The User in the Risk File: Bridging Medical Safety, Usability and Risk Management

Hello and welcome everybody. Sean Smith and We are here Friday in Focus, and I'm going to share some announcements.

Shauna, can you make me as a co-host, please, so I can share?

Yep. The problem, Fra, is that that's hidden somewhere. Make co-host. Close this… There it is. Confirm.

Love Zoom.

Thank.

People. Okay, so where are we? Upcoming events demystifying the FDA. That already took place. This is today. We have with us a speaker, Dr. Juan Dacasch, and I will wait till he's… sitting here, I don't see him. But he'll be with us in a second. So, what I want to talk about, is clinical project management, the hidden complexity of medical device and IVD studies is coming up, and… You're gonna want to join that event. because it's gonna be really, really good. This is the wrong presentation.

And I sent you the correct one on Slack, Sean, this morning.

Right.

Yeah, this is the correct one.

So, clinical project management, the hidden complexity of medical device IVD studies. So, one of the most, like, misunderstood areas of running a clinical study is the project management. What does that mean? What does that mean for, for clinical studies, and how. How can you go about controlling for the costs on that and understanding where they are and what it what it actually means? Maria can probably explain it a lot better than I can, and we'll let her do that later. So then we have your commercial training as part of the HF validation. This is fascinating. I have never seen a session about this anywhere.

That the idea of commercial training, particularly for complex devices, things surgical robotics, neurostimulation, that sort of thing, implantable devices, how much training needs to take place for the surgeons or the clinicians that are working with the product, and what part of that needs to be regulated Jerry Chasansky, And is regulated, and falls under the HF, the Human Factors validation. Think about that for a second. Think about if you're training people, and almost everything is not in the IFU that you're training people on. Next Friday, you get the pleasure of hearing from me, but I'm also inviting our MLV members to join this, because what are we doing here? We're live building

a referral program, and what does that mean for you? That means more referrals for you, and how to get referrals, how, like… like, I was thinking about it, and I was like, everyone out there wants more referrals. How do we get them? And that's what the session is about. And, then, the following Friday, we're gonna have. Mark, from a startup in Germany. They're doing some really amazing stuff. And it's basically… what would the notified body say about your submission, you know, if they got to see it and they could actually advise? And then, obviously, at the end of the month… oh, by the way, I'm leaving today for Europe. I will be.

traveling with my brother in Northern Italy, Switzerland, and then flying to Prague for this event. So, yeah, that's… It, for me, I'm gonna introduce today's speaker. Oh. Juan Dacacas and tell you that he is here to, talk about his experience as Medical Safety Officer at, for, for orthopedics and aesthetics, two companies, really interesting stuff, as his long background. But before we do that, we start with our usual question, which was, Juan, as a small child growing up, where?

Jeffrey A. Please tell us what you dreamed of being when you were an adult.

Wow. First of all, hi, everyone. I really thank you for joining us on a Friday. And Sean, thank you for this space. I think it's amazing that we get together. To answer your question, after I went through the basics of saying I wanted to be a fireman or policeman. I… Wanted to be a father and a husband. And that's where, everything… started to to come together. But apart from that little nice poetic thing, I always want to be an actor, but I could not tell my heavy Mediterranean dad that. So, I pursued some other career, and that career was medicine, because I always enjoyed being that guy with the first aid kit in the backpack when I was… when I was in school. So, that's a long-winded answer for your question.

Well, I always say that it's important to be a human first, and to tell us who you are as a human being, so that we understand that, and then you can share your brilliant mind with us, and your expertise, and everything that you know about Risk as a. Medical Safety Officer, someone needs to mute.

There… confirm.

If you're not speaking, please go on mute.

I understand.

Danny Crew, if you could please mute your, there you go. Thank you so much.

Beautiful, thank you very much. Okay, so sorry for that. But yeah, why don't you share your slides, and let's get started with this event.

Okay. Can you… let me know when you guys can see it.

Yep. All there.

All right, perfect. So should we get started? Yeah.

Let's get started.

All right. So once again, thank you guys for being here. I see some familiar faces that I've had the privilege of working with recently. Specifically, I'll shout out to Jonathan. So if you see him, he has been one of my latest partners in crime in aligning. My medical clinical experience with the importance of the user. We met through LinkedIn, and we have been working on couple of projects together. We're in Boston. We're writing an article that hopefully comes out soon about the same topic. So, if you see him, find him on LinkedIn. He's one of those people that are brilliant in what he does.

And I highly suggest you contact him for any Usability or Human Factors question. And I'm sure there are a bunch of you here in the line that I also wish to connect through LinkedIn. And at the end, you'll see my. my, my QR code, so you can scan it, and, I use LinkedIn a lot. Sean, thank you for this opportunity. I will be talking about how to incorporate what I have learned as a clinician, into, risk assessment. And how is risk assessment the bridge? Between pre-launch, post-launch, making it a true life cycle of a product. A little bit about myself. I am an orthopedic surgeon by training.

