#422: ISO 14971 and the Road to Risk Management with Edwin Bills

This episode takes an unexpected journey with Edwin Bills, a lead expert in medical device risk management and a key contributor to the development of ISO 14971. Edwin shares stories from his diverse career, beginning with his time as a submariner and electronics technician in the Navy, where he worked on early GPS technology and helped build missile subs from the ground up. This unique hands-on experience taught him the importance of knowing every component of a system.

From there, Edwin transitioned into manufacturing, where he applied his skills in teaching electronics and later, quality management. He was trained by renowned quality gurus like Joseph Juran and learned about concepts like Six Sigma and Kaizen from companies like Motorola and Toyota. These experiences, which focused on process improvement, cost of quality, and efficiency, laid a critical foundation for his future work.

Edwin connects these seemingly disparate experiences to his pivotal role in medical device risk management. He explains how issues in manufacturing—such as supplier quality, process capability, and yield—directly impact patient safety and a company's financial bottom line. This holistic view, blending technical expertise, quality principles, and a deep understanding of manufacturing processes, ultimately led him to the committee that developed the ISO 14971 standard.

Key Timestamps

• 01:52 - Edwin's military background as a submariner and electronics technician, and how he helped build a missile sub from scratch.
• 04:27 - The transition from military to a career in manufacturing and quality management, including his training under Joseph Juran.
• 10:59 - Edwin’s role in supplier quality and his lessons learned from working with a wide range of suppliers.
• 15:48 - The impact of Kaizen projects and other modern manufacturing techniques on efficiency and quality.
• 18:18 - An incredible story of a supplier with an ownership mindset that maintained delivery schedules after a factory fire.
• 18:59 - Connecting manufacturing experience and quality principles to the fundamentals of risk management.
• 20:46 - The crucial link between quality, patient safety, and a company’s financial success.
• 25:28 - How to apply statistical and manufacturing improvement tools in the medical device industry.
• 27:20 - The value of hands-on experience and a holistic understanding of a product's lifecycle.
• 30:09 - How Edwin’s diverse experiences ultimately led him to a key role in developing ISO 14971.

Quotes

• "If you want to think about cost of quality, it, you know, what is the cost to the company of this? You've got to have safe and effective medical devices." - Edwin Bills
• "When you get a statistician to help you improve your manufacturing processes, you want an industrial statistician... When you get to the clinical trials, that person's not the right person to have. You want a research statistician." - Edwin Bills

Takeaways

• Manufacturing Quality is Risk Management: The principles of quality improvement, process control, and supplier management—learned in high-volume manufacturing—are directly applicable to mitigating risks in medical devices. A high-quality, reliable process inherently reduces the likelihood of product failures that could harm patients.
• Prioritize Quality Over Price: As Edwin's experience with the electric motor supplier shows, prioritizing the lowest cost can lead to significant issues, non-conformances, and a higher overall cost of poor quality. A supplier focused on quality and delivery will...

Episode Transcript

(00:00):
Welcome to the Global MedicalDevice Podcast.
Where today's brightest mindsin the medical.
Device industry go to gettheir most.
Useful and actionable insiderknowledge direct from.
Some of the world's leadingmedical device experts and companies.

(00:28):
Foreign.
Welcome back to the GlobalMedical Device Podcast.
My name is Etienne Nichols.
I'm the host for today'sepisode and today I want to talk
about ISO 14971 riskmanagement for medical devices.
Most of us are familiar withthis standard and probably have those
numbers memorized at some point.

(00:49):
But with us today is to talkabout this is Edwin Bills who is
a leading expert in medicaldevice risk management, quality systems
and regulatory affairs withover 35 years of experience.
He's an AS SQ Fellow andregulatory Affairs Certified professional
and he helped develop ISO14971 and ISOTR24971 standards.
Ed has co edited keypublications, taught for Amy, AAMI

(01:13):
and top universities, and nowconsults on FDA compliance and product
safety.
He also advises medtechstartups helping shape safer, more
effective medical technologies.
And he's going to be openingfor mdic.
He was just telling me alittle bit earlier but, but Ed, how
are you doing?
And what did I miss and get wrong?
Anything I should add or take away?
That's probably a littlehighlight, I know.

(01:36):
Yeah.
Well, and I wanted to mentionyour military service as well, but
I saw, yeah, wherever you wantto start.
What I'm curious is whatpulled you into the world of medical
devices and anything beforethat that you have drawn on.
I'd love to hear any of yourexperiences, honestly.
Well, actually you brought upthe military and that's kind of where
this all started.

(01:58):
Matter of fact, I got my, Iwas wearing, looking for something
to wear today and I have myshirt on from Guam.
Oh wow.
And I, I spent quite a bit oftime in Guam because I was, I'm,
I, I'm a qualified submarinerand I spent seven years in the submarine
service and part of it was inGuam because I served on a missile

(02:21):
sub and we operated out ofthere and went all over the Pacific
and so I was electronicstechnician there and worked on something
that today you call GPS.
But in 1965 we were using aNavy satellite transit satellite
system for navigation and itwas actually probably a little bit

(02:43):
more capable than GPS is today.
Wow.
But it took two 19 inch wide,6 foot tall cabinets to hold the
electronics.
That was back in the earlydays when I started electronics in
the Navy, I learned in tubes.
The first 24 weeks of ourelectronic school was tube theory

(03:06):
and Then there was four weeksof this newfangled thing called transistors.
And.
And so I went.
My first submarine was a WorldWar II diesel boat.
That's another story.
And we had serial number twoof the SS radar system, which was
a World War II radar.

(03:28):
So we had the second oneinstalled on our ship.
And that was in the.
In the 60s.
So it was a fun gig.
I learned a lot and then went,like I say, to the missile boats
and actually helped build one.
And that was really excitingbecause when I walked into our navigation
center on the ship, when Ireported aboard, it was bare steel.

(03:51):
There was nothing in it.
Wow.
There was cable running downthe side, the electrical cables and
everything.
And we brought in all theequipment, installed it, did the
acceptance testing for the Navy.
And so I knew what was behindeverything after it was installed.
And these new guys would comeon board and they didn't know what

(04:13):
was back there.
Oh, yeah, you knew down to thenuts and bolts, man.
That would be valuable experience.
It was.
It was amazing.
And that kind of helped me inmedical device manufacturing.
Well, how did, how did thattransition come about?
Because, I mean, that justseems like two different worlds.

