evolution over time
- Emerging trends in medical device regulations, such as AI and cybersecurity
- Improving quality management systems and achieving regulatory compliance
Steve Gompertz is a highly experienced leader in quality systems management and regulatory compliance, currently serving as a partner at QRx Partners. With a career spanning over 30 years, he has held significant roles in various companies, including Pelican Biothermal, St. Jude Medical, Boston Scientific, and Medtronic. At QRx Partners, Steve specializes in quality systems development, project management, and engineering automation, helping companies navigate complex regulatory environments and enhance their quality management systems
Links:
Steve Gompertz - LinkedIn
QRx Partners Website
About Being An Engineer
The Being An Engineer podcast is a repository for industry knowledge and a tool through which engineers learn about and connect with relevant companies, technologies, people resources, and opportunities. We feature successful mechanical engineers and interview engineers who are passionate about their work and who made a great impact on the engineering community.
The Being An Engineer podcast is brought to you by Pipeline Design & Engineering. Pipeline partners with medical & other device engineering teams who need turnkey equipment such as cycle test machines, custom test fixtures, automation equipment, assembly jigs, inspection stations and more. You can find us on the web at www.teampipeline.us
Hey everyone, the next PDX event is coming up. So mark your calendars for Thursday, July 18. And this one is free and online. And this event simulation expert and 35 year veteran of the industry Krishna right tour will teach you the basics of setting up running at evaluating results of FPGA simulations. That's finite element analysis, you'll learn how early stage simulation can help reduce development times decrease warranty costs and reduce part weight through topology optimization. This is not a sales pitch. As with all PDX events, the training will focus on practical skills you can apply to your work right away. Krishna will also be doing q&a To answer your real world simulation questions at the end. The webinar will cover what simulation is and how you can leverage it in your product development work, how you can save time and money for your company by doing simulations early in the product design cycle, how to make what if design changes, and quickly see their effects on product performance, and how to optimize your design for things like weight, strength and size. Again, this PDX event is online and happening Thursday July 18. Register for free on the wave dot engineer. That's the wave dot engineer. See you there. Hello, and welcome to the being an engineer Podcast. Today we have the privilege of speaking with Steve Gompertz, who is a highly experienced leader in quality systems management and regulatory compliance, currently serving as a partner as QRx partners. With a career spanning over 30 years he has held significant roles in various companies including Pelican Biothermal, St. Jude Medical, Boston Scientific and Medtronic. At QRx partners, Steve specializes in quality systems development, project management and engineering, automation, helping companies navigate complex regulatory environments and enhance their quality management systems. Steve, thank you so much for being with us on the podcast today.
Steve Gompertz:Thanks for having me.
Aaron Moncur:All right, well, what made you decide to become an engineer?
Steve Gompertz:It was almost more of a whim But it stemmed from growing up, I liked anything technology. You know, I kind of grew up at the beginning of the space age. So I wanted to be an astronaut. Yeah. And then somehow I ended up wanting to be a geologist. And then I found that they don't make any money. I'm gonna get into cars. And I thought, oh, I want to design cars. I guess you you're you become an engineer to do that. And so you went off to college, actually, initially, I applied to be an electrical engineer. And I ended up graduating as mechanical engineer. So I mean, our lives always take these weird turns. And when you're that age, you're not particularly mature yet. I don't think a lot of thought goes into it. It's just sort of where does the breeze take this leaf? Right? And during freshman year, we had this engineering 101 class. And part of that was every week, they brought in a different department head to talk about the life of that type of engineer. And like I said, I had gone in as an electrical engineer, and two weeks in the head of the electrical engineering, electrical engineering department does his presentation. And that's when I was first in a panic because it was like, that doesn't sound at all like anything I want to hear at college, and I hate the idea of when I signed up. And then luckily, the next week was the mechanical engineering department head. And I went, Oh, that sounds really good. Nice to switch over to mechanical engineering. But then what's funny is I've never held a position as a mechanical engineer.
Aaron Moncur:Okay, I was going to ask you kind of got into engineering because of cars. Did you ever work in automotive?
Steve Gompertz:I did not. So, you know, what I was in love with, I think was the shape of cars. Not like, I mean, I'm a Motorhead. I have a high performance car. But I wanted to design the shape of cars. And at the time, I didn't realize there was like this thing called industrial engineering or industrial design that just wasn't major than was really what I probably should have gotten into. And I got out of college and I got a job as a software engineer, developing Computer Aided Engineering applications for electrical design. So I never really got away from the electrical side thing, but here I am with my mechanical degree. And I'm coming up with tools for electrical engineers. It all worked out.
Aaron Moncur:So looking back over your career, if you had it to do all over again, would you make any different choices as far as like, fundamentally what you know what your expertise, your background is in?