And I did joint reconstruction. which is hip and knee replacement, basically, and trauma surgery. The whole shebang, wearing scrubs, working at a hospital. In 2012, Depew Synthes recently formed. and they needed somebody in their Medical Affairs Department to join their joint reconstruction department. So that was what I like to call the Lego piece analogy, or Tetris analogy. I just… It was very weird, and it just fell right into place. I did not know what I was getting into, but I did, and I said, I'll do this for a couple of years and see where this goes, and here we are, 14 years later, two companies later.

So, I worked, for Depew Synthesia and Medical Affairs for 4 years, and then J&J, being a unicorn that it is, opened up a safety department all the way to the top, completely alienated from the commercial side or R&D side, which made sense to me, and I wanted to be part of that organization. So I applied for the job of being the global medical safety officer for joint reconstruction under the safety umbrella, and I got the job and worked there for 6 years. And that's where I was able to be part of The robot that started, at that time. being incorporated into J's and J's surgical platform for joint reconstruction.

And then everything has been, you know, very quick there and then. Four years ago, I decided to move away from the place that I used to live, which was Warsaw, Indiana. After 10 years, a lot of winters, and I got the opportunity to join an aesthetic company. Now, I know the initial reaction, I had the same one, how did you go from ortho to aesthetics? Well, it happens to be that aesthetic products also are Class 3 devices. So, everything that I had learned in J&J. from idea generation to post-market surveillance, and then some, was highly applicable, if not fully applicable, to the aesthetic business, because it's also a Class III device.

And there's also a user, and there's also risk management, there's also regulatory, there's also multiple regulatory pathways, etc, etc. So, recently, unfortunately, my job got terminated, due to a reorg in the company, and, and I am now, in between jobs, let's call it that way. So, if you know of anything, or anyone, or anybody that would. Like to have my background join their company, please let me know. Sean, I'm going to turn to the second screen so you can. take it away, Or do you want me to run the poll as well?

What kind of device do you work on? So let's let the audience go ahead and put in the chat, you know, as many numbers as you want or write NA, obviously, if, you know, does not apply to you. But we want to get an idea of like who's in the audience and, you know, who we're talking to. Right. Okay. Software. Advice, software, 3 and 5, home use devices, combination products, right? 4, software, 5, 4. 23453. Yeah. Okay. So it's good to know who's in the audience. You know, we we have other surgical instruments. Have not seen a one yet, but I did.

I did. I did see one.

Yes.

Judy, is it Julie or Judy? It went really quick. Ruthie.

Oh, that's okay.

Newman.

Yeah, so…

Luke.

Yeah, Jose has been all over the place or is all over the place. He does everything all at once.

Oh, yeah, he's one of the people I wanted you to meet.

Oh, awesome.

you Yeah, this is, this is great. So, yeah, you know, the reason for the poll is, you know, we want to understand who's in the audience, you know, what you care about, and, you know, if anybody has, you know, anything that they want to pipe up and say at the moment, go right ahead. But. Obviously, you know, so you can go to the next slide. People are.

Sounds good. All right, so… I'm sure you guys know that recently, FDA is doing something that maybe should have been done a long time ago. And it is framing. Risk management. towards the user of the product. Now. Whatever that means to you, the word user, is very important. Because this will expand your vision. And I will go through this a little bit further on the presentation. From, like, a forest to the trees and trees to the forest approach. When you're writing your, some people call it DFMEAs, some people call them risk assessment matrices, but basically it's your documentation of risk. So you should not be calling it URA, that I learned. It's the URRA.

Which is the use-related risk assessment. And… What it is, is… Just like your risk management, it's going to be a list of every task that a user performs. And if you have your worst-case mentality on, like I do from a safety perspective, which is not very fun as a dad, but as one that works with risk, it's very good to have, because I read in a book that whatever can go wrong will go wrong. So, the fact that you have it on your risk assessment doesn't mean it will go wrong. It means you thought about it. So this is great because it brings in the thought process into a document, regardless if it's going to happen when the product hits the market.

And… From a risk perspective, And fulfilling the definition of risk. You want to know what potential harms are going to happen from a basic risk management perspective from ISO 14971. You want to know, What is the harm? What is the severity? And from a risk management perspective, ISO 14971, the probability of occurrence. And when you do that, and when you work with your team. You know you're going to come up with a number, then you're going to come up with your mitigation strategies, and then based on those mitigation strategies, you're going to come up with a final number, and that's going to give you the stoplight analogy, red, yellow, green, and then you would do whatever you need to do to bring it to as low as possible.

So from your risk management perspective, have that idea in your mind and think about the use-related risk assessment, right? It's going to be the same discipline, but it's going to be applied to what people do with a device. When I look at complaints from devices, and I don't have pharma experience, it's only devices. You always see a classification if it's product dependent or procedure dependent. And that should tell you a lot. Procedure dependent is really the use of the product or the tools you're using in your intervention. Now, if you go to the far right and think about Usability Engineering, right, there is the IEC 62366-1 that Sean

Very recently showed us, you know, how Europe is trying to make standards a little bit more easy to understand, to cope with, to use. Making, obviously, the clarity that standards are not law. Right? They are standards, but a lot of people follow them because they are written in a way that makes sense. So thank you, Sean, for sharing that. And I highly suggest you get a chance to read that article that Sean posted on his LinkedIn page. Now, in here for Usability Engineering. I'm not an engineer, right? I have been on the field, I'm a user, and you'll see it when we get to it. But I have the privilege to work with engineers and have meaningful discussions, not to prove who's right or wrong.

