Latest / Let's Talk Risk! Podcast / LTR 165: Managing Characteristics Related to Safety

Transcript

LTR 165: Managing Characteristics Related to Safety · Let's Talk Risk! Podcast

• 0:00
  SPEAKER_01: Hi everyone, and welcome to Let's Talk Risk.
• 0:03
  SPEAKER_01: I'm your host, Naveen Agarwal.
• 0:04
  SPEAKER_01: Here we explore how leaders in medtech navigate uncertainty,
• 0:08
  SPEAKER_01: make better risk-based decisions,
• 0:10
  SPEAKER_01: and build safe and effective products to support sustainable growth.
• 0:14
  SPEAKER_01: Guys, today,
• 0:15
  SPEAKER_01: as you are settling down,
• 0:16
  SPEAKER_01: we are going to have an open discussion,
• 0:19
  SPEAKER_01: which means you can participate in this discussion if you like.
• 0:22
  SPEAKER_01: You don't have to, but you can join and we can have a very open discussion today.
• 0:28
  SPEAKER_01: We do not have a guest, unlike our previous shows.
• 0:31
  SPEAKER_01: Today, I do have a topic in mind to share with you.
• 0:35
  SPEAKER_01: It's something that I have seen in my practice, people struggle a lot.
• 0:39
  SPEAKER_01: There's a requirement in ISO 14971 to outline the characteristics related to safety
• 0:48
  SPEAKER_01: for your medical device.
• 0:50
  SPEAKER_01: And most people, I think, struggle with that.
• 0:53
  SPEAKER_01: The reason is That they follow the guidance in ISO-TR24971NXA where there's a list
• 1:00
  SPEAKER_01: of questions and they just find an answer, yes, no,
• 1:04
  SPEAKER_01: or maybe some description to those questions to the extent they are applicable
• 1:09
  SPEAKER_01: to their devices and they file them away.
• 1:11
  SPEAKER_01: And it doesn't really appear anywhere else in their RISC file.
• 1:15
  SPEAKER_01: So it's very hard to see.
• 1:17
  SPEAKER_01: How those characteristics related to safety are actually recognized and managed.
• 1:22
  SPEAKER_01: So what I have seen is people follow NXA and they just make a list,
• 1:26
  SPEAKER_01: type up the answers to those questions and file them away.
• 1:29
  SPEAKER_01: Now that is not the intent of this requirement in ISO 14971 that NXA is supposed
• 1:36
  SPEAKER_01: to be used only for guidance.
• 1:39
  SPEAKER_01: And why is that important?
• 1:40
  SPEAKER_01: It's important because if we don't have a good understanding and a full list
• 1:45
  SPEAKER_01: of characteristics related to safety,
• 1:48
  SPEAKER_01: How are we going to actually try to control them?
• 1:51
  SPEAKER_01: How are we going to try to control our risk, implement those risk control measures?
• 1:57
  SPEAKER_01: And that is where the biggest problem is.
• 2:00
  SPEAKER_01: So there is no linkage, no traceability.
• 2:02
  SPEAKER_01: So I'm going to talk about that today.
• 2:04
  SPEAKER_01: First of all, what are characteristics related to safety?
• 2:07
  SPEAKER_01: It's a term that is a little bit vague, but intuitively we understand.
• 2:12
  SPEAKER_01: So some of the examples could be, let's say, sterility.
• 2:16
  SPEAKER_01: Your device has to be sterile.
• 2:18
  SPEAKER_01: And that is a characteristic related to safety because if the device is not sterile,
• 2:23
  SPEAKER_01: you can have biological contamination and you can have infection if the device comes
• 2:27
  SPEAKER_01: in contact with the patient or there's a prolonged contact.
• 2:30
  SPEAKER_01: Another example could be that your device uses electrical energy for its therapeutic effect.
• 2:37
  SPEAKER_01: And the energy level could be too high or too low.
• 2:39
  SPEAKER_01: So that is a characteristic related to safety.
• 2:41
  SPEAKER_01: If you have a diagnostic device,
• 2:43
  SPEAKER_01: for example, accuracy of results is characteristic related to safety.
• 2:47
  SPEAKER_01: Inaccurate results can affect the diagnosis and timely treatment
• 2:52
  SPEAKER_01: or it can cause incorrect treatment.
• 2:55
  SPEAKER_01: And let's not forget the use environment.
• 2:57
  SPEAKER_01: For example,
• 2:57
  SPEAKER_01: you have provided information for safety and if those instructions are incomplete,
• 3:02
  SPEAKER_01: ambiguous or inaccurate,
• 3:05
  SPEAKER_01: then People can make mistakes even though they don't mean to.
• 3:09
  SPEAKER_01: They are not trying to make mistakes on purpose, but they are confused.
• 3:13
  SPEAKER_01: So these are some of the examples of characteristics related to safety.
• 3:16
  SPEAKER_01: Now, you have to make sure that to the extent possible, you define their acceptable limits.
• 3:25
  SPEAKER_01: So we were talking about electrical energy.
• 3:28
  SPEAKER_01: If the level is too high or too low, what does that mean?
• 3:30
  SPEAKER_01: That means there should be a minimum or a maximum limit.