Steve Gompertz:No, I mean, it's been a great career, it's been fun, it's been a great ride, we've gotten a chance to do a lot of really interesting things and become an expert in a number of areas. And I've been able to change that, you know, over time, so no, I wouldn't I wouldn't change a thing. I suspect if I, like I said, if I had been aware that there was this concept of industrial design, I probably would have gone down that path, and maybe stayed in that. But I wouldn't have ended up where I am today, loving what I'm doing.
Aaron Moncur:Yeah, yeah. Well, one of the things that you are an expert in is regulatory compliance. Can you tell us just a little bit about what what even is regulatory compliance within the context of medical devices? And then what have been some of the big changes that you've seen in that field? Over the years? Yeah,
Steve Gompertz:yeah. So the idea is, right, when you're doing something that can have, you know, adverse effects on patients, and the people using things like medical devices, to your medical professionals, or even lay people, you know, obviously, the government steps in and goes, Look, we need to make sure you're doing this right, right. The word safe and effective are what drive everything in that industry. And to make sure that there's no, you know, wiggle room there, they're gonna go, we're gonna give you the regulations and tell you exactly how we want that done. Or at least, here's the guardrails, right, and the constraints you have to operate with. And so as a consultant, what I do is to help make sure companies understand how to interpret that in med device, is actually pretty hard. Because the concept of a med device is so broad, the regulations are not not particularly prescriptive. You know, everything from a tongue depressor to a piece of software as a med device. Right? So how would you write prescriptive rules for how to manage those different things, and pacemakers and heart valves and catheters. So the device side of things is actually pretty complex. And that's where consultants like myself come in is the guy, when you look at your situation, read between the lines ago, here's what they're going to expect, or how they're gonna expect you to apply this rule, which doesn't give you a lot of detail. Yeah, make it work and drive you to make sure that you've got a safe and effective device. I think what's changed a lot, obviously, 34 years into med device, after seven years in computer software. I think, you know, there's obviously has been a lot of change. One that comes to mind is kind of the globalization of regulations used to be, you know, it wasn't that other countries didn't have their own regulatory processes, but not all of them did. And whatever countries did have them basically just copy the FDA USFDA. Now, over time, as they've matured, and they've started to realize what do they want out of this? They're not all going oh, FDA doesn't always get it correct, at least from our perspective. And so we're seeing, you know, a lot more proliferation of regulations where countries have taken the time now, to understand what they think it requires to protect their population, find the manufacturers make sure the manufacturers do the right thing. So it's been interesting to see that evolution of the regulations in different geographies. And to see some of them like particularly the in Europe, to see it really starting to take the lead, you know, we're now that's kind of what everybody's copying and where things are going. And where it's a lot harder. Used to be was considered the FDA was pretty tough, and it was hard to get devices into the US market. So you always went somewhere else first, we're actually seeing that is now completely switched. A number of our clients are your base, and they don't want to get on the market. They want to get on the US market because it's now the easier market to get into.
Aaron Moncur:I've heard a number of people comment about the increasing. What's the word I'm looking for? Severity is not quite the right word. But I guess strictness of the guidelines imposed by regulatory bodies, the FDA. Do you think that we strike a good balance at this point? Or is the balance tipped? towards the side of maybe it being too robust, a set of regulations and it's impeding innovation?
Steve Gompertz:No, I don't think it is impeding innovate. I mean, the you know, the manufacturers are always scream, oh, this is going to slow us down and you're just going to make it harder for patients to get the benefits. But I don't think there's any data that actually supports that. That that's happened. When you look at like I said the EU medical device regulations are certainly more prescriptive. And you might consider they're more burdensome, they're taught they are definitely tougher. But when you study them and you go, but they're right. It shouldn't be hard. Right? If I'm going to put a pacemaker in somebody's body, that shouldn't be an easy thing to get on the market, right? Yeah. And as we advanced the technologies, and now you add in things like AI and ML, it's getting much more complex and you go on, yeah, there probably should be some really tight rules about this. So I think they're on the right track. And that's why other countries are now following the US lead, and I suspect the US we will as well, in the not too distant future. It might have some impact, you know, on cost of design and development for new devices. But again, it's long yet it probably shouldn't be hard for certain, right? And everything gets classified, right? So if your class one low risk, you're making tongue depressors, the rules haven't changed. And everywhere, it's still pretty easy. You get the class two, class three, and then some go to four levels or five levels. But as you go up, the risks go up. Yeah. And not surprisingly, them yeah, the hurdles get higher as they should. And if you want to play in that field, then yeah, you should be prepared for what it's going to take to get there sure, can't come up with a new implantable device, and think that that's going to happen in six months. I know your engineers could probably do a design in six months, but they have to prove that it works. Right. And that it doesn't do any harm.
Aaron Moncur:Yeah. So let's talk about that, that the trade off between speed aggressiveness of your development, innovation, and and the need for, you know, rigorous quality control. Is there a project that you're able to share where you had to help a company strike some balance between the rigorous quality control and innovation and developing something quickly? And how do you find that balance?
