Inside Out Quality

The 2013 Baxter Mold Incident --Lessons in HEPA Filtration

Aaron & Diane Season 1 Episode 4

In 2011, Chris Wall, an HVAC technician at a Baxter Healthcare manufacturing site in North Carolina inspected the HEPA filters in a room where sterile Saline IV Bags were filled. He found them to be contaminated with mold. Management stopped him from changing them for 2 years, until he finally reached for help to the FDA. The result was a warning letter and 18 million dollar fine. 

In this episode, Diane and I meet with Tony Lee of AT Analytical (http://www.atanalytical.com/) to discuss the 2013 incident of mold contamination, and how to prevent mold contamination and ensure customer safety.   


Aaron Harmon:

Hi, I'm Aaron Harmon.

Diane Cox:

And I'm Diane Cox Welcome to Insight out quality.

Aaron Harmon:

both Dan and I build and implement quality systems in the biotech and medical device industry. But we often get asked, Is this really necessary? Do we know if we are doing too much too early? Or do we even need a quality system?

Diane Cox:

Our goal is to explore questions like these through real life events and experiences shared by our guests from various regulated industries. We will show you why quality is not just about compliance and how when it's done right, it can help your product and company improve lives and make a difference.

Aaron Harmon:

In 2011, Chris wall an H back technician at Baxter healthcare manufacturing site in North Carolina, inspected the HEPA filters in a room where sterile saline IV bags were being filled. What he found looked like mold growing on the filters, so he did the right thing and began replacing them until local management found out and told him to stop. He ended up having to leave five filters in place that still had mold. The management not only told him to stop, but ask them the future. He and his team record only that discoloration was seen. They were not allowed to use the word mold. Chris pushed management to allow him to change the remaining filters for the next several months on successfully. Finally, he contacted an attorney. And 16 months after first seeing the mold, the FDA arrived for an inspection because of his efforts. The results, the FDA found several mold species growing in the filters and issued a warning letter, the company settled lawsuits at a cost of $80 million, and Chris ball was paid 430,000 for acting as a whistleblower. For the management that kept the filters in place. Five of them lost their jobs because of their role. HEPA filters keep air clean. by trapping 99.97% of the particles that are greater than point three microns. They trap bacteria, fungi, pollen and large viruses so that the air where products are produced is essentially sterile. For companies producing sterile medical products, the FDA requires air colinas, which is outlined in ISO 14 644 and USP chapter 1116. With us to dive in further on this topic is Tony Lee, the owner of AT analytical located in Baltic South Dakota. Welcome to the show, Tony.

Tony Lee:

Well, thanks for having me.

Aaron Harmon:

So you're a local expert in HEPA filtration. Can you tell us about your business and how you support companies with clean air?

Tony Lee:

Sure can. So we're a 12 person company located up in Baltic. And our licensing and quality support is really funneled through Sita, which is contained environment testing Association. And with that, we actually have four teams of technicians that go out and test it within the five state area. So we'll dive into more of the testing later. But there's several tests that we do to try to help mitigate that type of stuff.

Aaron Harmon:

How did you get into HEPA filtration? It's out of curiosity.

Unknown:

Yeah, we got into it listening to our customers, I've got a biomedical electronics degree by trade. And just you would go to sites and they would really be struggling with Hey, we had a certification company come in, and they failed this, this and this. But we have nobody to do those repairs or to do that follow up testing and and so as listening to our clients that really got us going into the HEPA filtration testing,

Aaron Harmon:

that sounds like a good practice to listen to the place that customer it has treated us well.

Unknown:

Yes, thank you.

Diane Cox:

So if we could start with some basics, Tony, how do HEPA filters work?