But to come to common ground on how we can document the potential risk. of the device, inclusive of how it's being used, the environment, and everything that can go wrong around the use of the device. And that includes the experience and subject matter experts in Usability and Human Factors Engineering to design and cope with the formative and summative studies. So really, FDA describes this as a way to systematically identify use-related hazards, keyword. and to estimate the use-related risk. So for you guys that have worked on risk assessment, the word hazard should pop up and say, hazardous situation, and then, the risk.

Right? The output of severity and probability, and the probability split into P1 and P2. That is for risk management to your left. So, also, from a usability perspective, it is the document where you're going to bring in a cross-functional Collaboration between your human factors group. your engineering group, if it's not the same, because sometimes you have your designers there as well. Then your medical people that I'm gonna… we're gonna call them medical affairs, but it can be your medical safety officer, it can be your medical affairs person, it can be a clinician, it can be whoever wants that has experience in the field.

Then you're going to have your quality engineer, which is also very important, and you get together and you need to find the evidence to support what you're trying to say. Now… This has to be… For me. a bridge. between pre-launch and post-launch, or if you want to call it pre-commercialization and post-commercialization. I have learned that a lot of companies have a different team split in two, and those teams don't talk together. For me, that doesn't make sense because risk management has to occur in the full product life cycle. That's the key. It's a cycle. So maybe you see this as a line going from left to right.

But it has to go back. into the risk assessment. I have a way of saying this, which is. Pre-launch, Kind of. creates And post-launch updates. And this has to be done. Consensuously, and it has to be done with the evidence in the field. So obviously you're going to have a concept at the end. Somebody has an idea, somebody has a new product, somebody has a new anatomical location where they want to use your product, right? So you have to list what's going to change from what you have. Is something new? Are we acquiring a company? What do they have? What do they know? Right? And you have to understand, okay, this is a device, this is how it's used, where it's used, what can possibly go wrong?

Then you get together and say, all right, how are they using this product? And if they're using this product, you want to understand by observation, what are they doing right? What are they doing wrong? And this is where the formative studies come into play, and it'll come out with a lot of output. Now, the critical tasks have to be defined, right? And these are… Critical to quality, critical to safety, but you have to make sure that you define your critical tasks. And you have to mark them, right? based on your standard of what you consider could cause serious harm. Now, we all know the S12345, some companies go from 1 to 10, so understand and define what is your threshold. And don't make it fixed, because there are some critical tasks that may not go into that perspective.

from a clinical perspective, or from that perspective, from an engineering perspective. And these are where the meaningful conversations take place. And obviously, you do all the verification, validation, and then regulations, regulatory evidence, pre-launch, and then you get the green light, it's on the market. And a lot of people think, well, we're done. Nope. Safety starts here and post-market starts here. And this is where you have to make sure that your complaint handling unit, your case management, whatever it is that is feeding you. the real world evidence of the product inclusive of the use and adverse events secondary to use, right, are being brought into play.

Where do you document this? Well, you go to your risk management that you thought about. So I like to call the risk management to your left as a crystal ball scenario. Right? It's hard because you don't know exactly if it's a new product, what's going to happen. But you've got to bring your clinical and your engineering and all of your thought process to it and have those discussions. But then when you throw it out in the market, what is really happening? What did we see? What did we not see? What did we miss? What do we have to bring back in? Right, so once again, this is a living process, so your Usability Risk-Related Assessment has also

a living component of it, because it has to be updated as the new risks and hazards are discovered. Now, you were mentioning training previously, and I think this fits in really nicely with what we're going to see in a couple of slides. How do you think about training? What is normal training? When do you stop training? When do you include more training? When do you train the trainer and you're good with that, right? How do you make sure that the training is being done accurately, safely, following your instructions? So all of those questions are valid. Now. Those can go back to your user-related risk assessment.

this person has not retrained in X amount of months. What's the risk of that? What can happen? What's going to give you your answer? Well, what you find out in real-world evidence. I don't know if my good friend Sergio is on the line, but he and I worked in J&J, to… develop a program that is called trust, and that means the timely, relevant youth related safety training. Now, if Sergio is on the line, make sure you, you also find him. You can also ask me, but I haven't been part of that for a long time. But this is the safety related to the training, which is so important for our panelists that are going to be talking about this. I think this is a great topic.

Now… the URRA decides three things. The tasks that are critical, we talked about it, right? And the task that is not there, well, you can not mark it critical, because it doesn't make sense, right? It has to be there, and you have to figure out a way of marking it as critical, just like you mark the critical to quality aspects. Then you're going to have to bring in the results of what your summative study is testing, right? How you validate it from those critical tasks. And figure out if you have missed something or not, and bring that in. And then how much human factors really does FDA ask you to submit?

And this is important, because… A risk-based amount. is something that can range from a rationale and a summary of human factor activities to complete a report of validation. But then for higher risk devices and devices with critical tasks, FDA will typically expect the human factors process or the URA to have formative results, summative validation, and the evidence to support it. So the URRA really identifies the critical tasks, use-related risks, formative studies, improve the design, all of the validation that shows how intended users can perform a specific task, critical or not, safely in a real-world scenario.