Steve Gompertz:Yeah, wow. Yeah, I would say every branch, everybody, like businesses are in business to make money. On the opposite side, the regulatory bodies couldn't care less how much money you make, as long as what you do doesn't cause any harm. So it's always that balancing act. And, you know, part of what we hope we provide is finding the correct pathway. Because there are multiple pathways, there's not a one size fits all on the regulatory world. And you can definitely put yourself on a pathway that's too hard or harder than it needs to be. So we help companies try to see the right place. But then, once we realize, look, this is the appropriate pathway for you, then it's like, there's no getting around this. And that's when you get a chance to challenge you know, their company ethics on things, you know, that they get it that they're gonna have to do the right thing here. The, my experiences, most companies understand that and are fine with that, and they might make the right decisions. But you definitely see where they struggle with, you know, should we really reduce that risk? Well, the requirement is that you will, you can try to dance around it, but the end of the day, you can either deal with it now, or you can go down the pathway, and at the last second of FDA deny you access to the market, because he didn't do the right thing. And that usually, if you've got people around, he's going to do the right thing for the right reasons. And that threat of that they're going to put all this money there, because it's not cheap to get a device on the market. And there's that possibility that the FDA will say, No, they that hammers on pretty quickly, they get that and they go, yeah, I could have dropped millions of dollars, by not taking the right advice, and suddenly find that I'm nowhere near the finish line. Where Yeah, yeah, they usually get it. So you can usually turn that you want to use the word greed, or desire fake money into the driver that gets them to do the right thing. You might want to do that sooner. But you got to well, we could jeopardize the entire investment. Yeah, right. They usually get that
Aaron Moncur:You've worked at some large medical device companies, Boston, scientific Medtronic, who of course have their own procedures and protocols for how they do quality and adherence to regulatory requirements, things like that. There are also lots of companies out there who are small, you know, startup medical device companies who have to go through the same have to adhere to the same FDA regulations. But I have to imagine that in a small startup, things are a little bit different than in a giant company like Medtronic or Boston Scientific. How is your experience working for these larger companies influence your overall approach to quality management and regulatory compliance?
Steve Gompertz:Yeah, it's important to understand that there isn't a one size fits all and when I Uh, before I became a consultant, you know, when I've worked or applied for jobs at smaller companies, they'd often you know, meet with the CEO or the, you know, the head of quality wherever he's going to report to. And they'd always say, you're not going to come in here and put a Medtronic size quality surveys, because we're not. And I go, I understand that, right? And no, there isn't a one size fits all I'm not just because there was at Medtronic doesn't mean that that's all I know. And I'm gonna put their system in place here, because you're right, that won't work for you. And as part of what we offer to our clients today, where we focus direct focuses on small to medium companies, and we make it clear that we're, we see this as an evolution, when you're in the pre revenue stages, and you don't even have a prototype yet, you do need some parts of your quality system, you don't need the whole thing. And the parts you have don't need to be quite as robust yet, that will come over time and will evolve the system with you, as the company evolves and succeeds. And then as you get bigger, that adds different levels of complexity. When you get the multi site multi nations, you know, then yeah, your quality system has to evolve. You know, it's one of those things, one of the requirements is that you have to do periodic reviews of your quality system, and you're looking at the quite the three points are, is it suitable? Is it adequate? And is it effective? Everybody gets the third one effective, right those at work, it's the suitable and adequate is where you're kind of wiggle room is where that's that's the regulatory bodies and the standards bodies saying, we're not going to give you a one size fits all. But you need to figure out what is it you need to achieve the intent of this standard of this regulation. So it's always about right sizing it to the company and the point in time. But there's so many variables, things like the risk of the device, if you're a small startup, and you're developing a new neurological device. To some degree, it doesn't matter that you're not Medtronic, what you're doing is just as hard as what Medtronic would do in that same space. Yeah. So they can't really back off, you know, the regulatory bodies can't back off the rules for you, just because you're small.
Aaron Moncur:right? Still has to be safe, doesn't matter how big you are. Right? Right.
Steve Gompertz:And there's a sort of, there's what they call the the acknowledged state of the art, which is different than the normal, you know, definition of state of the wire, which we think of bleeding edge in regulatory space, recognized or acknowledged state of the art means what's common right now. And you have to at least do, what's the current standard of care. And it doesn't matter how expensive that's going to be for you, and how big you are, what your resources are, you can't come up with something less than what's already on the market. Right? That's, that's one of the barriers, you know, that might seem unfair. But on the other hand, you go find my patient. That's kind of what I want.
Aaron Moncur:Yeah. What are some of the common pitfalls that you see companies falling into? And how can they be avoided?
