Julie Omohundro If the pattern you are referring to is recalls due to "abnormal system behavior or unexpected user-initiated actions," I'm not sure I agree with the process improvement expert. I'm a fan of the FMEA process, but I don't think it's a silver bullet. I think it's Achilles Heel is that there is no way to know whether you have identified all the modes and effects. In that case, I think the abnormal and unexpected are most likely to get overlooked.

Beyond that, I don't know...design engineers work in mysterious ways, but I don't think I've ever seen one use requirements-based V&V without also using FMEA. Aren't these more like two sides of the same design coin?

Kris Kelly Hi Julie,

I understand your perspective and I agree, FMEA (by itself) isn't a silver bullet - it needs the rest of the process aligned with it to increase the capacity to identify the (unexpected behavior) hazards that exist for every product/system.

The use of FMEA’s provides the “Failure Mode” part of the input to Risk Management and Mitigation. FMEA’s do not address potential ”Non-Failure Mode Use Errors”. User Needs from the Intended Use Statement provide inputs into the Product Risk Assessment (PRA) to guide in the identifying of Hazards and Harms resulting from “Mis-Use” or “Unexpected Behavior”.

Integrating the PRA with the FMEA’s dramatically increases the likelihood of identifying & mitigating recall generating failures. Medical Device Manufacturers who don't spend the time (with the right team members) to identify/mitigate all the Hazards/Harms caused by Unexpected Behavior are doomed to remain on the FDA Recall List.

Kris Kelly From my experience there are two psychological contributors to teams missing the boat.

1) They see this activity as adding too much time to the project, (too many process layers), and it'll put them way behind schedule - a 'slow start' in their minds. "We don't have time for that..."

Sound familiar?

2) This is an even larger problem - they don't know "how" to do this properly. Which means they can't sell the benefits (recall prevention) and they're worried they'll look bad. It's completely understandable - who needs public (or even private) ridicule, right?

However, as I've experienced on large scale medical device/hospital system projects - when you understand the beauty of this process - then it's far easier to negotiate the upfront time required to implement this program, and start the project right. After you get started, and everyone in the room has that "Ah ha" moment, the push back slowly fades away. It's a great moment.

Clarisa Tate The FMEA the article seems to be talking about is no longer the old FMEA we know. This is more hazard and risk assessment/analysis. Hence why more and more auditors talk about normal fault risk assessments then ask about abnormal fault risk assessments.

FYI I've seen those ah-ha moments on people, however it doesn't necessarily mean they'll now adopt the new way of thinking. Change can be very difficult for some people.

Ginger Cantor Amazing that FDA keeps approving their PMA devices without a culture change, even when they keep getting Warning Letters. They do have deep pockets and a history of this.

Excelente artículo, en Chile debemos hacer una normativa mejor para los dispositivos médicos.

I see a different issue. Risk management is a larger process than "design engineers create a DFMEA and justify the risk is beneficial p, ys manufacturing team creates PFMEA and sometimes a control plan to justify manufacturing is releasing good stuff".

Risk management needs to start with a structured process to identify failure potenials. A tool such as Fault Tree Analysis (FTA) is far better at find failure modes. It challenge product and process designers to determine how things can fail, ideally a component or code section at a time.

Jacqueline -KAUCK I think it's not as black and white as discussed here. I think we should give a thought to the initial process of bringing devices to market. Sometimes everything works great in lavatories and initially testing but when used in many patients there is always the factor device failure - after all it is a machine with all possible failure rates AND we have the factor human also -one on patients side and one who uses/ implants it. I totally agree if companies know about a possible risk beforehand they should immediately stop shipments and inform customers and should also be held accountable for this. Furthermore if they knowingly put something into market before a product is ready just because they have a financial reason in mind ONLY that is not ETHICAL nor can it be -in my eyes -excepted.
I believe evidence based medicine is a better way to go here to really protect our patients, customer and also the company's - This will contribute to Patients safety and at the end this is what we all try to achieve - at least I still have the hope that's what our goal is. After all it could be YOU or one of Your loved ones ending up with one of those products.

Jacqueline -KAUCK I think it's not as black and white as discussed here. I think we should give a thought to the initial process of bringing devices to market. Sometimes everything works great in lavatories and initially testing but when used in many patients there is always the factor device failure - after all it is a machine with all possible failure rates AND we have the factor human also -one on patients side and one who uses/ implants it. I totally agree if companies know about a possible risk beforehand they should immediately stop shipments and inform customers and should also be held accountable for this. Furthermore if they knowingly put something into market before a product is ready just because they have a financial reason in mind ONLY that is not ETHICAL nor can it be -in my eyes -excepted.
I believe evidence based medicine is a better way to go here to really protect our patients, customer and also the company's - This will contribute to Patients safety and at the end this is what we all try to achieve - at least I still have the hope that's what our goal is. After all it could be YOU or one of Your loved ones ending up with one of those products.