Now, FDA expects evidence proportional to the risk, and posed by those potential use errors. So, if you go to the far right of your product. And you're finding XYZ findings on your real world evidence. FDA now, regardless of its product complaint, patient complaint, pharma complaint, whatever it is, they're going to backtrack all the way from that issue. to your risk management, and make sure that you thought about it. And they're gonna trace it, reverse engineer it, from the far right. to the left. So, it's a cycle, but still, you understand what I'm saying. Now, there are 9 aspects of your URRA, right? Now, you can… this is what FDA shows, but that doesn't mean it's the only ones that

you can have. If you find another column, if you find mitigations that you want to add, just like our risk management files, you're welcome to do so, right? But then you have Task ID, you have to you know label it somehow, right? Then you have the specific user task. What is a person doing with their device? Then you think about this is where that left is. expands in your Excel file, and you have different sublines, because it's a possible use error, right? How can things go wrong? Then, what is a hazardous situation? Which is important. Some people confuse, and sometimes I do the situation with the hazard itself. Right? The hazardous situation right is where the error

Puts the patient or user in harm's way. Then what is the harm? Well, we all know that. We quantify it. And then what is the severity of harm? And we defined, based on our standard. Is it a critical task? Then you have your risk control measures or your mitigation, your validation method, and that's really it. Now, some of you may say, wait. think something's missing. So… Which of these columns is not in the FDA? Now, I can't see the chat, Sean, but if you wanna… Go ahead and tell me what people are plotting there. It'll be great.

Yeah, go ahead and put that in the chat, what you think. 3, 2, 2, 2, 2, 2. who.

I think we got it. So yes, you are 100% right. The majority is right. And when I first saw this, I went, wait, hold on. Because I'm not used to working risk management from a usability perspective based on an FDA guidance. Now, this… we were doing this. We were including this, but we were going, like, wild, wild west on this. And we were adapting the risk management of our ISO, to Usability. And we were including the probability. And then I was very curious on why does the FDA Not bring in probability as part of the test. Because they define a critical task by how badly someone could be hurt.

Yeah. For the FDA, a critical task is a user task which is performed incorrectly or not performed. That would or could cause serious harm to the patient or user. where harm is defined to include compromised medical care. As every definition from the FDA, you can Attack this, interpret it the way you want it. But, basically what it's telling you… is that harm and severity columns are going to decide what you validate. And this is going to be. Clinical judgment? And a cross-functional discussion? On how to bring that in. I will go into why probability is not included. I still believe

That there could be a way. Of understanding the probability. But not using it in your calculations as you would do for normal risk. Because as you will see in the future slides, you have to avoid mitigation by math. And if you remember, risk is defined by the severity times the probability, and the probability is P1 versus P2, and we'll go through this. And you can really use the numbers you are thinking of the probability in use. to adjust those numbers in order to make your overall assessment look okay. And I think this is where FDA is going, in a sense of, don't use it. Maybe document that the probability is low, but don't use that number in your formula.

Now… Sean asked me to talk about a surgical robot. And… I'm going to tell you that in my over 20 years that I Decided to go into a surgical field. Inclusive of my 14 years in devices. I've had the fortune of living through Computer-assisted surgery. If you guys have been around, you know what I'm talking about. That went… and then went all the way down, because it took too long, you had to pin The patient invasive took too long. So I was part of that and it was painful because the setting up of the device took longer than the surgery itself. So right then and there was not going to work.

In ortho, Arthroscopy of the hip started. And I had a chance to see people train on this, and I still get palpitations on how they thought it was the same as a knee that has been going on for ages and it works, shoulder ages and it works. On the hip, it's different because it was new, so they were doing a lot of things that I was… really… Stressful about. And then robotics. Now, big difference here. Robot and cobot. Robot is doing, I'm going to say it very simply, right? You engineers, excuse me. Robot does what it does. A cobot helps you. Do what you need to do. The… Cobot that I worked with for J&J,

Is designed to place itself based on anatomical measurements at the position you want it to cut. But it does not activate the cut. You do it. You're the trigger finger. A robot would do everything and cut. A cobot positions. Makes you say, I'm ready, go for it, and then you do it, right? So it's basically you doing what the robot is suggesting. You can override it. Right? Which is actually a good thing. But let's go through a little bit of data. Now, look at the article at the bottom. This is a heavy-hitting article for robotic surgery, in which a lot of this information, plus the MOD database, which I'm not a fan of, but it provides data that most of us use as evidence, right?

In this article. Between the year 2000 and 2013, in medicine we would call this a classic article, it's 13 years old, you saw 10,624. Adverse events on robotic surgery. Now, Sean gave me that number. I said, oh, hold on, what else can we make out of this? Well, you've had 144 deaths. You had 1,391 patient injuries. You had 8,000 device malfunctions. Now, let's break this down a little bit. You might say, 1.4% deaths, that's nothing. If it's your mom.

If it's your friend? that dies in a robotic surgery, that's your 100%. So think about it. And this is why percentages for me are tricky, but they're used. So you have to be very clear on how you interpret them. And, you know, the injuries, the device malfunctions. Stopping at device malfunctions from my experience, if you're working on post-market surveillance. You have to make the correlation. Of which of the device malfunctions actually caused patient harm. So, if you have a complaint handling unit that is split. between product complaints and patient complaints, you have to make sure that down the line, they are coming together, and they're making the assessment that of those product complaints, these are the patient complaints. Now, that is what FDA is going to be looking for, and it makes complete sense.