Steve Gompertz:Yeah, a lot of times, you know, or startups are often, you know, some bright engineers, they've got a brilliant technical idea. They don't really understand the quality and the regulatory requirements. And they're not thinking about the why and the how well they're thinking about how you're not thinking about the why and the what behind them. And so you get this sort of checklist mentality, just tell me what they want me to do, What documents do I need to produce? What activities do I need to perform? And you want to often see that, and that's how the quality system is built is just around sort of this checklist, do this, then do this to do this. And nobody's thinking about the reason why they're doing those things. And that's really, really important slides. Because it isn't just a matter of having these deliverables at the end of the day. It's how you got there. Right? That it should it's part of a process and the whole thing is a system and all the pieces have to work together. And so too often what I see is nobody's taking that process view that system view of their quality system, and it's just a pile of procedures and activities that are not really well connected. And then they can't figure out why is it not very efficient? Why does it not seem to produce the right results, it produces results, but we still have recalls that we still, you know, you know, run into problems where our tests fail, and it's slowing down our ability to get on the market. And it's because you didn't take a chance to step back and see the bigger picture. It's hard to see the whole system, right. But we often talk about, you know, companies will sub optimize, right, because they get focused on one process ago, make that perfect. And I was just on another webinar today, and I've been using this in an article I've got coming out soon, where I talked about it's kind of like, you know, I want my car to go higher performing and I'm driving an economy car. Putting racing slicks on the back of that economy car isn't gonna make it go any faster, or perform any better. Matter of fact, that might work against you. As a mechanical engineer, I could think of the stresses you're gonna put on the axle. wasn't the drive train? And will they fit in the wheel wells, things like that. And too often, that's what happens in quality systems is we don't take a look at the entire car, we look at this one piece that's maybe not working well, or we, that's the part we know. And we want to make that better. And that's the only thing we focus on. And then we're wondering, why doesn't that work with the rest of the system was because you didn't think about the system and connect it all together?
Aaron Moncur:So talking about the entire system? Obviously, quality control system is is a large, complex thing, or at least it can be, and the scope of our conversation today will not allow us to get into every detail of that. But if you could, can you kind of pick out like the main components and walk us through? What does it look like establishing a quality system. And earlier you mentioned that, you know, pre revenue, early stage development phase, you don't need the full system in place, but you need some parts in place. So So kind of walk us through the evolution of, you know, starting a new medical device design, and what what are the kind of the core elements we need to have in place? And at what points do we need to start putting those core elements in place?
Steve Gompertz:Yeah, so it's funny that I mentioned I was working on a webinar today, it was one that our company was putting out, and we were talking about all the foundational or fundamental issues behind all the standards and regs. And one of the core, you know, pieces of that are tenets is this process approach. And what that comes out, as you know, a lot of people are familiar with this concept of Plan, Do Check Act PDCA cycle, known also as the Deming short cycle. And that's what every quality standard and regulation is trying to tell you to do. You have to plan for what you're going to do, right? You You know, you have to then do it. Right execute to your plan, monitor what you're doing, get some data and check, and then act on that and make an improvement. Update the plan. Do it again, check it again, right. And it's the that's where continuous improvement comes in. That's the basis for every quality system is and that's why you need to see the connection points. Now getting started, right, yeah, you don't need all of the components yet. And maybe you're not ready to do, right. And maybe there's certain pieces of checker hack that aren't ready yet. But there are certain basics. We like to say in the regulated space. If it isn't documented, it didn't happen. So doc control and records control is one of the first things you put in place. First, you architect your system, you write a quality manual, and you say this is what we think our quality needs to look like, so that it's suitable and adequate for what we're doing. And then the next thing is we have to control documentation records have to be ready for that it's while we're going to design something, so then we need design controls, what is the design and development process going to look like? But we don't need to worry yet about production controls, we probably don't yet need supplier management until we start farming any work out. Because we're probably not going to build a factory right away, we're probably gonna bring in contract manufacturers. But that's what I mean the evolution right, first, get our stuff under control, does management understand its responsibilities? Do we know how to manage our documents in our records? Do we know how to do the right things in designing and developing our product or device that it's like, Okay, now we think we've got a design and we want to start building it. Now we need to engage contract manufacturers, component suppliers. So now we need our supplier management process in place. That's the next step. And maybe at some point, then we're going to go to full production, either ourselves or we're at our contract manufacturers, but we're going to do production controls, you know, and then you keep building those pieces we don't need. We don't need complaint management yet, because there's nothing on the market for people to complain about. Right. So don't implement that until you're actually a market. We don't need post market surveillance, because we're not on the market yet. But when I talk about the evolution, that's what it looks like. Cool some point once you're on the market than the entire system has to exist.