• 3:33
  SPEAKER_01: And we should see that limit specified somewhere in your risk management file
• 3:37
  SPEAKER_01: and controlled appropriately.
• 3:39
  SPEAKER_01: This is what we mean by traceability.
• 3:41
  SPEAKER_01: So traceability goes two ways.
• 3:44
  SPEAKER_01: And I'm going to keep on speaking, guys.
• 3:46
  SPEAKER_01: If I don't hear from you,
• 3:47
  SPEAKER_01: you can share your thoughts and comments here in the chat,
• 3:51
  SPEAKER_01: or you can join me in the studio.
• 3:53
  SPEAKER_01: But I'm going to keep speaking on this topic today because it's an important topic
• 3:57
  SPEAKER_01: and people struggle with it.
• 4:00
  SPEAKER_01: So Traceability on two sides.
• 4:02
  SPEAKER_01: First, we need to connect these characteristics related to safety to hazards
• 4:07
  SPEAKER_01: and hazardous situations.
• 4:10
  SPEAKER_01: And then we also need to connect them to our design,
• 4:14
  SPEAKER_01: manufacturing, and even supplier controls through FMEAs.
• 4:18
  SPEAKER_01: So we should have full traceability in the risk management file
• 4:21
  SPEAKER_01: to these characteristics related to safety.
• 4:22
  SPEAKER_01: And generally,
• 4:24
  SPEAKER_01: generally speaking,
• 4:25
  SPEAKER_01: I know there are pockets of excellence in the industry,
• 4:27
  SPEAKER_01: but But generally speaking,
• 4:29
  SPEAKER_01: in my experience,
• 4:30
  SPEAKER_01: I have not seen that linkage properly established in the risk management file.
• 4:35
  SPEAKER_01: And it's just characteristics related to safety is another document somewhere hiding
• 4:41
  SPEAKER_01: in the risk management file.
• 4:42
  SPEAKER_01: And it's from that point of view, not very useful.
• 4:46
  SPEAKER_01: So we want to see linkage to hazards, hazardous situations and harms.
• 4:50
  SPEAKER_01: And you might be using a different instrument, something like a hazard analysis tool.
• 4:56
  SPEAKER_01: And you have to establish Limits to the extent possible.
• 5:00
  SPEAKER_01: If there are qualitative limits or quantitative limits, they can be done.
• 5:03
  SPEAKER_01: And those limits should be specified, for example, in your risk analysis documents.
• 5:08
  SPEAKER_01: And we also made the point that there should be linkage to your hazards
• 5:13
  SPEAKER_01: and hazardous situations and harm identification.
• 5:16
  SPEAKER_01: You might be using a different analysis tool.
• 5:18
  SPEAKER_01: That's fine.
• 5:20
  SPEAKER_01: And there should also be linkage to your design,
• 5:22
  SPEAKER_01: manufacturing, and supplier controls, for example, through the FMEs.
• 5:27
  SPEAKER_01: So the question you have to ask if you are practicing risk management in
• 5:30
  SPEAKER_01: the industry today, how are you managing those characteristics related to safety?
• 5:35
  SPEAKER_01: How have you documented them?
• 5:37
  SPEAKER_01: And can you trace those characteristics throughout your risk management file?
• 5:42
  SPEAKER_01: So let's talk about why that traceability is important, right?
• 5:45
  SPEAKER_01: First, we want to make sure that our risk analysis is complete.
• 5:51
  SPEAKER_01: And secondly,
• 5:52
  SPEAKER_01: if there is Some drift in those characteristics over time,
• 5:55
  SPEAKER_01: which will happen in the post-market phase during the life cycle.
• 5:58
  SPEAKER_01: When we get complaints in the post-market phase about device-related issues,
• 6:04
  SPEAKER_01: even if harm has not been reported,
• 6:07
  SPEAKER_01: we have to take a look at those characteristics and see
• 6:11
  SPEAKER_01: if our controls are drifting.
• 6:13
  SPEAKER_01: And if we are not able to have good traceability in the risk management file,
• 6:17
  SPEAKER_01: we are not going to be able to do that.
• 6:18
  SPEAKER_01: Okay, so we have,
• 6:19
  SPEAKER_01: I think with, I should be able to To bring you on stage, Vid, can you hear me?
• 6:25
  SPEAKER_01: Yeah, I can hear you, Naveen.
• 6:27
  SPEAKER_01: Okay, great, great.
• 6:28
  SPEAKER_01: Thanks for joining me, Vid.
• 6:29
  SPEAKER_01: Let's start by having you introduce yourself briefly, and then we'll continue our conversation.
• 6:34
  SPEAKER_00: Sure.
• 6:34
  SPEAKER_00: I work for Fresenius Gabby,
• 6:36
  SPEAKER_00: a senior principal quality engineer,
• 6:38
  SPEAKER_00: and work on disposables as well as infusion systems and some blood TCT products
• 6:44
  SPEAKER_00: under design controls.
• 6:46
  SPEAKER_01: Wonderful.
• 6:47
  SPEAKER_01: Thanks for joining me today.
• 6:48
  SPEAKER_01: We are talking about characteristics related to safety.
• 6:51
  SPEAKER_01: I hope you have an interest in this area,
• 6:53
  SPEAKER_01: but let me ask you by just inviting your thoughts on this topic.