All right. Robots. I think, make every use-related risk assessment problem a little bit bigger. Right? Six or more people around one system. Long training. The setup. As I was part of this robotic system, We went in. For weeks at a time, just to define. What is the best setup for the surgery? Now you can imagine you have to bring surgeons in. It can't be only me. You have to bring. Users on the field and say, oh, look, if you place it here, I can't move my shoulder. I can't use the drill. I can't hammer, right? The sterility of the system. How do you drape it? All of this has to be included in your thought process.

As I told you before, and I like to say it really loud because I'm proud of it, I'm a user or was a user. I know what it is to have those tools. I know what it is to have a reamer that doesn't work, a reamer that goes blunt, a drill that gets heated, drill bit that falls into the patient, all of this, right? I know when it is, when the light goes off in your OR, when an earthquake hits, and you're in the middle of the OR, and you might be saying, oh, that's fantastical. Well. Book recommendation, The Devil Never Sleeps, write that one down, because it will tell you a basic premise. That no matter if it happened once, it doesn't mean it's not going to happen again.

So, if it happened once, what are you doing so it doesn't happen again? Or… If it happens again, are you prepared for it? Now, for a robot specifically. The surgeon is only one of the users, right? You're going to have surgeons that are maybe, or, you know, other people that are using the console, or they're using the screen of the thing that's outside of the surgical field. You're gonna have your bedside assistant, your scrub nurse, the in and out of the OR, the circulating nurse, the anesthesia team. If you're in a student hospital, you're gonna have Double that by two, you're gonna have maybe a robotic coordinator from the company, you're gonna have OR staff, you're gonna have a bunch of people, right, that are there.

And those bring in their own risk. So think about all that. That's why you need somebody that has been a user. Now, Sean, let me know if I understood this correctly. But this little box that you're going to see throughout the next slides is called apply it to your device. So think about it. List every person who thinks about your product, or touches it, or is involved in the device when you get into your OR as a surgeon, or as an engineer, think what can go wrong in all of those areas, right? Now, remember this. I'm focusing on an OR, but devices are also in the hands or wrists of elderly population now, right? So, how do their cognitive and general know-how of using.

iWatch or a system that will tell you that your blood pressure is this and that. How are they understanding this? What do they need to do? What does a device do? Does it call 911 immediately and they're coming in and you see somebody drinking coffee? Right? All of these things are very important. You have to understand what is the device, who is the user. Right. My dad fortunately passed away two years ago, almost three, and he was one of these older tech savvy. I want to know everything about technology. But let's face it, he he wasn't. He just pressed buttons. Right. What is that going to do to the device? Right. Think about scamming. Right.

Why is there an elder population being scammed more? Because the use of technology, they believe in it. They think it's true. You have won a million dollars from the Prince of Congo. You've seen those emails. Well. Somebody can just press click. You press click, all of your information is gone. So think about that scenario of somebody, even elder surgeons now. that are 60, 70 years old, and when I say elder, listen. I already have my AARP card, so… Everything 50 above. I don't feel like an elder, but if you're way up there, 70 or 80, some of them are still operating. How does a vision? of their eyesight affected. So think about all this, right? Each group needs its own user profile. Now.

Task and User Errors. 1,557 reports in this same article described of a broken or burnt piece of instrument falling into the patient. In almost all of them, what happened next? So if you think that the surgery continued as planned and the team stopped, well. That's one. 2, the team stopped to search for the piece and retrieve it. That's 2. And 3, the patient was converted to open surgery. So let me give you an example. Let's say you're doing an abdominal surgery with a stapler. So if Sergio is here, he's going to relate to this. And all of a sudden, you're inside, and the stapler, for whatever reason.

The clamp at the end breaks off. And that little piece, False. You see it on the screen. Right? What do you think is going to happen next? Put it on the chat. One, two, or three. Sean, let me know what people think.

3, 2, 1… 3… 3, 3, 2, 2, 3. 3…

So we're getting a… we're getting kind of a mix between…

We got a four. Thanks for the four.

Is that 3 plus 1, maybe? So… that can't be…

Each one of these are what?

So here's, here's, remember this is based on that article down there, right? And I understand where you guys are thinking that the patient was converted to open surgery, but I will tell you the answer, right? The answer is two. And the answer is two based on this article, because your human nature is to retrieve it before you have to modify your approach. And… I will tell you one thing. When, for you guys that are not familiar with a joint reconstruction for the hip, you have the stem. And then you have the head. that in the majority of implants has to be impacted into the stem. And then you have the cup, which is the one that goes to the pelvic, the pelvis, the pelvic portion of the hip, right? And then you bring that in, and then you do the biomechanics, and that's it. Sometimes, this head