Aaron Moncur:Yeah, great summary. All right, I'm going to take a short break and share with the listeners that the being an engineer podcast is brought to you by pipeline design and engineering where we don't design pipelines, but we do help companies develop advanced manufacturing processes, automated machines and custom fixtures complemented with product design, and r&d services. You can learn more at Team pipeline.us. The podcast is also sponsored by the wave an online platform of free tools, education and community for engineers. Learn more at the wave dot engineer. All right, we have the privilege of speaking with Steve Gompertz, Gompertz today, and Steve can Steve can you share with us maybe a couple of the the key trends in medical device regulation that companies As should be aware of in the coming years.
Steve Gompertz:So like I said, I think the EU is going to drive a lot of the direction here. And one of the one of the big things that changed with the new medical device regulations that came out 2017 was more of a lifecycle view. If you look at the FDA regulations, and a lot of countries regulations, and what was in place in you, before, he was all about, how do I get on the market, and then there was a little bit what happens after the market, right, and then they started to go, Well, your responsibilities don't end, when you cash a check to somebody bought your device, you have a responsibility once that device is on the market out there in somebody's body, or being used over time. So they definitely added a lot more requirements around so yeah, post market. Now, here's a whole bunch of other things, you're nowhere near done with your responsibilities. And it's actually pretty complex now. And then, like I said, I think that's where everybody's gonna go, because they realize, yeah, these devices, they last longer, right? You know, when you get an orthopedic implant, now, they can easily last 20 years, you get a pacemaker and the battery technology keeps getting better. And the the intelligence in the firmware keeps you from overusing the, the available energy in the in the battery. So now, you know, the Pacemakers can easily last 1215 or more years, wow, that's a long time, you know, that that's going to be in somebody's body, you know, you need to account for, how are you going to deal with that future. And stay on top of how that device performs, because of course, it performs great when it goes out the door, and it's first implant their first use. But once that pacemaker is in somebody's body, what is it doing five years from now? Right? How do you prove that it will continue to do that, that will perform as designed for that time. So I think that's, you know, that that shift in thinking is a big deal. And, of course, the one that's you know, hot off the presses right now, of course, is AI and ML, every device, every segment of the device market is now affected by this. Devices that before, never saw, you know, a volt of electricity, now have sensors and software associated with them. Things like, you know, implantable orthopedic, orthopedic implants, were always purely mechanical, it was all about, you know, biocompatibility and mechanical stuff. Now, they're building sensors. So they can transmit information about your movements and your joints, to software. And that software is probably now going to have an AI element or machine learning element. So it can interpret what it's getting from you. We are starting right to see the device start to take on some of the role of the medical professional. And that definitely scares the pants off of the regulatory bodies. Machine learning in particular, ai ai is interesting. little short, aside here. AI is not new, we think it's new, because it's an all the press, but the term itself goes back to like the 1930s. And really any advanced, you know, or complex piece of software could be considered AI, right? It's the machine learning that makes the difference. And in the regulatory world, what they're used to doing is having you prove my device does this within these constraints. And they go yes, that's what we agree to we approve or clear you just for those things, what's called the intended use. And it stays that way. And if you want to change any of that, you have to go back through the process. Now you're telling us you're gonna give us device that on its own will change go outside those guardrails? The regulatory bodies don't know what to do. Yeah, yeah. And that's kind of the new frontier. It's all very exciting, because there's a lot of good that can come out of that. But like there's, it's it's stuff we don't fully understand yet. We see it if you play at all with any of the, the AI tools out there, and they talk about now they hallucinate, right, they come up with information that's not really reliable. I just saw a term slop, that they don't necessarily get it wrong, but the data they're basing it on is not particularly good, right? garbage in garbage out kind of thing. So, you know, the technology itself is not really advanced far enough. And we don't have control over it. And I think, you know, that's that's what worries, regulatory bodies. And I think that's the next big challenge here. The other one that's really a hot topic these days is cybersecurity. That's become a huge requirement with FDA and I'm sorry, my camera just went fuzzy there for some reason. But they're very serious now about making sure that if you're using any sort of software, with your device that you have addressed the security issues But regardless of whether it's connected to the internet, right, there's still the worry somebody walks up to it and gets into the software and changes things. Sure, yeah, that's become a big part of doing your submissions to the regulatory bodies, is demonstrating how you're dealing with cybersecurity.
Aaron Moncur:Going back to AI for a second, you may or may not have any direct experience with surgical robotics. But I think that's a really interesting area where AI is going to play a huge role, because now you have intelligence in these these surgical robots that can assist the physicians, the surgeons in making decisions, right, the robot might see something in the body that it connects with 87. Other things it knows about this data that it's been programmed with, and the surgeon might not necessarily make those connections, but the AI and the surgical robotics might, and then either suggest something or just do something. Do you have any, any anecdotes or any commentary about AI, specifically within the context of surgical robotics?