• 6:57
  SPEAKER_01: How do you see this being done in your risk analysis?
• 7:01
  SPEAKER_01: What are some of the challenges you face or some of the best practices you have?
• 7:06
  SPEAKER_01: Just anything.
• 7:07
  SPEAKER_01: It's an open discussion.
• 7:17
  SPEAKER_00: Characterized in a certain way that we can readily use in our FMEAs or risk management files.
• 7:23
  SPEAKER_00: So the risk management file is not always current based on the data that we receive
• 7:27
  SPEAKER_00: from post-market.
• 7:29
  SPEAKER_01: Yeah, absolutely.
• 7:30
  SPEAKER_01: And this is the point I was making before is that if we don't have connections
• 7:33
  SPEAKER_01: and linkages in our risk management file from top to bottom, it's hard to link it.
• 7:38
  SPEAKER_01: Can you think of any example that you could share with us
• 7:41
  SPEAKER_01: where you have seen this play out in practice?
• 7:44
  SPEAKER_00: Yeah, certainly.
• 7:47
  SPEAKER_00: Many times in human factor studies or something related to software,
• 7:52
  SPEAKER_00: there are various defects associated with software,
• 7:55
  SPEAKER_00: with the usability engineering,
• 7:57
  SPEAKER_00: and all of them get lumped together the way we characterize them.
• 8:00
  SPEAKER_00: And certain usability use cases are different and have different risk severities,
• 8:07
  SPEAKER_00: but we all characterize them similarly.
• 8:09
  SPEAKER_00: So that's the only example.
• 8:10
  SPEAKER_01: Yeah.
• 8:11
  SPEAKER_01: No, I think that's a great example.
• 8:13
  SPEAKER_01: Like if we don't have that level of detail, And granularity.
• 8:17
  SPEAKER_01: And let's just tie it to our discussion that we are having today
• 8:21
  SPEAKER_01: from characteristics point of view.
• 8:23
  SPEAKER_01: If we want the user to take certain actions and they may be interfacing
• 8:27
  SPEAKER_01: with our device,
• 8:28
  SPEAKER_01: whether they are actually handling hardware components or parts,
• 8:33
  SPEAKER_01: or they are interfacing with the UI in terms of the software, right?
• 8:37
  SPEAKER_01: We are expecting them to take some actions.
• 8:40
  SPEAKER_01: Now, a lot of it depends upon our IFUs and safe information and training and whatnot.
• 8:46
  SPEAKER_01: And in many cases,
• 8:49
  SPEAKER_01: we have actually on the screen a menu-driven workflow that they are supposed
• 8:53
  SPEAKER_01: to follow.
• 8:55
  SPEAKER_01: And some of those steps may be critical.
• 8:57
  SPEAKER_01: So that is where we have to define these critical tasks in our user risk analysis.
• 9:02
  SPEAKER_01: And what I'm hearing you say is that we don't
• 9:06
  SPEAKER_01: So if a certain user action is a characteristic related to safety,
• 9:11
  SPEAKER_01: as we are talking about,
• 9:13
  SPEAKER_01: it should be identified very specifically and shown in the risk management file,
• 9:19
  SPEAKER_01: top to bottom.
• 9:21
  SPEAKER_01: I think that is a key point we are making.
• 9:23
  SPEAKER_01: Does that sound kind of reasonable to you?
• 9:25
  SPEAKER_01: Does it make sense?
• 9:26
  SPEAKER_00: Yeah, it does make sense.
• 9:28
  SPEAKER_01: So your example is a great example.
• 9:30
  SPEAKER_01: I think there are many characteristics related to safety that involve the user device interface.
• 9:35
  SPEAKER_01: Now, a point I want to make is the standard is actually quite clear on this.
• 9:38
  SPEAKER_01: The standard says that to the extent possible,
• 9:42
  SPEAKER_01: and I'm paraphrasing,
• 9:44
  SPEAKER_01: we should define the qualitative and quantitative limits for safe operation, right?
• 9:51
  SPEAKER_01: I've not seen that.
• 9:52
  SPEAKER_01: I've just seen people writing up answers to NXA questions and filing them away.
• 9:58
  SPEAKER_01: There is no min-max limit.
• 10:00
  SPEAKER_01: There is no, yes, no limit.
• 10:02
  SPEAKER_01: I mean, even sterility assurance, right?
• 10:04
  SPEAKER_01: You should be controlling your sterilization operation to a certain SAL.
• 10:09
  SPEAKER_01: And that limit should be specified as a risk control measure.
• 10:12
  SPEAKER_01: We don't do that.
• 10:13
  SPEAKER_01: So thinking about these tasks,
• 10:15
  SPEAKER_01: I think we should be challenging ourselves during the design phase.
• 10:19
  SPEAKER_01: What is the zone of safe operation?
• 10:22
  SPEAKER_01: And have we defined those limits in relation to the hazards analysis?
• 10:25
  SPEAKER_01: So I don't want to take too much time on this topic.
• 10:28
  SPEAKER_01: It was a topic I had in mind I want to introduce.
• 10:30
  SPEAKER_01: But since you are on stage,
• 10:32
  SPEAKER_01: I want to give you an opportunity to bring up a topic or question you have in mind
• 10:36
  SPEAKER_01: that we should talk about today.