Falls into the patient behind the scan, well, under the scan, but behind the glutes because you're operating from that end and you think, well, that's an easy retrieval. No, it's not. That thing goes deep, and… You're literally in there for a while. Sometimes in the abdomen, you try to bring your scope and try to pinch it. If it's right there, everybody, nobody moved. You bring in your clips and you try to remove it. So the answer is two. It starts there, right? So a machine failure or a device failure creates a human task. Right? Now, look at the… look at the… like, the box on the right, and we're going to go back to what this Usability thing is all about. So…

The user task, right? Obviously, you have to… Check the instrument, the tips, everything before the surgery. Now think about what can go wrong, right? The people that do this are on shifts, they get tired, they get hungry, they get bored, right? Sometimes, you know, things happen, right? So what would you do? You maybe introduce a second verification like you do with technology, right? You have two sets of eyes. Oh, but that increases cost, that increases time. Yeah, it does, but… you gotta look at your data to see if that's actually going to work. So if anybody of you is a Six Sigma guy, I'm a green belt, think about the process, become the process, and learn to see in what areas of the process you can… you can…

Alleviate these potential things happening, but Include this in your URRA, right? The possible use error, right? The, you know, the piece.

Can't hear him at the moment. Can you hear him? -

I think you froze.

Postdoc. Post-op X-ray? Whoop! What is that? Can you guys hear me still?

We can now go back 30 seconds.

Can you guys hear me?

Yes.

Okay. Sorry. Internet stuff. So I'll go back to the example knee surgery. You're done, you close up the patient, and every surgery has a post-op X-ray. So the patient's in recovery room, the X-ray comes back. I think now they're waiting for the X-ray before they send the patient, but this is my old school way of doing things in a third world country, right? You couldn't wait for the X-ray to come back, you'd send the patient back. To the recovery room, and then you pull up the x-ray, or you see the x-ray, and there's a… You know, piece of metal that's not supposed to be there, so what do you do, right? Do you go in, do you take it out, is it going to be important, what's worse, is it worse to reopen the patient, is it actually going to make a difference, should we wait until the patient.

All of those scenarios have to be incorporated in the possibilities of the use of the device. So, what is the hazardous situation? What is your potential harm when you, you know, you decide to leave it? When you decide to extract it? The severity? Define it as a critical task. In this case. You'll make your decision. I think it's a yes, depending on, you know, my mind exercise. What are your risk controls? Right? And then how did you validate this? So when you apply it to your device, for each way it can fail, ask what the user has to notice and do next. And this is a task and it needs a row. So Six Sigma it, that's what I like to call it, become the process.

Draw your flow chart. I'm a drawer. I like whiteboards. And then I'll put them in Visio or whatever, but I like to draw and think about things that can go wrong. It really provides that tree approach so you can, that forest approach so you can focus on the trees. Now… when you launch the product, you have to track this, and you have to get your feedback. This is your feedback loop. This is where post-market surveillance and real-world evidence, from the field has to be incorporated. Now, if there's a non-conformance that you detect, if there is something that triggers a threshold, whether you have to escalate it, do an HHE, health hazard evaluation, and determine if you have to do a field action of some sort, right?

And this is very important. Now. With my friend Sergio, we've discussed, well, we need to incorporate this in our IFU, but then regulatory comes back and says, well, that requires a full submission. And then you go, so? And you go, well, that's gonna take 5 years, and we have to put the product on hold. So… there's a… there's a hiccup there, right? How do you document this to include it in your labeling without having to do another submission? Now, I'm not a regulatory expert, but I've been part of these conversations, and this is a subject… a full other subject, for another conversation, but think about it.

Now, the harm column needs someone who has actually seen the harm. Right? Somebody that has some type of relevance with the potential harm and severity that you are analyzing, right? So, the procedure was interrupted. What is the harm that relates to longer anesthesia? Is there a harm? Right? A longer open wound, right? Your infection risk has to be bumped at least one point, right? But then how are you going to substantiate all that, right? When you convert, for example, you're doing a robot, right? And you have this old-school surgeon that the hospital makes him use a robot because they bought this $1 million machine. He's not happy that he has to use it. He's using it.

He, as any other human, gets desperate like all of us do, gets frustrated, throws that thing away and decides to go manual, right? What does that imply? Look, 780 reports of this happening, right? So how does this get quantified if the FDA considers this as a harm? Electrical arcing. So for those of you that don't know, electrical arcing is the unintended discharge of electrical energy through the air across an unintended pathway, right? Could create a spark, could create high temperature, right? In robotic surgery, energy can jump from one place to another, and, and this is, this is important now.

Every time we have these discussions with engineers, I picture Star Wars lightsaber stuff, and it's not that way, right? You have to really have somebody, an expert, and tell you. how does this energy dissipate in the OR, and what is a real risk? Now, you may have burns, you may have radiation, you may have a bunch of things that you have to think about. If you reschedule surgery, because something happened and you can't operate that patient. Well, there is no harm specifically done. And some companies say no surgery, no harm. But the delayed treatment. Right? Or even in this case, they had put the patient under, and they couldn't complete, and they had to put it again. What does a second anesthetic procedure mean to the patient, and what is that risk?