(04:33):
It is.
But I took that electronicsand started teaching electronics.
And I was doing it in avocational school setting after I
went to college and got mydegrees and all that kind of stuff.
After the Navy, Navy college,then I started teaching.
And then this guy came in oneday and he's friend of mine, he says,
you know, our company needssomebody to teach electronic maintenance

(04:59):
for our numerical controlsystems that we build and, and sell.
They were.
I was working for CincinnatiMill, which was the machine tool
builder for the world duringWorld 2 and after.
And they had a division thatmade the controls for all the equipment.
So robots, the.

(05:21):
Oh, we even did composite machines.
We built the machines thatbuilt the B2 bomber.
We built the machines thatbuilt the B1 bomber.
Wow.
We had machines at McDonnellDouglas in St. Louis before Boeing
bottom that had a 250 footlong bed for making wings bars for

(05:43):
airplanes.
Yeah, I mean, there was.
It was an amazing world that Igot to see and I got to be on aircraft
plants and all different kindsof places.
So again, it expanded mymanufacturing experience.
And then I was there and I. Ilived in southeastern Indiana, and

(06:03):
this plant was on the otherside of Cincinnati and was 65 miles
to work.
And I was sitting at home oneday reading the newspaper people,
I'm not sure they rememberwhat a newspaper read, but there
was an ad in the paper for aguy doing supplier Quality.
Well, I had.
While I was in the numericalcontrol business, they.

(06:25):
My boss had come in one dayand said, we want you to teach quality.
I said, frank, how do youspell that?
He said, we'll send you to school.
And so they sent me to school,all right, under some guy by the
name of Joseph Duran, who wasone of the Preem people.
Yeah, I was in Duran's class.

(06:47):
And really, he was.
He was 82 when he was teaching.
And we spent a week with himand part of the class, he took groups
of about six, and we had ameal together, just a bunch of us
right around the table with.
With the guru, you know, youbow down to.
But it was.

(07:07):
It was a great experience.
And then I got trained by aguy that worked on Six Sigma.
Motorola, invented that khakiboat for Motorola.
And then there was anotherguy, was a industrial engineer who.
I always forget his name.

(07:28):
Anyway, he.
He was really good.
He did the red X theory and abunch of stuff, and he never wrote
anything down.
He would go into a company andconsult and help them, and.
And that was their job.
Huh.
So Khaki wrote the book abouthis information.

(07:49):
So this is kind of interestingto me because if you had that personal
relationship with JosephDuran, I would imagine that probably
played a part when you werehelping with ISO 14971, with the.
Didn't have the trilogy.
Quality planning, qualitycontrol, quality improvement, Pareto
principle.
All of those things were his.
I have most of his books.

(08:10):
Yeah, yeah.
Okay.
And I did a presentation inone of these conferences on.
This is where this connectioncomes in.
When I was.
Was working at this companyand they had me teaching quality.
Well, they used.
Duran had a 16 videotape series.
People may remember videotape vaguely.
Yeah.

(08:31):
And they were set up so thatyou were doing a development project
and there would be a video,and you would watch the video and
then you would figure out, howdo we implement that on our project?
So we developed productfollowing this model, which came

(08:51):
from the quality planning andall that kind of stuff, the trilogy.
So we did it.
We implemented it there.
And so I was working with oneof the guys who was from Purchasing,
Fred Woods.
Fred and I ended up writing apaper on.

(09:13):
I don't think we called itvendor quality at that time.
So.
Can we stop a second?
Yeah, sure.
You want me to pause?
You mentioned the vendor quality.
Yes, we did a paper on vendorquality, which we presented at one
of the Duran's quality conferences.

(09:34):
And so I'm sitting there doingthat, and this newspaper thing came
out And I looked in there andthere was an ad for a guy doing supplier
quality.
Oh, I can do that.
I just did the paper.
Yeah, yeah, I got my ASQquality engineer certification at

(09:55):
that point too.
But I was working in.
So I sent in my applicationand got hired.
And it was four stop signs and12 miles from my house, which was
a whole lot different than the65 miles on the expressways and all
that.
Life changing.
Yes.
So I took it, I accepted.

(10:17):
And then I walked in the doorof this plant and I was there for
a little while before Ifigured out we're doing medical devices.
And there's.
There was the 78 version of820 at that point in place, which
was just manufacturingquality, where they hadn't introduced
the 96 version yet.

(10:38):
So design controls went, yeah,so I'm doing this supplier quality
stuff.
And we had a lot of.
Of suppliers.
I got in 200 different plantsin US, Japan, I know, Mexico.
I was in a bunch of places.

(10:59):
How did you.
I'm curious how with that manyplants, was there any consistency
or standardization acrossplants as far as their approach?
Well, no, these were oursuppliers and we were getting steel,
integrated circuits, hydraulic stuff.
I mean, it was across theboard a lot of different materials.

(11:22):
And so I met all thesedifferent plants looking at things.
I went in a die casting plantonce, and my habit from my Navy days,
I walk in the door, look downat the floor, look up at the ceiling.
Okay.
You know, and themanufacturing plant, to get just
a sense for their organizationand all that, I looked up in the

(11:45):
die casting plant and thefloor with.
Had all kinds of dirt in it,stuff hanging and all that.
And, and these die castingmachines, they would expel smoke
and all, you know, it was a process.
And I told the guy, I said,you know, this is, this looks like
this could be a problem.

(12:06):
And.
And the floor was not muchbetter, but at least there was a
lot of traffic on it.
So I went on and went on mybusiness, went to other plants and
I heard from this place, theysaid, you know what you told us about?
Yeah.
Said, well, one of ourmachines burped and it expelled some

(12:28):
hot metal and all this dirt inthe ceiling caught on fire and burned
the roof off the plant.
Wow.
So I put that in memory and itprobably wasn't another year.
I was in a different diecasting plant.
Walked in, looked up at theceiling, looked down at the floor.
I said, you guys need tobenefit from my experience.

(12:51):
I said, last place that lookedlike this, they had a fire and Burned
the roof off the building.
You need to clean this up nowor we won't be doing business with
you.
Yeah.
So.
And another problem that Iencountered was when I went into

(13:12):
this company, they had hired anew electric motor supplier and they,
they made hospital beds.
So there's a lot of motors.
It was three or four motors ona bed.
Yeah.
So I'm looking at that and,and those motors.
We were having some problems,you know, and they said, yeah, we

(13:34):
just, we just brought thissupplier in because the other supplier
was too expensive.
Oh no.
And so I worked with them andworked with them and trying to get
improvements out of theirprocesses and all this kind of stuff.
And finally I went to my bossand said, this ain't working.