Steve Gompertz:Yeah, I don't know that. It's that it's pertains only to surgical robotics, I think that is a good example, right? The AI again, has great promise for things it can do, right? Because it can pull together pieces of information at a faster rate and see trends that the human might not pick up on. But in the end, it's not much different than being the surgeon themselves, right? Every surgeon doesn't operate at the same level of proficiency, and competency. They've all had different experiences based on their different, you know, patients, and what they know, is different, right? And then they do their best to share that information through articles and conferences, the way all of us do. I think the benefit is AI can at least put all that together and not lose it right? And actually, but it also then comes down to wait, what were the experiences, the AI learned from the same as if the surgeon only ever, you know, operated on white males? Right? What does it really know about the full patient population that that raw robot is going to be, you know, operating off of and the same thing as you feed the data? The training database to AI, diversity is becoming a big concern there. Yeah, garbage in, garbage out. So how do you make sure that the AI is seeing a broad enough spectrum to make these right decisions, and that the information is not being fed any bad information? Because they're certainly you know, I teach at the universities and I warn my students about even just getting up pick a pedia. Right? Forget about all the AI stuff they can do now. But if those reminders, uh, you know, you're just Googling and you're getting a list of white papers. Some of them are pretty awful white papers from they sound wonderful. And the people seem to have the right credentials. And they come to the wrong conclusion. Now you've got a artificial intelligence, and you're if you're letting it gather all of that information, even bad stuff, it's gonna make some bad decisions, right? There was, you know, the article just last week about Google. And its new AI. Right, the question about how can I make sure that the cheese stays on my pizza better? And it suggested glue? That's what I mean, right? And they went back and looked at it. And they said, because it picked that up somewhere where, for whatever reason, whether somebody was being sarcastic in a post somewhere, we're actually method. And the AI picked up on that one. Well, I saw it online, it must be true.
Aaron Moncur:Right? It's funny now when we talk about it, but not so much when the robot cut your spleen out unnecessarily.
Steve Gompertz:Exactly. Right. And that's, that's the challenge. And that's what the regulatory bodies are trying to figure out is how are we going to get control over devices that are utilizing this, there's a framework that's being developed around that they don't necessarily have the regulations yet, but they understand the types of controls that are going to need to be in place, you know, what it boils down to, we're not going to be able to allow device manufacturers to have machine learning based tools, you know, are devices that are just free to update themselves as they go. I always refer to it, they're going to have to phone a home, you have to get permission. So they can capture this information. And at some point, the manufacturer is going to be required to review. All right device, what is it you think you've learned? And then we through? No, don't use that? Yes, you can leverage that. And then we can go back until FTA. Here are the things we're going to let the device change.
Aaron Moncur:That makes sense.
Steve Gompertz:You have to still control your releases of the design changes.
Aaron Moncur:Sure. Right. All right. Well, can you talk to us about maybe a success story where your company QRx Partners has has helped a company improve their quality management system, their regulatory compliance? Maybe walk us through what that process looks like if you can, in the context of a story that you can share. Yeah,
Steve Gompertz:we had a client that was growing pretty quickly. And they were really outgrowing their quality system, what they had in there was pretty typical of when they were to start off, it was almost more of a facade. Right? That's checklist. Yes, we have a quality system. Yes, we have procedures for this. And they hadn't really thought about how to use them and why. And as the company was now growing, right, that was great. It was okay, when they were like 20 employees, and now there's 250, the whole thing is now caving in itself. And we're asked to revise, you know, revamp this quality system. And it was a complete rewrite, basically, we're gonna burn it down and start over thinking about the system, and then all the processes. Now what was indicative was, when we came on board, they had just come out of an FDA inspection, FDA left them with for inspectional observations, they were certified to ISO 1345, which the quality standard for medical devices. And so they went through annual audits from their certification body. And those audits typically produced about a dozen to a dozen and a half major findings. That sounds
Aaron Moncur:like a lot. Is that a lie? That's a lie. Yeah, that is a lie.