So apply to your device. Make sure that harm and severity are informed, are reviewed with somebody that has been in the field with some experience, and it's knowledgeable of real world evidence. So briefly on probability, because this lives in the ISO 14971, and you have the P1 and P2. Some companies don't use this. This is recommended to be used. Why? Because they have two different definitions. P1 is a probability that you have a hazardous situation. P2 is a probability of that hazardous situation ending up in a harm. You multiply those together, that gives you a number, you multiply by severity, you get the risk. That is usually how it's done. But it's a careful use of numbers. So don't use your probabilities without substantiated evidence to

lower the product of what you're doing. So, you may have a severity of 5, But if you only use one P and say that the probability is very low, so you give it a 1, then you have a 5. But if you incorporate the hazardous situation and its probability, and then the probability of harm occurring secondary to that situation, you have a very different thought process, which is a little bit more robust. When you have risk controls, well, this is going back again to training, right? And I'm really glad to point out to that conference or panel that you're going to have in the future, Sean, if you guys weren't here, make sure you

you join in, because that's going to be very interesting, right? So, your first choice is always going to be that your product's going to be safe by design, right? It's going to be perfect. Nothing's going to happen. Well, that's not true, right? Something is going to happen. So, what are your protective measures if something happens? What are your mitigations that you're building in? I like the word pokey yoke, this. I mean, I heard that a lot from our engineers. They were coming from the automotive industry, and were working on ortho products in Indiana. They came from Michigan. So, all of these things that you do to mitigate this not breaking, this not turning, is

Fantastic. It's great. It's great learning for me that I never had that opportunity to learn about. And then. How do you label that and how do you train people and how do you inform them? And this is where we're talking about the you know, how does that get regulated now? Labeling and training for the FDA, they rely on the user to remember or refer back to the information, and sometimes labeling might be unavailable during use, and sometimes, let's, you know, address the elephant in the room. Some of us surgeons don't read the IFU, right? Because IFUs, when you have a surgical procedure. I'm gonna tell you…

The potential things that can go wrong. But sometimes in surgical procedures that use implants and instruments, you're going to have not instructions for use per se, but as part of your labeling, your surgical technique. Now, this is the coloring, this is a book with drawings that has how to use your stuff for this particular company. So think about it. I've been using X company for my joint reconstruction. My success rate is high. The hospital decides to cut the contract with that company, and now I have a new company. Oh, it's still a joint reconstruction. It should be the same. No, it's not. The tools are different. So what's the learning curve?

What's going to be the risk of that curve of your surgeon using a different company, right? It's going to have the same areas, macro, but the micros are going to be different. So how do you bring that in? So count the rows where the only control is IFU or training and try to define them a little bit better. Again, from this article. Field reports, really. By closing the loop, looking at post-market and evidence will tell you what you have missed. Let's face it, some people don't think of this as important. Some people just acknowledge the article in their post-market surveillance and say. We are reducing this as low as possible.

But look at these numbers. 334 system restarts in the middle of a procedure. Right? 780 conversions to non-robotic surgery. We went through that. And 270 procedures rescheduled. So as I was telling you before, right, remember the forest to the trees, right? you find the general perception, and this is also kind of like a funneling technique. What's your general area? What can go wrong in that general area? And this can be the prepping of the instruments the night before, right? Whatever you want to think about it. Then bring it concentrically in. And start defining those areas, and start defining those rows.

Now, you go from macro to micro, and every macro will have its own sub-micros that can also be sub-macros. I know it's a lot of words, but think about it. You got your Broader side of the funnel, what can go wrong, and you do all that work. Then you go to your next phase of surgery, right? So a different color, right? Okay, what can go wrong here and what can go wrong here? Be careful to make sure that those are connected because it's still the same procedure. Now I will tell you this. After my presentation in Amsterdam I was really Taken aback by how people with all of this AI influx believe in

the medical or health output that it provides. And this is very dangerous, right? A study done by J &J, my former boss's boss's boss, G .Joe James, presented, and he said that three out of 10 answers were actually accurate. Three out of 10. Right? That is very low for people that are using your Claude, your ChatGPT, your Copilot, your Geminis to ask them medical questions. We all do it, right? Hey, I have a pain, what can go wrong, right? If you don't know how to interpret the answer. you are introducing error. That is foreseeable misuse. Are you incorporating that in your, in your risk assessment?

Because people think that everything that comes out of technology is perfect. Well, or they know better. Well, that's that I think it's debatable. And… Kind of a take home message, right? When you go back and think about this URRA, right? Think about who is your everyday user. Remember the funneling down. Think about… What has failed and how had you, what would you need to do to recover, after that failure, right? Then. You bring in somebody in the field to have a conversation on the quantifying of harm, and it's not subjective, right? It has to be substantiated. Don't use mitigation by math. Make sure you have rows for training, and how are you going to define it? And remember about closing the loop, right? What is your post-market, real-world evidence telling you that you had foreseen in the beginning of the loop?

Now… Questions that we might not have time to go over, because we have 4 minutes, and I want to be very respectful of everybody's time, but please… Look me up on LinkedIn, and we can go through them, right? for your documentation, who should sign the severity column, right? Who should sign the full document? And when you sign the full document as the healthcare in the mix, are you acknowledging all of it or a certain bit? Or you have a percentage of that decision? How do you quantify this, right? So make sure you have that defined. when, you talk about training, how are you going to validate that, right? How are you gonna… how do you make sure that's documented?