(13:54):
He says, yeah, we got a lot ofnon conformances and returns and
everything.
Don't.
We said we got way too many.
So we had to go back to theoriginal supplier who was flawless.
That is so valuable.
That is so valuable.
Yeah.
So, you know, this kind ofstuff is going on.

(14:16):
I'm learning more aboutregulations and things and, and finally
I was the only supplierquality person there and.
And we had a new purchasingmanager that came.
Well, he was moved from someother job and accounting and purchasing
manager, but.

(14:36):
Okay, well, I'm not so sureabout this, but he was brilliant.
He understood that.
First you got quality.
Second, you got delivery.
Third, you got price.
He said, that's how I'mmeasuring each of my purchasing agents.

(14:57):
Quality, delivery and price.
And the whole world changed at that.
Point because previously itsounded like price might have been
first.
Oh yeah, yeah.
Typical.
A lot of companies purchasing.
Short sighted thing.
Yeah, yeah.
So Jerry turned everything around.
Jerry Kretzman, he's retirednow, but he did a great job of changing

(15:21):
the focus.
And we were doing a lot of stuff.
We were doing Kaizen projects.
And our first Kaizen project,the CEO of the company was on the
floor underneath somethingdoing some welding.
But he had been raised.
He was the second.
No.
Was he?
Yeah, he was second generationcompany ownership there.

(15:43):
And his, his dad had had himworking in the factory floor with
him is growing up so.
Doing everything.
Yeah, yeah, yeah.
So he could be.
He could do welding.
Yeah.
And I said we need to changethis process because this isn't right
over overhead welding from the floor.
We got to find some fixturesand change.
So the Kaizen projects did that.

(16:04):
We did a.
Did a whole bunch of reallymodern techniques back in.
That would have been, gosh,19, but I went there in 78.
Yeah, well that's, you know,manufacturing So I was a manufacturing

(16:24):
88.
That's when I went.
They.
They're still.
You know, kaizen seems to belike a new thing.
If you go to a newmanufacturing facility.
I mean, it's.
It's been around for a long time.
Kaizen, kata, kanban, all theJapanese terms that caught on.
So.
Well, we did process mappingon a wall in the factory.
Just meat paper, or we justpainted the wall.

(16:48):
Yeah.
Wall white and put the wholeprocess and all the movement that
occurred.
Looking at how inefficientsome of the things were, because
we'd do an operation, thenmove the thing to.
To the warehouse and thenbring it out of the warehouse and
do the next step.
And then, you know, somovement, the amount of miles that

(17:10):
a part moved in a plant was unbelievable.
Wow.
So now you started with thekaizen processes and some of the
other Japanese methods.
Streamlining the processes,improving the flow, trying to.
To cut down on.
Just in time.
Right.
Those kinds of things.
Because we had.
Great.
Just in time.
But I'll tell you a storyabout jit.

(17:32):
We had a plastic injectionmolder supplier whose job was to
deliver on Tuesday, everyTuesday, what we needed for the next
week.
Well, on Saturday, this planthad a fire, and it was in a small
town.
They had inadequate power,water supply.

(17:53):
The plant manager told thefire department to spray on the molds,
keep the molds cool.
I can lose the rest of theplant, but I can't lose the molds.
Yeah.
On Sunday, they had the molds requalified.
On Monday, they were in adifferent plant producing parts.

(18:14):
On Tuesday, the truck arrivedwith a regular delivery.
Wow.
That's.
That's unheard of today.
I mean, I. I can't imagine.
That's incredible, but thepower of ownership.
Right.
Right.
It was a.
A series of small injectionmolding companies that were owned

(18:34):
by somebody at the top.
You know, he owned each ofthese individual plants, so he had
the opportunity to move it toa different plant to meet his schedule.
Yeah.
And made it happen.
So it's interesting.
So I started my career inmanufacturing, actually in steel.
Then I came to medicaldevices, got to work on neurosurgical
and cranial fixation equipment.

(18:55):
Night and day world.
Completely different worlds.
Yeah.
And I can trace the value of that.
I mean, even pre.
Pre manufacturing engineer, Iworked on a machine, and that was
incredibly valuable.
Just working on the machine,with the rapport I had with the people
on the floor to get thathonest feedback on how things should
move.
I could just imagine someonelistening, saying, okay, yeah, I

(19:17):
can recognize the benefit ofall these manufacturing things, how
does that play into risk management?
And so I'm curious what yourtake is.
I could, I could, I'mconnecting some dots here, but I
want to see what your take is.
Yeah, well, when you haveproblems in manufacturing, you know,
you have a process that maybeis not quite as capable as you would
like.

(19:37):
Maybe your CPK, use the term,is maybe 1.0 and, and it's churning
out parts regularly that meetyour requirements.
Right.
So the shipment comes and youopen the box and you start sampling
on incoming inspection.

(19:58):
And gosh, this stuff, it's notup to stuff.
We're gonna, we're gonna haveto sort it and then sending the stuff
that doesn't meet requirementsback and tell them to rework it,
you know, do a scar to use it.
Supplier directive, action request.
Yeah, we typically, you know,this is something, I'm glad you bring

(20:20):
this up because this issomething I mention every now and
then with the medical device world.
We, when we talk about riskmanagement, we're talking about just
patient safety.
Typically most people outsideof medical device world, you talk
about, if you're in projectmanagement and you say we need to
look at our risk management.
They're talking about scope,schedule, budget.

(20:40):
If it's safety related, you'llhave some safety in there, but they're
all covered.
Is that an issue?
And what are your thoughts?
Yes.
Top management's focus is what money.
Right.
Meeting their requirements.
Well, if you don't meet yourquality requirements, if you have

(21:03):
safety issues, what does that do?
That costs you money.
A typical recall of a FDAproduct is multi millions of dollars.
That goes direct to the bottom line.
And untold in reputational damage.

(21:23):
That was the next thing I wasgoing to.
Yeah.
Because all this stuff istogether and Duran talks about that
cost of quality.
That's what one of the thingsthat he and his quality handbook,
which is sitting on the shelfbehind me here, he has a whole chapter
on cost of quality and theimpact of failures and, and also

(21:48):
the cost of, of doinginspection and testing versus the
cost of doing repairs and thenthe, the failures in the field that
really impact you a whole lot.
So all this stuff, if, if youwant to think about cost of quality,
it, you know, what is, what isthe cost to the company of this?