Shocking

From reading this article I sense a perfect pill or device should be attainable or exist. If this were the case, we all should have IQ’s of 175 along with being Olympic Athletes. We are human beings with unique DNA and conditions. To assume one can design this outcome for all inputs is setup for failure. I suggest consideration is given to the input to the variability of infinity to the process of design. Unfortunately sometimes you can’t make situations up. However, through FTA, FMEA, it helps when measured against what is known. Remember how these are ultimately developed,on disaster. Engineering of buildings got better after 911. We have great people who can do better, discussions like this can improve outcomes, carpe diem!

Richard Bollinger Kris Kelly, your "...hidden critical defects…" look like risks with high Detection ratings. Seeing them in one device/component would be reason to search in others. And, applying resources and expenses appropriate to their priority. Be careful though, ISO 14971 does not approve, no matter how reasonable or useful.

Shlomo Hillel

Proof of Concept and Device Design translation development clinical Research and monitoring to test safety and effectiveness of test device guided by GCP GMP GLP is rarely mentored. Sadly, Safety recalls just waiting to happen putting patients at serious risk.

Medtech veterans at Aproio claim that an “informed design” workflow is required.

Kris Kelly These are all wonderful comments - there's nothing better than sharing thoughts & ideas with intelligent people.

Here's my two cents regarding PRA/FTA/FMEA: PRA's and FTA's are Product driven which start at the top and work down. Whereas FMEA's are design driven, start at the bottom and work up. They're two techniques used to identify risk and mitigations...we believe in using both of them.

As for devices being shipped with risks - many devices are shipped that way. The task is to determine if the risk is acceptable; oftentimes Human Factors Studies directly contribute to risk mitigation, if they're properly conducted and the HFE's are listened to.

Simone Rudolph-Shortt Interesting

Rafael Simões de Oliveira As someone working with Software Quality, you should know that, when comes to software, there is no such a thing as "fully tested". This kind of statement reassures the common misconception "quality=testing".

Nate Badii Excellent article. Thank you for sharing.

Kris Kelly Hey Jim - I wanted to shoot a reply because I really liked your comment - I hope this feedback from my Process Colleague will shed some light on our approach. :)

Unfortunately, there are no perfect pills or silver bullets. Using various techniques for products and processes we systematically execute analysis, evaluation and control of possible risks.

The goal is to produce a “perfect device” and have it last forever. Much the same as having a goal to win a race. Not attaining perfection does not preclude us from aiming for it.

Hence, shipping a product does not depend upon perfection but on the overall residual risk acceptability of the identified and mitigated risks.

Cheers!

Kris Kelly As an aside - I was a Lead SQE Contractor on a large Medical Device Project back in 2014, where a Fortune 100 Biotech firm was marrying two systems for European Customers, (hospital enterprise system with 500+ point of care devices) which hadn't been done with this technology before.

I was able to utilize the process described in the article above, for this (almost 2 year) Project, and we successfully installed the system at 6 large sites in Europe.

All 6 sites have been operating without any reported issues (beyond initial configuration & implementation) since 2015 and they're very happy with this large, networked system.

For further details, please feel free to reach out to me.

Salvatore Domenic Morgera It appears that the specific device recalls cited are due to poor verification and validation (in other words, quality control) and the consequences can be seen to be highly detrimental. We can expect to see more and more recalls, as companies race to roll out neurodevices and other medical devices. In consulting to medical device companies, I have stressed the importance of understanding the fundamental underlying phenomena before proceeding to device design. For example, there are devices that treat the brain as a black box when it comes to neurostimulation and neuromodulation. Some are FDA approved and produced by companies whose scientific/engineering staff could not even answer my basic questions regarding nerve fiber activity, nerve fiber interaction, and population synchronization under exogenous electric field stimulation signatures. Unbelievable!

Very good article! I like Professor Salvatore Morgera’s comment. Very relevant!

Cole Stryker McCann I've considerable experience in product design industry qualification & limitation testing in the oil and gas industry and have seen this scenario play out time and again because the right questions were ignored and glossed over early on in the component's design.

With a long time interest in the medical device industry, I am wondering, how difficult is it to test a device that is exposed to the caustic environment of the human body and interacts with its many systems?