• 10:38
  SPEAKER_00: I don't have anything at the moment.
• 10:41
  SPEAKER_00: Okay.
• 10:41
  SPEAKER_01: Okay.
• 10:42
  SPEAKER_01: That's no problem.
• 10:43
  SPEAKER_01: As we go along,
• 10:44
  SPEAKER_01: if you have any questions or comments, you can definitely bring them up right here.
• 10:49
  SPEAKER_01: Okay.
• 10:49
  SPEAKER_01: So let me touch upon that.
• 10:52
  SPEAKER_01: Why is it so?
• 10:54
  SPEAKER_01: Why am I making such a big deal about this?
• 10:56
  SPEAKER_01: We already talked about the post-market a little bit, right?
• 10:59
  SPEAKER_01: Let's take an example.
• 11:01
  SPEAKER_01: Let's say that you are getting reports of infection.
• 11:03
  SPEAKER_01: They are very vague.
• 11:04
  SPEAKER_01: You don't know what is going on.
• 11:05
  SPEAKER_01: You don't even know if it's related to serenity or it could be something else.
• 11:10
  SPEAKER_01: Is it related to kind of errors being made, use errors in the operating environment?
• 11:15
  SPEAKER_01: As we just talked about,
• 11:17
  SPEAKER_01: if you don't have those line items clearly specified in your risk management file,
• 11:22
  SPEAKER_01: it'll be very hard for you to estimate the rate of occurrence from
• 11:27
  SPEAKER_01: the post-market data and compare it to what you would expect and connect it back
• 11:33
  SPEAKER_01: to your manufacturing operation or the design.
• 11:36
  SPEAKER_01: So the connection to post-market surveillance is a problem.
• 11:40
  SPEAKER_01: If you don't have full traceability.
• 11:42
  SPEAKER_01: On the other end,
• 11:43
  SPEAKER_01: I have seen,
• 11:45
  SPEAKER_01: and actually there are several warning letters on this,
• 11:47
  SPEAKER_01: where suppliers which supply raw materials or services or components or parts
• 11:53
  SPEAKER_01: which are safety critical, they make changes on their ends.
• 11:57
  SPEAKER_01: And you have absolutely no idea what happened.
• 12:00
  SPEAKER_01: You don't have their FMEAs in many cases.
• 12:02
  SPEAKER_01: You have not integrated those FMEAs in your product FMEAs.
• 12:06
  SPEAKER_01: So the connection is broken.
• 12:09
  SPEAKER_01: Now, during the design and development, I think people are very diligent.
• 12:13
  SPEAKER_01: They do a lot of good work,
• 12:15
  SPEAKER_01: make sure suppliers are qualified,
• 12:18
  SPEAKER_01: but the drift happens when the product isn't in the marketplace.
• 12:22
  SPEAKER_01: Over time, it happens.
• 12:24
  SPEAKER_01: And as I mentioned,
• 12:25
  SPEAKER_01: there are actually a couple of warning letters already
• 12:27
  SPEAKER_01: where supplier-related changes did end up creating a safety scenario in
• 12:33
  SPEAKER_01: the finished device.
• 12:34
  SPEAKER_01: So that is a lifecycle risk management issue.
• 12:38
  SPEAKER_01: Where if we don't have traceability of these characteristics related to safety all
• 12:43
  SPEAKER_01: the way to our suppliers, we are actually quite vulnerable to those changes.
• 12:49
  SPEAKER_01: And I know you will say, fine,
• 12:51
  SPEAKER_01: I have it in my supplier quality agreement that they're supposed to inform,
• 12:54
  SPEAKER_01: but they still don't inform.
• 12:55
  SPEAKER_01: And look, we do the supplier quality audits and they are all compliant.
• 12:59
  SPEAKER_01: I hear you.
• 13:00
  SPEAKER_01: But supplier quality audits are focused on ensuring compliance to requirements,
• 13:05
  SPEAKER_01: not necessarily looking into their process capability or the change history
• 13:09
  SPEAKER_01: and whatnot, not to the level of detail.
• 13:11
  SPEAKER_01: And I feel personally that as we move into the world of QMSR
• 13:15
  SPEAKER_01: and FDA becoming extremely sensitive to risk-based controls throughout
• 13:20
  SPEAKER_01: the quality management system,
• 13:21
  SPEAKER_01: one particular area of vulnerability for all of us would be in the supplier controls
• 13:26
  SPEAKER_01: and the purchasing controls.
• 13:27
  SPEAKER_01: So that's another topic I want to talk about.
• 13:30
  SPEAKER_01: And if you guys, again, have any questions or comments, you can write that in the chat.
• 13:35
  SPEAKER_01: Okay, looks like we do have a question here.
• 13:38
  SPEAKER_01: Thank you for this session.
• 13:39
  SPEAKER_01: I have a more general question.
• 13:40
  SPEAKER_01: What advice do you have for an entry-level quality or regulatory person?
• 13:45
  SPEAKER_01: How can they get trained on risk management aspect of medical device?
• 13:49
  SPEAKER_01: So the best way to deal with that is, in my mind, almost like a 60-40 ratio.
• 13:56
  SPEAKER_01: You know, 60% should be on the job training.