(22:13):
You gotta, you gotta have safeand effective medical devices.
And where I was gonna go onthe CPK thing.
Yeah.
Is if we're manufacturingmedical devices, they should be made
to a CPK of 2.0, not 1.0.
2.0.
Is six sigma.

(22:35):
Yeah, that 99.6 or whatever.
Yeah, we should be at least atthat point.
And I used to have, from my, Ihad another class on statistics from
a PhD who studied underWheeler in, in Knoxville who happened
to work at the bomb plant,worked at Oak Ridge.

(22:58):
He had some very interestingstories about going in and trying
to improve processes andproduction, saying, no, we don't
have time because we're onlyat 15% yield and we got to get products
out the door.
And you know, and, and wedon't have time to do this statistic
stuff.
It's going to waste time.

(23:19):
15%, yeah.
And they had finally, youknow, they were so upset about this
and they said, okay, you goteight hours process for eight hours.
After eight hours they had theprocess running.
The statisticians came in andhad it at 85%.

(23:41):
Wow.
Wow.
And they turned around andsaid, well, how much time do you
need.
Man?
In two weeks they had itrunning at 99 point something.
Wow, that's impressive.
So there is that.
And they had an example ofanother company you may have heard

(24:01):
of called Toyota.
Sure.
We have a control chart fromToyota that they handed out in class.
And the control chart was onthe most important component on an
automobile.
It was called a cigarette lighter.
Cigarette lighter productionwas running at 8.0 because they never

(24:24):
had to stop the process.
They could keep it running allthe time and it would just pump out
product and it would meet ourproduction needs just in time.
Yeah.
And we don't, you know, they,and, and I had a year chart and you
could see all these Japanesenotes on the thing pointing to places

(24:47):
and you could see a change andall this along the way.
So they were constantlyworking on the process at 8.0 to
improve it.
Wow.
Wow.
Yeah, that was Toyota beforesome of the new management came in
and it's kind of slipped.
And that's actually somethingI was going to say is you don't hear
about that kind of improvementvery often.

(25:09):
Or at least maybe I'm just nothere talking to the right people.
But 15% to 85% in eight hours.
I don't know if there's aresistance to change or what it is,
but it seems like things movemuch slower in the medical device
industry.
And what are, you know, howcan we actually apply some, some
faster moving improvements,would you say?
Well, take some of the toolsthat the Japanese did you, you've

(25:33):
got experience with, with allthe stuff that I did with Kaizen
and those kinds of things,improve your processes.
But at the same Time use thosestatistical tools that are out there.
But there's an importantconsideration here.
When you get a statistician tohelp you improve your manufacturing

(25:53):
processes, you want an, astatistician that is an industrial
statistician.
Not some guy that comes fromthe college campus somewhere that,
that does research.
You want somebody that doesthat and then when you get to the
clinical trials, that person'snot the right person to have.

(26:15):
You want a, a research statistician.
There's from the side.
So you gotta have the rightperson helping you at various points
in the process.
That's a really good pointbecause I could see the difference
in, I mean just even inmechanical engineering, if you bring
somebody out from the collegewho is purely theoretical and they
have no manufacturingbackground, they may try to injection

(26:36):
mold apart.
That has a, you know, a curvethat you physically is impossible
and just not recognize that.
Yeah, yeah, yeah.
Well, when I worked atMelikron we did injection molding
machines, control.
So I, I know what you'retalking about.
I'm sure you've seen somedesigns that were so fancy.
But one of the things, one ofthe things as a new employee they
trained you to do.

(26:57):
From our division we got sentdown to the main machine plant and
we learned programming and welearned machine operation.
We, we had to turn parts, wehad to machine parts and all that
kind of stuff then go back upand make the controls that, that
did those jobs.

(27:18):
Yeah.
So that was invaluable.
That's the way it ought to be.
Absolutely.
So I had seen all kinds ofchips being made.
You have big, big piles.
Aircraft plant in wasNashville was Avco.
I think they had a big machinemaking a lot of titanium parts.

(27:40):
And the titanium chips were,were like hip deep on the floor.
But they came around and theygot every chip off the floor.
Because of the price of titanium.
Yeah, yeah.
They had conveyor belts thatran under the machines that would
extract them.
But then they would go cleanup and, and bring all these.

(28:00):
Send them back to the smelterand re read.
Yeah.
Yes.
So go ahead.
Well, it's just so manyexperiences I got to have.
I got to watch a compositemachine making parts that go into
like the 787 is all composite.

(28:21):
The B2 I mentioned.
Yeah.
And was it the.
Was the F117, I think was a.
It was the first one we made.
But our composite machine,when we installed that at the Boeing
plant, the guys that did theinstallation walked in the front
door, they were blindfoldedand led into the plant in the Middle

(28:45):
of the plant.
And when they took theirblindfolds off were curtains that
went from the floor to theceiling all the way around the machine
that they were working on.
They wanted to go to therestroom, back to the blindfold,
lead them out, lead them back in.
So it was that secret at that point.
Wow.
But, you know, Watson, composite.

(29:08):
But tape is about the width ofa dollar bill.
We used to say the a dollarbill size of composite machines,
composite materials worth morethan a dollar.
Yeah.
So this was because it was oneof the first using composite material.
And then you laid it indifferent directions because the

(29:28):
strength was onlylongitudinally right then.
So you had to lay a stripe,stripe this way and then turn the
machine and turn the machineand turn the machine.
The machine was 16 axes.
Wow.
Oh, man.
7 axis is the most I've everworked with.
That's awesome.
We had 16 axis machines makingcomposite stuff.

(29:48):
Wow.
I, you know, I never.
I've never seen.
I can't believe I've notlooked that up to see how those made.
I used to do destructivetesting on custom aircraft interiors,
and we destroyed so muchcomposite material.
This is, you know, as recentas 15 years ago, I guess.
So it's a little bit morerecent, but.
Wow.
So anyway, all of these management.