Steve Gompertz:And then they were doing some internal audits, and they weren't doing a great job with Earl, but even their internal audits, were identifying, like, 24 issues every year. Wow. Right. So that's a whole bunch of things that need to be fixed. And it's expensive. That's a lot of resources, right? It's a lot of time to investigate those, figure out the root causes, figure out the right solution. But the solution in place, prove this a solution work. And this is why non conformance is expensive. So we came in and said, Look, we're gonna burn the entire quality system down and rebuild a new one that is all connected, as I talked about, we're going to show you how the process interconnect. We're teach everybody about the why. Right? There's a there's a difference between education and training in an organization. I train you to follow a work instruction, step one, step two, but I have to educate you on why you're doing that. Right. Why does this regulation exist this way? Why doesn't regulation say it that way? Why have we designed it the proper way that creates awareness of what the risks are, right. And risk based thinking is a big part of that we built that into the quality system as well. But we built this much more holistic, quality system, where now everything is interconnected, we've done a better job of educating the employees on their role, and why they're, you know, doing things within the quality system, including employees who didn't think they were part of the quality system, like sales never sees themselves for the quality system, yet, they're a huge part of the quality system. And now you've got everybody truly engaged. This takes time, this didn't happen quickly. This took about three years to make this evolution, which happened to coincide with the next visit by the FDA. And they weren't zero findings. Wow. And over that three years, they got down to the most recent at the time, the FDA committed that the most recent certification body audit found one minor, non conformance. Amazing. Now internal. And we taught them to be more critical of themselves, they were still generating a lot of internal findings. But we were clarify classifying them as well. What kind of finding were they? Typically, when an auditor looks at as three things one, did you understand the requirements to? Are you actually following the procedure? And three, does that produce good results? You don't want number one, because it means you don't understand the requirements, right. And that was a thing, a lot of their findings were have that kind of first level that they weren't even understanding the requirements. Number two is not great, because it means people aren't really paying attention to the procedure. They're not executing to what's documented. But that's correctable. And then number three is kind of like this is where we get refinement and optimization. So you'd prefer that you find mostly in that third level. And that's where they got to so they were finding a lot more stuff on their own. But it was all typically at that second and mostly at that third level. Yes, we understand the requirements. We're never finding instances where we don't understand the requirements. We're almost always following that procedure as written. Or we know how to change it as needed, right if the conditions change because they will and now we're down to refinement of are we really as efficient and effective as we would like that process? Yeah, so that was a huge evolution for them. Yeah. When I don't have the numbers at hand, but I did a workup of the cost savings and it was hundreds of 1000s of dollars a year.
Aaron Moncur:Wow. Incredible. That makes it pretty easy for you to market right I mean, look what we
Steve Gompertz:cost the poor quality and it's just amazing. You know what that can cost? Yeah. And versus what it costs and it sounds like a lot it was my three years of work the burned out quality system. We do it we educate everybody but you that, that avoidance of several$100,000 a year, easily paid for. Yeah. This is why Philip Crosby, who was one of the quality guides, his book was quality is free. And that's what he was on. Yeah, the cost of poor quality is always going to be more than what it's going to cost you to get it right. And
Aaron Moncur:that maybe doesn't even account the new opportunities that may they may have been able to pursue because those resources were freed up to spend time on other stuff.
Steve Gompertz:Exactly right. Your resources are now not distracted with redoing things, right? Rework is always bad. This is a form of rework. I have to go in and correct or, you know, figure out new corrective actions for my processor, there should have been right in the first place. Yeah, those resources are not working on the things that would generate money. Yeah. Well,
Aaron Moncur:I imagine they're going to be folks listening to this episode who say to themselves, this sounds like an interesting path. I might like to pursue this or learn more about it, what do you think are some of the most important skills or attributes behaviors that quality management professionals need to have in the medical device industry?
Steve Gompertz:So that I talked about the system's thinking, right, you have to be able to see broadly and see the entire system. And that's hard, right? That's complex. And, you know, I always say it's seen as the proverbial sweater, you pull the thread, the whole sweater comes with you, that's what a quality system is like. But you have to understand, yeah, if I pull this part of the thread, what's gonna come along with it. So that being able to see from a systems perspective, and then you're always a problem solver. So excellent problem solving skills is going to be important. Because there is no, there's no point at which you declare success, and then go, we're done. We have the perfect quality system, some sort of condition, some variable is going to change tomorrow, and some part of your system is going to break, rules are going to change, the conditions are going to change or whatever. And then with problem solving, the most critical aspect of problem solving is root cause thinking. And if you don't actually get to the correct root, cause your solutions will never actually solve the problem. And that's the hardest one. And that actually, that comes back to did you even define the problem correctly? Right, so those are the kinds of skills can we get past? Can we get past the symptoms, and what I call the apparent causes, and see the system and figure out where the root cause actually was? Because the root cause is almost never where you observe the Pro.