• 13:59
  SPEAKER_01: Find a project that gets you involved in the risk management process,
• 14:03
  SPEAKER_01: even though it may not be directly in your area of responsibility.
• 14:07
  SPEAKER_01: Ask around because how can you get involved?
• 14:09
  SPEAKER_01: If somebody is working on creating these risk management files and documents, get involved.
• 14:14
  SPEAKER_01: You will learn on the job what the problems are.
• 14:18
  SPEAKER_01: Do more of it if you are interested.
• 14:21
  SPEAKER_01: The remaining 30-40%, I think you should invest in a little bit of training.
• 14:26
  SPEAKER_01: And I know there's a lot of training out there.
• 14:28
  SPEAKER_01: I also offer some training on my website.
• 14:31
  SPEAKER_01: I'm not going to pitch a lot of that here.
• 14:33
  SPEAKER_01: If you are interested, you can ask me, you can reach out.
• 14:36
  SPEAKER_01: But you should invest in your training.
• 14:38
  SPEAKER_01: Now you will say,
• 14:38
  SPEAKER_01: you know,
• 14:39
  SPEAKER_01: I don't have time and it takes a lot of money and my manager is not able
• 14:44
  SPEAKER_01: to sponsor me.
• 14:45
  SPEAKER_01: But look, I know this is an important topic.
• 14:47
  SPEAKER_01: I want to build my own competence.
• 14:49
  SPEAKER_01: So I would say find something that is affordable.
• 14:53
  SPEAKER_01: And that is useful and practical.
• 14:54
  SPEAKER_01: And go for it.
• 14:56
  SPEAKER_01: Invest in yourself.
• 14:59
  SPEAKER_01: So here's a philosophical perspective I want to share with you guys.
• 15:02
  SPEAKER_01: Imagine a company,
• 15:03
  SPEAKER_01: a medical device company that you work for,
• 15:06
  SPEAKER_01: reduces their R&D budget to, let's say, almost zero.
• 15:10
  SPEAKER_01: They say, we're not going to do any R&D anymore.
• 15:12
  SPEAKER_01: Will they be in business?
• 15:14
  SPEAKER_01: I don't think they'll be in business anymore.
• 15:15
  SPEAKER_01: Four, five, 10 years down the road, they'll lose all the competitive edge.
• 15:19
  SPEAKER_01: We have to look at our own career the same way.
• 15:22
  SPEAKER_01: If our manager is not supporting us, invest in yourself.
• 15:27
  SPEAKER_01: You know, set aside some funds every year.
• 15:29
  SPEAKER_01: I know budgets are tight and you have so many other things to take care of.
• 15:34
  SPEAKER_01: But think of it as investment in your own career.
• 15:37
  SPEAKER_01: So I know I went on a tangent, guys, and these conversations are meant to be like that.
• 15:42
  SPEAKER_01: They are meant to be very informal and casual.
• 15:45
  SPEAKER_01: And I hope you get value out of them.
• 15:47
  SPEAKER_01: So let's get back on track.
• 15:49
  SPEAKER_01: Okay, let's talk about risk management.
• 15:50
  SPEAKER_01: Again, I want to open the floor.
• 15:52
  SPEAKER_01: If you guys have any other questions or comments you want to raise, please bring them up.
• 15:56
  SPEAKER_01: And Vid, again, since you're on stage, you are welcome.
• 16:00
  SPEAKER_01: I have
• 16:02
  SPEAKER_00: one question.
• 16:03
  SPEAKER_00: So when we do the risk assessment rate for any particular defects or hazard
• 16:09
  SPEAKER_00: that we have come across for medical devices,
• 16:11
  SPEAKER_00: which are repeatedly used with disposables,
• 16:15
  SPEAKER_00: how do you, what's the correct way to characterize the PODME?
• 16:19
  SPEAKER_01: Can you repeat, what's the correct way to characterize what?
• 16:22
  SPEAKER_00: The PODME, the probability of defect malfunction error rate.
• 16:26
  SPEAKER_01: Okay.
• 16:26
  SPEAKER_01: So probability of occurrence.
• 16:28
  SPEAKER_01: Yes.
• 16:28
  SPEAKER_01: Wonderful.
• 16:29
  SPEAKER_01: So If I understand your question correctly,
• 16:31
  SPEAKER_01: you are asking that you have,
• 16:33
  SPEAKER_01: let's say, a device that is used repeatedly using disposables.
• 16:37
  SPEAKER_01: And you have to characterize the probability of occurrence
• 16:40
  SPEAKER_01: of maybe device-related defects or maybe harms being reported to you.
• 16:44
  SPEAKER_01: What is the correct way?
• 16:45
  SPEAKER_01: Did I understand your question correctly?
• 16:48
  SPEAKER_00: Yes, more so on the defect.
• 16:50
  SPEAKER_01: Okay.
• 16:50
  SPEAKER_01: So the best way to do that is go back to the basics.
• 16:54
  SPEAKER_01: The basics teach us what is the number of defects per opportunity.
• 17:00
  SPEAKER_01: Per million opportunity, per 10,000 opportunity, 100,000.
• 17:04
  SPEAKER_01: So what is an opportunity in this case?
• 17:06
  SPEAKER_01: An opportunity is a use.