Unknown:

There's three main ways that a HEPA filter works, the easiest one is impaction, right. And this is definitely for your larger particle sizes. Because you have a weave, and that we even that large particle, that air flow is actually carrying that particle. And literally, it's kind of like a coffee filter, right? It cannot get through the weeds. So that is the easiest type of trapping of particle there really is. The second one is interception. And interception is the particle can actually be smaller than the weave. But it literally then hits that fiber membrane if you would an actual individual fiber, and it adheres to it. That works for anything, you know, especially in the smaller particle sizes, because a lot of these filters are going to be a three nine, or a four nine filter. So that truly means you your smaller particle sizes technically could fly right through that week. So that's, that's why interception is so important. And then the last one was actually very controversial for many, many decades was diffusion. Diffusion is actually using because when air travels across the filter, there's a small static charge that builds up on that silica membrane if you would, I shouldn't call it a membrane I should call it a fiber and then particle is so small in size, it's actually attracted to the actual silica fiber. And that's your main three ways to capture particles within a HEPA filter.

Diane Cox:

Got it? I didn't, I didn't realize there were three different kinds. That's awesome. So is it common to find mold in HEPA filters,

Unknown:

if monitored correctly? Absolutely not. And that's that's really the frustration when I when I was rereading this article and stuff, because mold takes very good conditions to start developing. And air traveling through a HEPA filter is not a good condition, right? If you're trying to grow the mold, right, we put it in an incubator, we put it in all these ideal conditions to actually get it to grow. So so no, it is definitely not, you have to have a repeat situation over and over and over again, or for a very long period of time, before you're going to get that mold to start affecting that HEPA filter. It kind of comes back a little bit like New England compounding, not to bring up another situation, but how their air filters were located so close to a recycling center. They were those filters were constantly getting overloaded or losing its filter capacity, that then that mold and all that stuff was able to start leaching through the filters. So again, a filter only has a very finite amount of capacity. And when that capacity is used up, it has to go somewhere, and therefore it's going to go downstream.

Diane Cox:

Sure that makes sense. So how often do they need to be monitored or inspected?

Unknown:

So it really comes back to your governing body. In general, FDA is every six months, if you're starting to have issues, they might ask to go every quarter. And then if it's USDA, they're usually annual, it's going to really be dictated by your governing body or if you're trying to troubleshoot or work through a situation.

Aaron Harmon:

The reason the mold can grow from what I understand is that you're having to accumulate enough carbon and organic material in the filter to give it some kind of food, nutrients, essentially, that needs to grow, because I'm assuming it probably can't grow really well, just off the silica

Unknown:

correct. The other aspect is moisture. Um, I got the privilege to work with two different troubleshooting scenarios down in I call it the Southern climates, Atlanta, Georgia. Now it sounds silly. But to correct that cleanroom situation, we actually had to add a furnace to preheat the air leaving the cleanroom before it would go through the air conditioner, because you need certain parameters within your age factor to be able to capture that moisture. And you know, it can be something as simple as that adding a furnace or adding heat to that air so that you can get the proper delta t across the a coil of your age back to make it work. Because water usually is the contaminant.

Aaron Harmon:

So it's a lot more than just the filter. But it's all the aspects of the air.

Unknown:

Oh, absolutely. You know, if your H vac system isn't working correctly, the HEPA filter can't do its job. And vice versa, the H vac system can be working perfectly, but if you put in a subpar HEPA filter, it can't capture the particulates coming through the H vac system. Yeah, it's definitely a harmonized system.

Aaron Harmon:

In that case, was that recirculating air or fresh air coming in? Like how was that airflow,

Unknown:

there are guidelines to that. So at a bare minimum, we try to have like 12 air changes per hour of fresh air coming into a system. And depending on if you're doing more hazardous drugs or more volatiles, those all make a you know play a role in that. But anytime you can re circulate the air within a clean room. And it's already cleaned and polished air. So you want to reuse that as often as you can. Unless if you have that hazard or situation that you need to mitigate. So yeah, clean air is very, very expensive to make. And so yes, you do want to re filter it and reuse it as often as possible. We call them pumping dumps or a single pass system is for like a hazardous application. Let's say you've got 4000 cfm coming into a room, you're actually going to heat and or cool that air plus filter it, dump it into the room and then you're exhausting 100% of it. That's a very, very expensive way to operate a cleanroom. So

Aaron Harmon:

it sounds like air quality goes beyond filters and even some of the aspects of the air conditioning but then also maybe in design and how the entire facility is set up.