(30:09):
Yeah, all of these differentexperiences led you to help and be
on part of the committee thatwrote ISO 14971.
So tell us a little bit about that.
Well, I was in the office asmy supplier.
Quality hat on and coupledesks away, was a quality engineer

(30:30):
who was on the committeewriting the third edition of IEC
60601.
The company was veryinterested in standards.
Yeah.
So Mike came in one day and hesays, hey.
He says, we need somebody onthe committee that has some kind
of experience in decontamination.
And he said, you do that.

(30:52):
Because I was.
I was in charge of mattressesat one point.
That was part of my job.
Yeah, we made mattresses.
So getting lots of patients onand off mattresses, you gotta decontaminate.
Right.
Makes sense.
So the hospitals use all thesechemicals that are really, really
scary to get to kill all the.

(31:14):
All the bugs, you know.
So we had to make sure thatwhen they put all these chemicals
on the surface would survive,you know, and eat holes in it.
So I had learned about that,and I had done mattress testing,
which is kind of interestingjob, was getting on and off the mattress
for eight hours.

(31:34):
I come home exhausted.
But no, I don't want to lay.
Down at any job, any.
Anyway, I got on the 601committee on the third edition.
I was working there, and I got.
Let's see, I got transferredto a different facility now as the
manager of quality and regulatory.

(31:55):
It was actually I gottransferred to Charleston, South
Carolina, and one of the.
One of the guys in purchasingthat I'd worked with had.
Had gone to the.
The company owned anothercompany that was in Charleston, and
he had left this company andgone to the Charleston company.
And I've told my wife abouthim going to Charleston.

(32:17):
She says, you know, ifanything ever would come up in Charleston,
we ought to think about it.
Okay.
So I came home one day and said.
Cheryl said there was a jobopening in Charleston because we
just bought the plant down there.
It just merged the two companies.
Yeah.
And I went in and told my bossthat I wanted it, and I think I got

(32:39):
it.
And we're going to be moving.
That's a great town.
Such good food.
Such good food.
So anyway, went to Charleston.
I'm down there.
And my same boss that moved meto Charleston came in or called me
one day and says, hey, our guythat was on the Risk Management Standards

(33:00):
Committee quit.
He said, you're on it now.
Oh, subpoena.
So what's Risk management all about?
So I was running the.
The for the whole company now,not just one plant, but I'm doing
risk management for everybody.

(33:21):
And we had a committee thatwas overseeing it.
We had our insurance agent, wehad our internal insurance person.
We had our doctor.
We had an inside medical guy.
We had our product liabilityattorneys inside the company.

(33:42):
And then.
Oh, yeah, we also had someengineers and Humber guys that came
in there.
Yeah.
And.
And we were doing.
We would bring in a productthat was supposedly ready to go and
look at the risk analyses andall the documentation.

(34:03):
And the rule was, you gavethem that.
The committee, all thedocumentation two weeks before the
session so they had a chanceto look at the risk analysis and
all the information.
So we're in there and we'relooking at this bed.
Sounds simple enough, right.

(34:23):
This was one that wasmotorized so that a.
Instead of taking two or threepeople to move it through the hospital,
you could do it with one set,a motor on it and some wheels on
the bottom that were drivenand all this kind of stuff.
So, okay, so let's test thisthing out, see how it does what it

(34:44):
does.
And we ended up taking it intothis hallway that was right side,
right outside our conferenceroom that had an incline.
So the one person is therewith me in the bed.
I was volunteered by thecommittee to Be in the bed and this
guy's pushing the bed.

(35:04):
And they said, okay, now whathappens if the, if the powered system
fails?
So they had put a switch inand flipped off the power.
And now instead of going thisway, the guy's going this way with
the bed pushing him.

(35:25):
What's the risk of this happening?
We never looked at that.
Well, your bed's not ready tosell yet, so you need to go back
to your risk managementprocess and fix that problem.
Because we're not going toclear this product for marketing.
Because we had to say at theend of the process, you know, when

(35:47):
you got to the overallresidual risk evaluation, that was
at the end of the process.
You've got a product, you sayit's ready.
Let us look at thedocumentation and concur with you.
This is the case.
Well, here's one you missed.
So residual risk, that seemslike one that companies probably

(36:09):
struggle with.
And maybe that's an assumptionon my part.
I'm curious if you have anopinion on one particular area that
companies struggle with morethan others.
But.
Well, the first one that they,they have is the use of fmea.
And you know about thatfailure mode effects analysis.
Yeah.
And they don't understand theproper use of tools.

(36:31):
They need to be usingpreliminary hazard analysis up front
very early in the pro.
In fact, as soon as you havethe intended uses and the patient
population, the environmentaluse, those kinds of things identified,
now you start risk managementand human factors both there.

(36:52):
You now in ISO 1345 in clause7, 3.3 C says the inputs, design
inputs are the outputs of risk management.
Well, what does that mean?
That means risk management hasto start before design input.
Yeah, yeah.
And where does, where does fmea?

(37:14):
Well, FMEA needs designoutputs to be done.
That's way too late.
That's where you cost moneyand time like this, you know, this
little.
The issue that we just talkedabout, you know, they didn't think
about.
Yeah, the motor failure.
So that would have shown uphopefully at fmea, but it apparently

(37:38):
didn't.
So you need to be thinking more.
Now another thing is that'soverlooked all over the place.
Most of the product standards,the safety standards, identify hazards
for you.
You can take that hazard andput it right in your risk analysis

(37:58):
because the risk is deemedunacceptable if it shows up in a
standard.
So now instead of trying tofigure out what's the probability
and the severity, you're pastthat point.
You're at unacceptable risk.
Okay.
Because it's there.
Yeah, yeah.
Then if you implement thesolution that they advise you in

(38:22):
the standard.
Now you, that's your riskcontrol measure.
You put that in.
Okay, I'm, I'm doing a, atraceability matrix.
That's what I'm doing along.
No, I, I, I'm following the columns.
Yeah, yeah.
And then if you have a test inthat standard that tells you whether
that is acceptable or not,then that's the final step in your

(38:45):
process that says that riskcontrol measure is effective.
So now you're done.
Okay.
You don't have to do anymore.
Yeah.
So there's all thosecalculations and gyrations and everything
that people go through to getto that point that can be bypassed.

(39:05):
That makes sense.
Now you're, you go ahead.
No, well, you mentioned thetraceability matrix and that seems
like a given in this conversation.
However, I am curious becauseof historical, you can go so far
back that, you know, I wasn't,I wasn't there for so many of these
things.
There had to be some risk management.
I mean you talk about being ina submarine in 50 something years

(39:27):
ago.
There's a whole lot of riskmanagement there.
Yeah.
I was in electronic school inthe submarine service when the thresher
was lost.
Wow.
And what was it, 225 peoplewere died from that incident and
they stopped production of submarines.