• 17:09
  SPEAKER_01: If a device is being used once per patient,
• 17:14
  SPEAKER_01: but let's say 5,000 times a year,
• 17:16
  SPEAKER_01: you have 5,000 opportunities for a defect to surface.
• 17:20
  SPEAKER_01: Does that make sense so far?
• 17:23
  SPEAKER_00: Yes, that makes sense.
• 17:24
  SPEAKER_01: Okay.
• 17:25
  SPEAKER_01: So the next question would be, Naveen, how do I find these opportunities?
• 17:28
  SPEAKER_01: How do I calculate?
• 17:29
  SPEAKER_01: I don't know.
• 17:30
  SPEAKER_01: I mean, I don't go to the clinics and ask how many times they are using the device.
• 17:34
  SPEAKER_01: You have to estimate to the best of variability.
• 17:37
  SPEAKER_01: And one way to do that is track the number of disposables being used.
• 17:41
  SPEAKER_01: So for example,
• 17:43
  SPEAKER_01: if you have a machine that uses a certain disposable part only once,
• 17:46
  SPEAKER_01: you can make some assumptions.
• 17:48
  SPEAKER_01: Say, if I sold 10,000 parts,
• 17:50
  SPEAKER_01: 10,000 of these disposable parts,
• 17:52
  SPEAKER_01: and I have X number of machines out there, what is an average usage rate?
• 17:58
  SPEAKER_01: That average usage rate could have a 95% confidence interval one way or another.
• 18:03
  SPEAKER_01: And you can say, hey, this is my zone of number of opportunities.
• 18:07
  SPEAKER_01: You can be conservative, you can be relaxed, you decide.
• 18:11
  SPEAKER_01: But you must estimate the number of opportunities applicable for probability of occurrence.
• 18:18
  SPEAKER_01: Does that make sense, Vid?
• 18:19
  SPEAKER_00: That does make sense.
• 18:20
  SPEAKER_00: So that's how I'm used to.
• 18:23
  SPEAKER_00: I was just curious if there was any other way we could also No,
• 18:27
  SPEAKER_01: so this is the way, but it's not without its problems, right?
• 18:31
  SPEAKER_01: Because if you ship 10,000 disposable parts, not all of them are being used.
• 18:36
  SPEAKER_01: Some of them are sitting on shelf.
• 18:38
  SPEAKER_01: Your machine,
• 18:38
  SPEAKER_01: if it can be used with a disposable part made by someone else,
• 18:42
  SPEAKER_01: they are probably using it.
• 18:44
  SPEAKER_01: So you don't know, you don't have that information.
• 18:47
  SPEAKER_01: But here's what I say.
• 18:48
  SPEAKER_01: You do the best you can to come up with the best estimate.
• 18:53
  SPEAKER_01: Knowing fully well that it'll have an uncertainty in your estimation,
• 18:56
  SPEAKER_01: and you document your assumptions,
• 18:58
  SPEAKER_01: and you test those assumptions using post-market data,
• 19:02
  SPEAKER_01: I don't know,
• 19:02
  SPEAKER_01: every quarter,
• 19:03
  SPEAKER_01: every six months,
• 19:04
  SPEAKER_01: every 12 months, and you can say, is our average rate of use still valid?
• 19:10
  SPEAKER_01: That's the question you're asking, and you need to refine that.
• 19:13
  SPEAKER_01: You cannot just estimate once and forget about it.
• 19:16
  SPEAKER_01: Okay, great question, Vid.
• 19:17
  SPEAKER_01: Thank you for bringing that up.
• 19:18
  SPEAKER_01: It is a big challenge, and Very relevant.
• 19:21
  SPEAKER_01: You know why?
• 19:22
  SPEAKER_01: Because I read warning letters as a hobby and you can tell,
• 19:25
  SPEAKER_01: you know, there's something wrong with that.
• 19:28
  SPEAKER_01: I have a sense over so many years having read all these letters that in general,
• 19:35
  SPEAKER_01: there's a perception that risk is being underestimated on part
• 19:39
  SPEAKER_01: of regulatory authorities.
• 19:40
  SPEAKER_01: I have a feeling.
• 19:41
  SPEAKER_01: It's not data.
• 19:43
  SPEAKER_01: It's an opinion.
• 19:43
  SPEAKER_01: It's a feeling.
• 19:45
  SPEAKER_01: But an informed feeling that in general, there's a sense that we are underestimating risk.
• 19:50
  SPEAKER_01: Whether through probability of occurrence calculation,
• 19:52
  SPEAKER_01: in fact,
• 19:52
  SPEAKER_01: there's a very specific warning letter on that to a big manufacturer,
• 19:56
  SPEAKER_01: or through severity assignment, and there are warning letters on that too.
• 19:59
  SPEAKER_01: So risk can be underestimated by both,
• 20:03
  SPEAKER_01: either underestimating severity of the outcome
• 20:05
  SPEAKER_01: because we don't understand hazardous situation,
• 20:07
  SPEAKER_01: or underestimating the probability of occurrence because we use
• 20:10
  SPEAKER_01: a very large denominator.
• 20:13
  SPEAKER_00: We don't use the proper denominator.
• 20:15
  SPEAKER_01: And we'll have to, I think, get better at that.
• 20:18
  SPEAKER_01: Really, it's an opportunity for us to make a difference in our organizations,
• 20:23
  SPEAKER_01: think differently, and apply different things.