If you're trusting AI, what can go wrong? Right? And think about it when you're creating your summative study testing for human factors. If you're incorporating AI because it's Everywhere, right? How are you going to document the use of it, and what can go wrong with it? When your complaint handling units or whoever receives your complaint gets it, what? What thresholds, or what triggers, for it to go back into your Usability Related Risk Assessment. And then, when a device has to be modified. How does that change your document and your human factors submission? And that is, all the… Sources that we, talked about today.

And then there's a little bit of a propaganda for me. I am open to my next role. That is my QR code for LinkedIn. I enjoy LinkedIn very much. Even… So that it's now got a little bit automated and commercial, and you… people think that you don't realize that it is an automatic response to your linking to somebody. But I am still myself when I answer through LinkedIn. I don't have a bot, I don't have AI on LinkedIn. I think it's a platform that really works. So… Let's keep in touch. And if you know of any opportunities, I'm open. So Sean, thank you for this space. I want to thank all of you guys.

For, for your time, especially on a Friday. I wish you all the best, and, enjoy your weekend.

He stole… he stole all my lines, so thank you all. I'm putting in the registration for next week. I want everybody to click through and register for next week, and then I want to show you this little trick that I just discovered. And I am… So, I am gonna go to LinkedIn. And… If we go to any event, on LinkedIn. So let's take the clinical project management one, and let's say that you're registered for this, view as member, right? There's a share button here. Once you've registered, what I want you to do is I want you to hit share, invite.

And then you can invite, as a viewer, up to a thousand of your first connections by doing this. You scroll over. You pick industry. From the industry, you pick medical equipment manufacturing. You show the results. This is going to show you your first connection results. You hit this button here. And you keep scrolling, and scrolling, and scrolling, and scrolling, and honestly, I just found this out this morning, that as a participant. So, here's what I'm saying. If you want to, invite people that you think are gonna be interested in Donowa's session on clinical project management, what you do is you go in and you cherry-pick from the list, right?

So this guy, clinical quality at Abbott, he should definitely be there. Right? You pick the people from your first connections, and you invite them. Now I want you to do that, for, Maria, but I want you to do that for me, and what I would love to see happen is for you to invite 5 people, just 5 people, to next Friday's call, who want to know what? They want to know. more events, so find the event that I put in want more referrals. So, what are we talking about next week? We're talking about US member.

So there's one attendee right now. I want to see that number go up while we're on the phone. Click on the link, say that you're gonna go, and then invite 5 people to this event. Why? Because the more people we have on the referral call, the more people can refer business to you. the more business you can refer to other people. Many of you are products and service providers in non -competing roles. Right? And did you know that there's something called a referral fee? And so, one way to boost your income in 2027 would be to have an active referral program. But who wants to build their own active referral program? Practically no one, right? It's time-consuming, but if you are part of one.

and you could refer business quickly and easily to other people within the network, that would make sense to you, I assume. So what we're going to talk about is referrals are a great source of business. Getting the referrals. That's another story. So how are you gonna get referrals? And how are people gonna validate you before they send you a referral? You know? And so what I've created, and I'll introduce this next week, is. A… audit for MedTech professionals, which basically asks, what does Google say about you? what is LinkedIn… what is LinkedIn… what does your LinkedIn say about you? What does your company website say about you? And are… and is AI going to refer you to people? Are they going to say, oh, you should reach out to

XYZ, who are they referring people to, and why? And so, if you run this audit that I've built on yourself, and I'll introduce it next week, it's gonna tell you, basically, what ChatGPT is telling people about you. And what it's doing and where it's looking and how it's assessing you and then. What, what we're talking about here is basically providing evidence. That you can do what you say you can do. It's a CER for you. It is a… it is an E-Star for you. As a MedTech professional. You say you're an expert because you have 30 years of experience, and how do I know that you're any good at what you do? Right? So how do you demonstrate your ability as a MedTech professional to the world and to AI, most importantly.

Right? And then how do you get people to refer business to you? So that's sort of what we're going to be covering next week. I hope that you'll join us, but I hope that you'll also invite the people that are responsible for business development within your organization to understand that the idea of getting referrals, all of this is changing. Everything about our business is changing. And that if you don't Adapt. You There's gonna be a problem. If there's not already a problem, right? And, so… I invite you to join in the conversation, I invite you to participate in the conversation, I invite you to, invite others to the conversation, but I think that this is a really, really important topic.

And it involves, you know, survival as a business, you know? So with that in mind, I thank you for staying over time. And hope… hope that you will tune in next week. I will be… Probably in Torino, Italy. with my brother on our vision quest, I will be visiting the Shroud of Turin. So, you know, immediately following that, you will see me. And, thank you, Juan, so much. For joining us. You know, if I were you and I were on this call, I'd be thinking, how can I get an hour of this guy's time to tell me, how I should, you know, get surgical robotic clients, get aesthetic clients, get clients in, you know, I mean, I would seriously be reaching out to him immediately and being like, I need to talk to this guy.

You know, and if you know anybody who wants to give him a job so that he doesn't have to move from North Carolina, please, please let him know his children will thank you. His. Family will thank you, and again, very happy to see all of you. Have a great weekend.

Thank you, Sean. Life emoji. Look, there you go. Bye.

Bye.