(39:47):
And that's how I got to thediesel boat.
I was, I was slated to go to amuscle boat but they, they had all
of us in the, in the chain.
You know, we were headed downroad because it, it took, I was in
training for two years beforeI got to a boat.
And, and then we, we went thatcircuitous route while they redesigned

(40:11):
the process.
There's a subsafe process iscalled, which is basically a quality
system thing.
And what it does is it saysanybody can report any defect anywhere
along the line.
And if you don't, you're in trouble.
Yeah.
You know, because that's lives.

(40:33):
Yeah.
Well, I'll give you an example.
I did when we built thissubmarine, we took it out on the
first sea trial.
The first sea trial, we tookit all the way down to test depth.
Now The HULF is 3 1/2 inchesthick of HY80 steel, which is high
tensile strength.
Yeah.
Okay.

(40:54):
And it's a 32 foot diameter hull.
Okay.
Okay.
So we're inside and we hadsome guys that were, had some experience.
I was in relative, I was myfirst missile boat and they said,
come on, we're going to godown the missile compartment and
we're going to do this thing here.
So that's where one place youcan go where you can go all the way

(41:15):
across that 32 foot.
There's not stuff in the way.
So I said now tie this thingon this side and tie this string.
Carry it all the way acrossthe other side and tie it there.
Get it as tight as you can.
Okay.
So we're, we're basically onthe surface with this tight string
across there.
I got the test depth thatdropped 18 inches in the middle.

(41:37):
Wow.
Hull compression.
Was that it?
Yeah, because there's 0.44pounds per square inch per foot of
depth on the hull from thewater weight.
Yeah.
Causes that.
So that's, that's the pressurewas lost.

(41:57):
Something happened that thehull was crushed.
Yeah.
I used to do the algebra toget the, the I guess depth on the
cotangent or whatever or 3218 inches.
That seems like a lot morethan I would expect.
Wow.
Yeah.
So.
And, and not only the hull but every.
About every what it was about18 inches.

(42:17):
There was basically an I beamaround the inside of the hull.
Strength of the hull.
So we've got these, these Ibeams and then you know all the way
around the circle to.
And, and the idea is of courseif your pressure's on equal from
all sides you all you have todo is, is resist that pressure and

(42:42):
you know you're okay.
Yeah.
But the thresher didn't.
Well, later on when I'm onthis boat on patrol at a Guam we
got a message that says submiss and it said the, the Scorpion
was missing, overdue.
And then a little couple dayslater we got a sub sunk message that
said the Scorpion was lost andthey had not been through the sub

(43:07):
save process.
They were built before it andthey were, they were rotating all
the submarines back through itto get them caught up and they didn't
get there in time.
Yeah.
So there was another 99 loss.
So you know.
Yeah.
And I know there are lots ofrisk control measures and it seems

(43:28):
like other industries havereally figured out two things in
particular that the medicaldevice industry still seems to struggle
with or at least consistentlyhas issues with.
One of them is project management.
And that's maybe a wholenother animal.
I don't know if you want totalk about that or not.
But the risk managementitself, that's the other thing.
And I don't know if you can mention.
Why is it just the differenttype with every devices so different?

(43:50):
I mean we're not building carsthat there's a standard way of building
something.
But I don't know, projectmanagement's really important.
I had an experience back in mynumerical control days.
We had a project manager whowe were designing a new control.
Okay.
It was a lot of electronics.
Right.

(44:10):
Much smaller and it was.
What was it?
That was for turning centers,I think.
And Lenny was assigned aproject manager.
So he, he drew a part chart onan E size drawing.
He filled that whole thing upbecause this was before computerized

(44:30):
project management.
Okay.
And over the, the three yearsof that project, Lenny wore out three
electric erasers.
Thank that.
Because one of the things onthat is you, you know, you run into
problems.
Typical projects, thingsaren't going to be ready on time.

(44:51):
So what do we need to adjustso we can still meet our date?
Well, Lenny, in that threeyears missed the final date, the
initial date that wasestablished by three days.
Wow.
Wow.
Okay.
So just so people listeningunderstand what that is.
PERT chart is the programevaluation and review technique chart

(45:11):
which is used, it's a projectmanagement used to visualize tasks,
their dependencies, timeline.
See the float.
Yeah.
Write a flowchart of theprocess and all the, the connections
to different things.
And all the input.
Yeah.
All the estimated dates andwhat, how long it's going to take
to do processes and all thesekinds of things.

(45:32):
And, and what's interesting,PERT there, PERC was used to build
Polaris submarines.
I see they had had a missilethat was over here being built, a
submarine over here being built.
And oh, by the way, we'regoing to cut this submarine in two
and add 150 foot section ofmissile tubes and we're going to

(45:53):
bring this missile and put init and this is all going to fit together.
Yeah.
And we got all this navigationstuff we got to do.
And that was my part, youknow, and nuclear power and so firing
and everything else.
Yeah.
So Admiral Rayburn was the guythat led that project and it got
done early.
Wow.

(46:13):
Yeah.
Yeah.
Project management, that issomething the military and construction
has figured out.
But medical device we're stillworking on.
Maybe not completely militaryhadn't been as good with it because
when they buy new products,they don't do quite as efficient

(46:33):
a job on that.
But that original design, thatwas an experience that they need
to go back to and look at thisand see what we can hear.
So if I go Back to either 1497or 24971 are the things, particular
sections that you feel likeget particularly misinterpreted or

(46:54):
ignored consistently.
Well, I talked about the FMEA thing.
Yeah.
Need to be using Other tools.
FMEA has a couple of problemswith it.
One it doesn't cover.
It only covers failures.
It only covers single fault failures.
14, 9, 71 says all failuresand it says normal condition.

(47:16):
When everything is workingright, what kind of hazards can you
have?
A scalpel when everything'sworking right, has a sharp edge and
there's a hazard there.
Yeah.
Okay, so that's a very simpleexample, but that's true of all medical
devices.
We have problems that we needto consider.
So FMEA doesn't cover that.