• 20:27
  SPEAKER_01: Great.
• 20:28
  SPEAKER_01: Anything else on your mind?
• 20:30
  SPEAKER_01: Since you're on stage, you're kind of a little bit visible to the crowd, but you don't have to.
• 20:33
  SPEAKER_01: I don't want to put you on the spot.
• 20:34
  SPEAKER_01: But if there's any other question or comment you have,
• 20:37
  SPEAKER_00: we still have some time.
• 20:38
  SPEAKER_01: No, I don't have one.
• 20:41
  SPEAKER_01: Okay, okay.
• 20:42
  SPEAKER_01: Okay, wonderful.
• 20:44
  SPEAKER_01: All right, guys, I think we do have a comment here from Eric.
• 20:48
  SPEAKER_01: That is a great perspective on continual improvement, continual learning.
• 20:51
  SPEAKER_01: Yes, Eric, thank you so much for sharing that perspective.
• 20:55
  SPEAKER_01: And honestly speaking,
• 20:57
  SPEAKER_01: if you think about it,
• 20:58
  SPEAKER_01: and if you're smart about selecting the training solution,
• 21:02
  SPEAKER_01: it doesn't have to cost that much.
• 21:04
  SPEAKER_01: And remember, training is not just Passing the exam and getting the certification.
• 21:09
  SPEAKER_01: Ask the question whether the training is actually helping you becoming confident
• 21:13
  SPEAKER_01: in understanding the basic principles and applying them correctly.
• 21:17
  SPEAKER_01: So it doesn't stop.
• 21:19
  SPEAKER_01: So let me,
• 21:19
  SPEAKER_01: while we have a few minutes here, wait, again,
• 21:21
  SPEAKER_01: you can stop me if you have any other questions or comments,
• 21:23
  SPEAKER_01: but I'm going to share with you guys a couple of quick pieces of information.
• 21:29
  SPEAKER_01: This is the reason why I do Let's Talk Risk.
• 21:32
  SPEAKER_01: Let's Talk Risk is not a training platform.
• 21:35
  SPEAKER_01: Risk is a knowledge sharing and a community building platform.
• 21:39
  SPEAKER_01: And if you're new to this, let me tell you a little bit more about that.
• 21:41
  SPEAKER_01: I started this maybe three years ago.
• 21:44
  SPEAKER_01: And the intent here is for everyone in the industry,
• 21:48
  SPEAKER_01: no matter what our level is,
• 21:49
  SPEAKER_01: what our experience is,
• 21:50
  SPEAKER_01: to become more confident in applying these risk management principles.
• 21:55
  SPEAKER_01: And how do we do that?
• 21:56
  SPEAKER_01: We do that through conversations like these.
• 21:58
  SPEAKER_01: These are very informal conversations that If you are new to this forum,
• 22:02
  SPEAKER_01: you can check our website and download the podcast that we have recorded over 160
• 22:07
  SPEAKER_01: of them.
• 22:07
  SPEAKER_01: Conversations like these with industry leaders and everybody very casual way,
• 22:11
  SPEAKER_01: not a webinar or a lecture.
• 22:13
  SPEAKER_01: We also do that by publishing a lot of practical content on the Let's Talk Risk website.
• 22:19
  SPEAKER_01: And if you're interested, you can reach out to me.
• 22:21
  SPEAKER_01: We can share the link with you.
• 22:23
  SPEAKER_01: It's almost nearly free.
• 22:25
  SPEAKER_01: Some of the content is behind the paywall.
• 22:28
  SPEAKER_01: And that is just for me to find some support to be able to do this.
• 22:32
  SPEAKER_01: So if you so desire, you can support me through a paid subscription.
• 22:35
  SPEAKER_01: And that has some other features, but really 70, 80% of the content is free.
• 22:41
  SPEAKER_01: My intent is for you to learn from it.
• 22:43
  SPEAKER_01: And what is that content?
• 22:44
  SPEAKER_01: That's content are case studies, thought-provoking questions, opinions.
• 22:48
  SPEAKER_01: I publish a Think About It series every Monday.
• 22:52
  SPEAKER_01: And all these things that we talked about,
• 22:54
  SPEAKER_01: safety characteristics,
• 22:55
  SPEAKER_01: traceability, supplier controls, they have appeared in the Think About It series.
• 22:59
  SPEAKER_01: And that is meant to inform you and also encourage you to think about these questions.
• 23:04
  SPEAKER_01: We are also going very heavy on QMSR-related topics.
• 23:08
  SPEAKER_01: I know this is going to be a very big topic already.
• 23:11
  SPEAKER_01: FDA has cited many,
• 23:12
  SPEAKER_01: many manufacturers,
• 23:14
  SPEAKER_01: citations from 13485, and many of them are explicitly risk-related in Clause 7.1.
• 23:21
  SPEAKER_01: So keep an eye on that.
• 23:22
  SPEAKER_01: Even if you feel like your paperwork is rock solid,
• 23:26
  SPEAKER_01: I would ask you to be more diligent about managing risk.
• 23:29
  SPEAKER_01: So that is what Let's Talk Risk is all about.
• 23:32
  SPEAKER_01: It's all about knowledge sharing and community building.