Rainer Moosdorf MD PhD We probably mix up different problems, which, each of them, seriously affect patients but have different causes. One is indeed, that some devices are released without clear proof of concept and I always mentioned that devices should be developed on request of patients and physicians. The market should follow the needs and not the other way around. Secondly, devices have to be tested intensively before roll out. This includes durability, material fatigue, continued efficacy and others. The third problem, which is obviously responsible for the recent Medtronic recall, is a failure during the routine manufactoring process. This happens beyond testing during development and indicates a problem of internal quality control. In many of these complex devices, special parts are manufactured by other companies and then put together and the more important is a meticulous quality control of the final manufactoring process.

Your proposed solution is to use an FMEA approach - which is useful enough, but hardly unique or cutting edge. What makes you think the manufacturers involved are not using it, or a similar approach?

Kris Kelly While the FMEA can uncover some hidden risks the majority of hidden risks will likely be uncovered from the PRA or FTA with inputs from HF studies, which is a key part of recall prevention.

Many organizations use the FMEA almost exclusively; ISO 14971 requires more than the use of FMEA's.

Identifying, evaluating and mitigating risks needs to be looked at from perspectives other than design. They need to be looked at from users and their unintended uses. Recall prevention can be greatly enhanced by identifying targeted testing in these high risk areas.

So yes, FMEA's are commonly used in the medical device industry but we go beyond only identifying bottom up high risk areas from design and include top down risk identification as a powerful addition in identifying high risk areas.

Agree completely with your approach. An FMEA/PRA/FTA also should look at other manufacturers recalls and the causes... in fact, since these recalls are a public safety risk, I believe it may be appropriate for the government agencies to provide the data to an industry team to study these types of recalls and relook at the standards and protocols for recommendations that can be added to the current design/test/"pilot lot" scenarios.

Proposal . In view of ..... must be taken into account
Possibilities in most manufacturing pacemakers are limited.
When the patient is using limited mobility possibilities, he / she will have spontaneous complications when he / she does an unusual motor exertion. Causing complications and abnormalities.
During the rapid movement of the nakedness ..
It must be treated as a natural person to lead a normal life

This is a correct correction of a pacemaker device that contains a sensitive, adaptive and responsive body with harmony and harmony with the patient when moving and working like any natural person.
2 - other than securing the brain area. It affects the stroke of the blood clotting
Through a controlled neck filter and force. During the patient's movements fast and natural

Art Pichierri Medtronic has been recalling devices for years. As a monopoly of specific technologies, their failures result in a replacement market whereupon they get paid the difference and the physicians get a fee to change out the units. It's a win win for them both.

Antonin Cuc In EU there are Medical Devices "CE" regulated by the Directive 93/42/EEC Medical Devices and the Tests of new products must be coordinated by the EN Standards and with supervisions of national Authorised Bodies with international mutual accepting Protocols "CE". There are principials mistakes, when the Product Medical Devices "CE" on common Markt EU are dangerous, despite there were false agreements with Protocols "CE". Such situation should be coordianted in EU as dangerous Disproporctions with principles EU and Tests-such usage must be immediately stopped with satisfactions for all Useres of Producers and of Authorised Bodies -and then with inforced principial Design Changes and new testing, by stronger Technician Requirements by changes in EN Standards for Safety usage. I have other tragic experiences - as a State investigator H+ S - when the Useres are Mass taking the perfect products illegaly - when the Medical staff Mass ignored Technical Laws EU+ exact product invoids!

Aren’t The FOod and drug Administration checking the validity of devices and isn’t there a study ? For Breast Implants they are so stringent and expect a 10 year study !

Christopher Smith It should not come as a surprise to anyone that problems are encountered with any medical device after being manufactured. Why? Because of the all mighty dollar. I've worked in aviation, then medical, and left both industries. I refused to give up quality for quantity, large companies & corporations don't. Profits will always come before people. Capitalism at it's best.

Roger Cepeda, JD, MBA, RAC I may be missing the point here, but I don't think a whiteboard session is going to undercover hidden defects due to unexpected uses. That's kind of what "unexpected" means. While it sounds nice in power point slides to say stuff like "we need to expect the unexpected" or "we need to see around corners," the truth is that all Class III devices pass a safety and effectiveness standard, and sometimes, things are missed. Perhaps it's due to greed and cutting corners, as some here have implied. I've seen it as well. But phen-fen killed people after making it through an FDA clinical trial process that is more stringent than most devices on the market. A compounder in New England was responsible for 50+ deaths and the FDA said they had no authority over them (but they do now). And back to devices- just because CAPAs show up as a top issue in warning letters, it doesn't mean we can change that trend by identifying its existence.