(47:39):
FMEA comes late in theprocess, as I said, design output.
So that's a tool that you canuse as a check to make sure you didn't
miss anything.
Yeah, but it still has gapsand holes in it.
Fault tree does multiplefaults, you know, it pulls the whole
thing together.
That's a nice tool that camein a software where that originated

(48:02):
and, and then the PHA that Italked about, that's really good.
The questions in Annex A of 24971, 37 questions there now asking
you what's in this product?
What, what kind of risks doyou envision from electricity?
Oh, electricity, yeah.
It's an electric product.

(48:24):
It's got a lithium ion batteryin it.
Yeah.
Oh, well, I need to thinkabout that.
You know, where the risk in that.
That's all stuff you can findat the beginning.
Yeah.
You know, so the standards,the questions on safety, you know,
your concepts for design.
It doesn't.
FMEA doesn't talk aboutusability issues.

(48:48):
So you got your URA toolrelated risk analysis.
Yeah.
Which by the way, should beintegrated with the traceability
summary because a lot of that overlaps.
Yeah.
With user needs and, and so on.
Yeah.
The front end of the userneeds and everything is real good.
But then it goes into yourrisk controls and all that.

(49:08):
Well, that's already in thetraceability thing.
So there's a.
And that's why I have harpedfrom the beginning we need software
products for risk management.
Yeah.
It's a database applicationtrying to track all this data and
make sure it all works together.
So that that's a problem.

(49:30):
And I know Greenlight has asolution in that area and there's
other ones out there as well.
But the idea, you know, ofspending a lot of time on our good
old spreadsheets, it's goodfor concept, but it's not good for
some of the complex productsthat we have now.
Yeah, that was one of mybiggest difficulties, I suppose,

(49:54):
and I'm sure you've seen thisas well.
But when I was, I used to be aproject manager for a drug delivery
combination product company.
So I, I was integrating therequirements from the drug company,
multiple drug companies, toour single, you know, medical device.
Try to integrate all of thedifferent requirements and tie those
to our risk management.
All done in Excel.

(50:15):
It's a struggle.
I don't know how it's donethese days, but that was one of the
things that I realized.
You know, I saw engineersleaving the industry because of the
documentation side of things.
I don't think that should bethe case.
We're losing a lot of talentbecause we're not giving them the
tools they need.
But maybe that's a different,different subject.
Well, the, the whole way weattack documentation is a, is a problem.

(50:36):
We wait to the end and go backand document reverse engineering.
Yeah.
And, and you mentionedcombination products.
I'm on the combinationproducts committee for risk management
and currently 14971 is.
It was reaffirmed.
It's good for five more years.
24971.

(50:58):
Well, we've got two projectsgoing on to expand 249711 is machine
learning.
Now question on that.
I thought 34971 covers machine learning.
Well, it's a US and Britain document.
Okay.
We're now expanding to ISO.
Okay, gotcha.

(51:19):
So it is.
We're taking that base andexpanding it to ISO.
So there will be 249712 whenit's printed.
Eventually when we update24971 it will roll into that so it'll
get a little bit thicker.
And then there's another oneon combination products by chance

(51:41):
which has just started.
And that's going to be 24971dash three.
Okay.
And that's coming from Amy'sTIR 105 which I was on the committee
to author that.
And we used the trace matrixfrom 14971 ideas to be the repository

(52:05):
of all the risk from the drugand the device in one place.
Yeah.
And we also talk aboutinteraction risks.
What if I put this, this drugin this plastic syringe?
What's going to happen?
Yeah.
What's chemical actions aregoing to be?
What are the risks of that?

(52:25):
That needs to be considered as well.
So we not only have the drugin the device, but the interaction.
So we have a system here thatwe're, we're developing and we need
to do risk management from asystem perspective for that.
And that's how they TRACEsummary helps pull all that information
in one place.

(52:46):
Love that.
That's good.
Is our IVDs, because it seemslike that would be another area that
deserves its own view.
I don't know if there'ssomething coming up as well.
IVDs, there's a absolutelybrilliant guidance in 24, 971 for
IBDs.
Okay.

(53:06):
And they take an IVD and pull it.
You know, here's how you do itin 14971.
And the, the two people thatwrote that, one of them still on
the committee was at Abbottfor 30 years, I think, and.
And she's.
She's great.
And the other guy retired.
He's in his mid-80s now, but he.

(53:29):
He was in IVDS forever too, sothey did a fantastic job.
But it's also got some ideasthat might apply to other areas worth
looking.
At, no matter what you're in.
Absolutely.
Yeah.
Yeah.
So two more questions and then I'll.
I'll let everybody get back totheir day.
If you've taught all over,from Amy to Virginia Tech and so

(53:52):
on, what's one concept, or Iguess that students or professionals
struggle with the most when itcomes to risk management?
Well, some of it is kind of basic.
If they're brand new to theindustry, understanding the difference
between harm hazards andhazardous situations and how that
works, which is to me, I'vebeen in it way too long.

(54:16):
Why did they have problemswith that?
And, and we took the.
If you've seen the sharkillustration of the different.
Well, anyway, we were at astandards committee meeting we used
to meet in Sanibel in Decemberwas our standards meeting.
And the problem was by thetime we got done with our meeting

(54:39):
and got outside, it was dark.
So anyway, we were trying tocome up with a way to explain the
differences there.
And somebody said, well, youknow, the shark swimming right out
there, well, you can see themout the window there.
And the shark is certainly ahazard, but if I don't get in the

(55:02):
water, no big deal.
No hazard.
Well, no, there's still ahazard there.
What you have there is ahazardous situation.
You have to get in the waterand get exposed, you know, and the
shark has to be somewhere nearby.
So there's your hazardous situation.
And then the harm is when youget bit.

(55:24):
So we were using that as asimple thing, and people kind of
poo pooed that after a while.
I know another instructor,Amy, she uses coffee, you know, the
hot coffee cup fromMcDonald's, like cardboard cup that
spilled somebody's lap, you know.
Yeah.
And then, then we got to more,you know, got the cap on the thing,

(55:45):
you know, all the different.
Now we got These kind of rightups have all kinds of protection
built in and everything.
So that, that's just one wayof explaining it.
But that's one in thebeginning that people have trouble
with and, and then trying toput all this stuff together in a,

(56:07):
in a chain.
Now the trace summary to me isit helps a lot to see.
Well, now I got these thingsand then this risk is unacceptable.
Well, this one's okay.
This is acceptable.
I don't need to do anythingfor this one except in Europe because
you have to do them all.
Yeah.
And then this one down here,this one's unacceptable.