• 23:35
  SPEAKER_01: If you are interested in participating, please reach out.
• 23:38
  SPEAKER_01: If you want to be a guest on this show,
• 23:40
  SPEAKER_01: it requires no preparation,
• 23:42
  SPEAKER_01: just your willingness to share your opinions and perspective.
• 23:47
  SPEAKER_01: And there's no right or wrong answer to anything.
• 23:49
  SPEAKER_01: We don't This is not a webinar or lecture.
• 23:51
  SPEAKER_01: You're not going to be showing slides or making a presentation or being worried
• 23:56
  SPEAKER_01: about whether your perspective is right or wrong.
• 23:59
  SPEAKER_01: Your perspective matters.
• 24:00
  SPEAKER_01: So raise your hand and say, Naveen, I'm interested.
• 24:03
  SPEAKER_01: And we will have a chat, 30-minute chat to see if that would work for you.
• 24:07
  SPEAKER_01: And you can come on the show to share your perspective.
• 24:11
  SPEAKER_01: Wait, final call.
• 24:12
  SPEAKER_01: Any other questions, comments before we close?
• 24:14
  SPEAKER_00: No, I think this was a very good discussion and I would like to be involved going forward.
• 24:20
  SPEAKER_00: I think it's on every Friday, right?
• 24:22
  SPEAKER_01: Every other Friday,
• 24:23
  SPEAKER_01: we meet at 11 Eastern and we took a little bit of a break last month or so,
• 24:28
  SPEAKER_01: but from now on, we'll have a lineup of speakers every other Friday at 11 Eastern.
• 24:33
  SPEAKER_01: So yeah, if you are interested, let me know.
• 24:35
  SPEAKER_01: You can reach out to me via LinkedIn or any other way, quite visible there.
• 24:40
  SPEAKER_01: So guys, first of all, I want to thank all of you for joining.
• 24:43
  SPEAKER_01: Thanks, Vid, for joining me here on the studio.
• 24:47
  SPEAKER_01: I really appreciate you participating and sharing your thoughts and questions.
• 24:50
  SPEAKER_01: I want to close with by sharing with you another resource that you may not be aware.
• 24:54
  SPEAKER_01: I have created what I call a thinking partner using ChatGPT's platform.
• 24:59
  SPEAKER_01: It is not going to give you a yes, no answer.
• 25:02
  SPEAKER_01: It is designed to use my knowledge sources based on over 10 years
• 25:06
  SPEAKER_01: of practical risk management experience to navigate,
• 25:09
  SPEAKER_01: to help you navigate through a risk-related question or topic.
• 25:13
  SPEAKER_01: It's a thinking partner.
• 25:15
  SPEAKER_01: If you say, hey, LTR Risk, the name is LTR Risk Coach.
• 25:19
  SPEAKER_01: If you say, Risk Coach, give me an FMEA template, it is not going to do that.
• 25:23
  SPEAKER_01: But it's going to ask you what it is that you're trying to do.
• 25:26
  SPEAKER_01: It's going to help you create your own template that might work for you.
• 25:31
  SPEAKER_01: So that's one of the features available to our paid subscribers.
• 25:34
  SPEAKER_01: And I think we do have a question here.
• 25:36
  SPEAKER_01: What is the premium content, Pankaj?
• 25:38
  SPEAKER_01: You are asking, just asking given its paid content.
• 25:41
  SPEAKER_01: LTR Risk Coach is available right now only to the paid subscribers of the LTR website.
• 25:47
  SPEAKER_01: It is very limited launch, but it is based on my knowledge and experience.
• 25:52
  SPEAKER_01: Another aspect of premium content really is some of the very detailed case studies that I share.
• 25:58
  SPEAKER_01: And they are based on warning letters,
• 25:59
  SPEAKER_01: analysis of more data,
• 26:01
  SPEAKER_01: They go heavy into analysis and maybe some predictive capability.
• 26:06
  SPEAKER_01: So some of them are leadership articles.
• 26:08
  SPEAKER_01: So some, I would say premium content is that type of a content, LTR risk coach.
• 26:14
  SPEAKER_01: And let's not forget,
• 26:15
  SPEAKER_01: I also offer 20% discount on my training products to people if you are
• 26:21
  SPEAKER_01: a paid subscriber.
• 26:21
  SPEAKER_01: So those are some of the benefits of the paid content.
• 26:24
  SPEAKER_01: Enough about that, guys.
• 26:25
  SPEAKER_01: I want to close by again, thanking all of you for showing up.
• 26:29
  SPEAKER_01: Show up in two weeks from now where we'll have a great conversation
• 26:34
  SPEAKER_01: about how AI is affecting career prospects for people.
• 26:38
  SPEAKER_01: Not only entry-level jobs, but also all levels.
• 26:41
  SPEAKER_01: I'll have somebody with great experience in the recruiting industry.
• 26:45
  SPEAKER_01: I respect them a lot.
• 26:47
  SPEAKER_01: And hopefully you will learn how AI is affecting our world today.
• 26:52
  SPEAKER_01: With that, I want to thank you again, Vid.
• 26:55
  SPEAKER_01: Thanks for joining me and hope to see you guys next time.
• 26:58
  SPEAKER_00: Take care.
• 26:59
  SPEAKER_00: Thank you.