Kris Kelly To answer Roger Cepeda, JD, MBA, RAC

We are promoting the process of identifying potential product issues using risk management techniques. Not only Class III but also Class II devices pass or have accompanying rational for all executed verification and Validation protocols before they are transferred to “On-Market”.

Roger is very correct that potential issues are often missed using the normal “Happy Path” testing. That is where we come in and using a risk-based approach that targets possible unexpected behavior we mitigate the identified possible issues and then prioritize (the greater the potential risk the more exhaustive the testing) non-requirement-based testing. We are able to then to some degree expect the unexpected and test accordingly. Granted, we are likely not to identify “all” unexpected potential issues but we will improve the product being shipped.

I'll finish my thoughts - Part 2

Roger Cepeda, JD, MBA, RAC Kris- I think we are on the same page, and the challenge is convincing management to take the time needed for the full V&V, rather than the minimum required to check the box with FDA or a Ministry of Health. I'm sure you've heard the excuses, "first, let's get it on the market, and then we'll add your ideas in a future upgrade," or "let's not allow perfect to become the enemy of good, as we need to bring this product out to hit our commitments." And, as you've pointed out, that short-term thinking becomes the root cause for "a stitch-in-time-saves-nine" missed opportunities. The only managers I've met who get this are the ones whose P&Ls were hit with the downside costs.

The ECRI doing a good role in issuing and keep updating the Recalls on most medical devices and technology while they also have a good database and we'll instructed module for the same among their services

Dan Golka Kris. This great. Can I share it in the Med Tech Directory?

Kris Kelly To answer Roger Cepeda, JD, MBA, RAC (Part 2)

This is Part 2 from above:

Again, Roger is correct in his statement about CAPA’s, but he is only looking at identification. CAPA’s attempt to both correct the issue and then prevent the issue.

The correction is realized as adequate in the short term and the prevention is more long term after successful effectiveness checks. CAPA’s are not closed until the effectiveness checks are completed successfully.

So we are not only identifying an existence of an issue we are also correcting the issue and then preventing the issue from occurring again assuming our effectiveness check is successful.

If an organization is using the CAPA process only to identify issues without correcting and preventing, they are wasting most of their effort.

Roger seems to define compliance as compliance to the letter and not compliance to the intent.

Proactive compliance to the Intent (via Recall Prevention) yields the Desired Results.

Kris Kelly Yes Dan - please do!

Kris Kelly BTW - I sincerely appreciate all the comments. Tthis ongoing dialogue benefits all of us. :)

Cases where people are injured / die, are as terrible where only 1 person is injured / death or thousands, cause it only takes one person from your love ones to join the statistics so your family can be in pain.

Is the responsability of people who work in this industry to work as if they were FDA and think out of the box when it comes to possible scenarios can lead to injury or death.

If we only do the basics of V & V, this will only bring the basics regarding patient integrity and safety.

Should more people (management) need to hit the bars (jail) ?.

Andy Pal Coming from the R & D side of the equation, I have seen contributing factors in the past that can lead in this direction. The process can get flawed very quickly.

Ship dates can be determined by earnings calls, not how long does it take to make. Thus that "magic date" must be met no excuses. From there a timeline is built in reverse to see how that date can be hit. The tasks and roll ups are pulled out of thin air based on educated guesses. Then the real challenges start to pop up. Often, those date swags do not include any provision for fixing anything. No turn time for bug fixes, no plan B for any subsection that needs a course correction. The goal simply becomes "do what we need to do to match the requirements". Requirement interpretation can get fuzzy to get past some of them. No matter what, development has to overrun the schedule. This kicks the can down to QA /V&V. They have no time, so they are not able to really test edge cases. They focus on matching tagged requirements.

Kris Kelly Perfectly stated Andy.

David LeVine For countless CEO's, founders, and execs, business can be too brutal to tolerate. Pressures from investors, Wall Street, boards of directors, competitors, employees, struggles within themselves. While we all have a moral compass, too often fear of failure takes center stage, at least until the golden dream leads to disaster as it too often does. Stepping back is always an option, sometimes discovered too late in the game.

Most often, slow is fast.

As a member of the risk & analysis team for medical devices and daughter of someone who lives with a defibrillator, this article is a nice remember to Development and V&V's teams. It's important to remember the output of these analysis is the patient & user's safety, its not just about a regulation requirement.
A year ago, when my father called from the hospital and ask me "which defibrillator should I chose?". I answered, pick the one from the company I known as a most challenging team I've met in term of validation.