Aaron Moncur:Interesting, that's great. All right. I think just maybe one more question here. And then we'll wrap things up. This is one I've been asking most of our guests recently. It's something that I'm very interested in what what is one thing that you've done to accelerate the speed of engineering? And most folks, when I asked this question, there's they're answering for from maybe more of a kind of broader, more more general context. But you and your experience that's very specific in the quality control world, I imagine, you might have some answers that are, you know, very specific to to quality, and regulatory, and how those things actually speed up the engineering process. You talked about how quality is free? Well, that's that's probably because it allows you to move faster, in some sense. But don't let don't let my comments influence you what, what's one thing that you've done to accelerate the speed of engineering,
Steve Gompertz:I would actually go further back in my career to when I was transitioning into the medical device industry, and I was still doing engineering automation tools. But now turning that more into project management techniques. And these were the early days of what we call then concurrent engineering. Right now, we always have multifunction teams, right cross cross functional development teams, back in the in the 90s. That was a new idea. And it was not obvious that there was a good thing to do, it was complex. This idea of like getting away from the throat over the wall engineering keeps engineering until they feel like they're done. And then they'll throw it to manufacturing and operations and get things done. Why would you want to involve them upfront, we're just gonna waste their time because we don't know what this thing is yet and what its gonna look like, now we know better, right? And so that right, because they're going to learn and they're going to help you come up with a good design that's actually manufacturable, serviceable, you know, whatever the downstream needs are going to be. And it will accelerate. But at the time, this was pretty controversial where I was trying to push for this idea of, we need to engage these downstream functions early in the process. It was it was a tough sell project managers. This kind of blew their mind because it meant that there was going to be work that they would throw away. It was going to be more iterative. Right. And so a lot of project managers just balked at it, it sounded kind of risky, because it wasn't very linear. And how am I going to manage my budgets and so trying to convince a VP of development that they should put their career on the line for this new idea was difficult. I eventually found a guy, the product development director, who had a small project, that wasn't really considered business critical. And he was kind of given carte blanche to figure out we do. And, you know, this was pacemakers, if a typical pacemaker budget at a time probably would have been on the order of three to $4 million, over about a three year period. And he was developed, he was supposed to be developing a simpler, cheaper pacemaker to go into underdeveloped countries where they don't have health care coverage and insurance, right? So how do you how do you deliver a pacemaker that's the same cost as a refrigerator, so that the family can decide? Do we get a new refrigerator or you buy grandpa off pacemaker, right? And so he had about a million dollar budget, and a year and a half to figure this thing out. And he had heard about some of my ideas, and I was demonstrating them with some of my IT projects of getting multiple functions involved. And he said, Yeah, I think I get this, I'd like to try it on this project. Can you help guide me through it? And so he was very open to taking the risks. And we talked about, you know, here's what the team needs to look like, right up front. And we're going to need to bring these people in, and there were people on in those downstream functions, who are waiting for these invitations. Right, they were really happy to hear you finally gonna let us talk upfront. Great. I want to be on that team. So we did that we ran it from everybody's in the pool at the same time, right? From the start, we're going to iterate Yep. You know, that didn't go the right direction, we're going to take a step back, oh, my gosh, the Gantt chart went the wrong direction. And then we're gonna go forward again, and back again, authority and right, and that kind of thing. And it sounds like that will be really wasteful. At the end of the day, the project took nine months, did successfully produce a device? And I don't think you use more than 600,000 of that budget.
Aaron Moncur:Wow, really? Huge saving.
Steve Gompertz:I got people's attention.
Aaron Moncur:Yeah. Do you think that there was something it was a fluke? Was it something specific about those people or that project? Or in your, you know, in your best judgment? Is that widely applicable?
Steve Gompertz:No, I don't think so. It certainly wasn't luck, or fluke, I think it was just, it's just natural, you start to bring in this cross functional team and these other areas of expertise. And the decisions just get better, because you're not narrowing your review. Right. Yeah, yeah. And so I think that's what they figured out. And that's why it's suddenly got really the norm. Because they've had, you know, this isn't particularly difficult to implement, isn't it? No, no, it's gonna be harder to manage your project managers have to get more comfortable with a little uncertainty in there. And management has to be forgiving of those iterations. But understand, we're gonna hit the finish line earlier, upfront, it may look a lot more chaotic. And you know, there's so much iteration going on, and it looks like it's taking too long. But all of a sudden, it's gonna get more and more efficient as it goes, and you're gonna get that a lot earlier. And now that's just how right now development is done.
Aaron Moncur:Yeah. Yeah. Wonderful. All right. Well, Steve, thank you so much for joining us today and sharing so much of your experience and wisdom with all of the listeners, how can people get in touch with you?
Steve Gompertz:The best way is through LinkedIn, either just through my name, or through my company, QRx partners, or there's our website QRxpartners.com. That pretty simple. But LinkedIn is usually our best approach. We like to encourage people to follow us. We do a lot of posting of really good, interesting information every day. Tuesdays is, I think, is the comedy days, we put a little funny stuff out there. But the other days of the week, it's it's useful technical information. Like today was the day we do our warning letter review, where we take a look at FDA warning letters and explain how not to get there.
Aaron Moncur:Mm hmm. That sounds tremendously valuable. Great. Well, Steve, thank you again, so much for for being on the show today.
Steve Gompertz:Yeah, thanks for having me. I appreciate the
Aaron Moncur:I'm Aaron Moncur, founder of pipeline design, and opportunity. engineering. If you liked what you heard today, please share the episode. To learn how your team can leverage our team's expertise developing advanced manufacturing processes, automated machines and custom fixtures complemented with product design and r&d services. Visit us at Team pipeline.us. To join a vibrant community of engineers online, visit the wave dot engineer. Thank you for listening