(56:30):
So I need to have a riskcontrol measure.
So now I gotta have a design input.
That's my risk control.
So risk control connects to design.
Develop now the risk controlmeasure and then I decide what that's
going to be.
I write a design output whichis a specification for that requirement,

(56:52):
which is a safety requirement.
And then I need to make surethat that actually shows up in the
product.
Yeah, that I'm.
Oh, I, I missed it.
It didn't get in.
Don't want to find that at the end.
You want to make sure it's there.
So now you have design verification.
Does that, doesn't it?
It looks at things to see ifeverything, if the outputs are covering

(57:16):
all the inputs.
And then the next thing, Well,I put this thing in here.
Does it work or not?
Well, is it effective?
I do design validation andmaybe I can do some tests in verification
that will show that.
But a lot of times it'sinvalidation, especially if it's
a usability issue, becauseusability validation is done as part

(57:41):
of design validation.
So we work together with thepeople doing usability a lot from
the risk management sidethrough the process and making sure
that the hazards that theyidentify as issues get in as risk,
then get implemented.
It inserted in the design andimplemented and then validated by

(58:04):
the usability validation atthe end.
So we, we, you know, this is ateam effort.
That's why I say, I alwayslike to say risk management is a
team effort, but so is thewhole design process.
Absolutely, yeah.
And it should be.
Yeah, right there alongsidethe design.
Well, in the beginning is notjust the design people.

(58:27):
In the beginning you need tohave manufacturing in there because
do they have to buy a new machine?
Well, that might take sixmonths to a year to get that on board,
you know, purchasing.
What kind of suppliers do theyneed to be searching for new things
that you're going to use?
What kind of environment isthis thing going into?

(58:48):
Medical environment?
Well, that's the medicalpeople's expertise.
They can come in and tell youwhat kind of things you need to be
considered there.
So there's, there's threepeople I came up with right away
that need to be on that teamat the beginning.
And then complaint people needto be on there because they need
to understand what thisproduct is going to be so they can

(59:11):
prepare for doing complaintinvestigations at the end.
What's this product going to do?
Where's it going to work?
You know, all those aspects.
And then let's see what else.
Complaints.
Well, that's examples.
I mean, yeah, it's, it's agreat example.
And to your point, people whoare just doing this in a spreadsheet,

(59:31):
that is a rough life.
You know, it reallymaintaining that traceability.
You talked about getting tothat, that harm.
And what risk control measuredo I have in place and is it still
in place six months later?
Because if those are separatespreadsheets, you know, Greenlight
actually does a really goodjob with their software with this.
Go ahead.
When manufacturing comes inand says we need to make a process

(59:54):
improvement to reduce cost, sowe're going to make a change to your
design.
Okay, what's the impact ofthat change?
Yeah, so my process has to beresilient and to be able to manage
change because it's going to happen.
There's no way around it.
Right.

(01:00:14):
Single mold injection.
You want to go through a multicavity and you know, okay, the gates
are here.
Why are they here?
That's a design output drivenby a design input, which may be a
risk control measure.
Absolutely.
Started with aluminum mold andnow I'm going to switch to steel.
Yeah.
You know, what's, what's theimpact there?
Yeah, your material.

(01:00:35):
What's the impact of using thealuminum mold that wears quicker?
Yeah.
Or in an anodization, youknow, for whatever reason, going
to a. Yeah, no, these arereally good points.
And if you don't have thatability to trace from the design
to previous risk mitigations,I mean, it's just a nightmare.

(01:00:56):
It really is.
And it's one of the reasons Ithink the medical device is not.
A lot of companies are notagile as they could be if they had
the ability to go back andlook and see.
Okay, this can change forthese reasons.
Or if we stay within theseparameters, we can move around.
Let me give you one more itemthat is really useful as our friends

(01:01:16):
at the Food and DrugAdministration have these wonderful
databases out there.
You can put your product codein the total Product lifecycle database.
And it will come up and showyou all the products that are built
under that same product codeand all the problems they have, all
the failures and everything.

(01:01:36):
Failure codes.
It's a tremendous tool.
So when you're starting thisnew product, you put your product
code in that database and thenlook up who your competitors are.
So you can tell marketing togo off and look at these different
competitors and get some, youknow, some dirt on them.

(01:01:58):
And then you look in there andsee, well, these are the kinds of
problems that we need to takeinto account when we design our product,
because these failures havealready occurred.
These are real.
There's your risk management beginnings.
Yeah, yeah, absolutely.
Yeah.
That's why I say pha.
The total product life cyclefeeds into that, too.

(01:02:19):
Your complaints on yourprevious products that you've had
that's similar or they use thesame processes, even a different
product, but it uses the samemanufacturing process.
Oh, well, I guess there's somecomparison there.
Yeah, that's a.
That's a really good point.
Yeah, that's a great tool touse, too.
I'm glad you brought that up.

(01:02:39):
I let us go over, but I couldtalk to you for a long time.
I mean, these are great stories.
This is what I. I live for.
This is the reason I do this podcast.
But I am going to have to.
I guess I'll have to shut usdown, but hopefully our paths will
cross before.
Before long.
Where can people go to findyou to learn more and go ahead if
you have any last words.
I'm on LinkedIn and I answerquestions on LinkedIn.

(01:03:01):
When you hang up, I want totalk to you a second husband.
Yeah, absolutely.
Okay.
Well, thank you so much forbeing on the show, Ed.
And like I said, hopefully if.
If I don't, I'm not gonna beat mdic, but I hope it goes well.
And before too, maybe wraps.
We'll run into each other atsome point, I'm sure.
Some point.
Yeah.
All right.
Okay, everybody who's been.
You've been listening.
We'll put some links in theshow notes so that you can easily

(01:03:23):
get a hold of Ed.
Obviously find him on LinkedInand we will see you all next time.
Thank you so much for tuning in.
Take care.
Thanks for tuning in to theGlobal Medical Device Podcast.
If you found value in today'sconversation, please take a moment
to rate, review and subscribeon your favorite podcast platform.
If you've got thoughts orquestions, we'd love to hear from
you.
Email us at podcastreenlight Guru.

(01:03:45):
Stay connected for moreinsights into the future of medtech
innovation.
And if you're ready to takeyour product development to the next
level, Visit us at www.greenlight.guru.
until next time, keepinnovating and improving the quality
of life.