Unknown:

Yes. So we call it a site assessment before we would ever help coach a cleanroom design. You'd have to go in you're going to find out if you're working with any hazardous chemicals or hazardous drugs. And then you're going to ask add that to your application. Because if we can recirculate 18 to, let's say, 30, air changes per hour, not only are we being more green and energy efficient, we're actually going to have a cleaner, clean room. But that site assessment is really, really important so that you know what the customer is doing, so that you can get a system that most fits their needs.

Aaron Harmon:

So you mentioned with hazardous drugs, so there's clean air to protect the product, but then also to protect the operators.

Unknown:

Yes. A clean room is not a cleaner, right, we actually break it down from the BMBF, fifth edition, USP 797. And then ISO 14 644. Each one of them really has their own little caveat that they require 14 644 for like, a lot of the manufacturers doesn't require to do a HEPA filter integrity test on the filter in the housing, but that you pass particle counts within the room, well, then you would have like USP 797, they require all five tests to be done to include HEPA filter, leak testing, particle counts, viable sampling, so that you get a more, I call it real world scenario of what's happening with that. The other aspect of testing is under dynamic condition or under static condition, dynamic condition has to do with Are you testing the room at real world application? Are you testing the room? How it's going to be used? Or is it just sitting empty? Because most of the time, US people are the contaminant? Most of the time.

Aaron Harmon:

So in terms of level of contamination, or air cleanliness, I'm familiar with ISO five, air colinas, ISO seven, eight, can you maybe give a little more in depth to that just by numbers?

Unknown:

Oh, sure, absolutely. So I so five in is really your pack or your primary engineer control or your hood, that's going to be the cleanest air, it's going to be unidirectional air, and it gives the best opportunity to protect the product. And within that column of packs, you have both a biological safety cabinet and a laminar flow workbench. Right, a biological safety cabinet then pulls the air away from the operator. Plus, it also provides HEPA filtered air across the product, and then re circulate 70% of that, plus or minus within the biological safety cabinet to give that added protection, but it's actually pulling air away from the operator, a laminar flow bench is it kind of really an open bench that's just flowing air at the operator. And at the product, because it's a non hazardous application, we're not worried about the person in that type of situation. Those are the two primary ISO five devices. And then you take a think of a bull's eye for playing with darts. And the very center is your ISO five, well, the standard states for like USP 797, that in the bull's eye is your ISO five, and then you have your buffer space, your buffer space is usually ISO seven. And why do we care about classifications? And is because in an ISO five device, it is established as recommended thresholds and 1116 of hey, in ISO five, I'm only allowed one colony forming unit a viable sample in my viable sample? Well, in an ISO seven, it's 10. In an ISO eight, it's 100. So as that ISO number goes up, you're actually allowed more contaminants in the room. So that classification is is huge. Because right? The tighter the of the tolerance, the lower the number, the lower the particles and the thresholds that you're allowed in that application. And then we also have what we consider controlled, non classified, that might be a staging area or something, or a support space to a clean room or a production setting where right, we don't want we're not going to go in and cut a board, we're not going to create a bunch of particles. But hey, this is our staging area to take into a classified space. And all that stuff makes a huge difference to protecting that ISO five or that bull's eye

Diane Cox:

as the ISO number goes up or down, I guess does the HEPA filter itself change the type of HEPA filter or is it air circulation? Or is it a combination of the two that might be I guess different or other factors that I'm not thinking of? That's a great

Unknown:

question. So ISO five requires unidirectional air to protect the product. When you get into an ISO seven you can have non unidirectional or we call it dilution. The HEPA filter itself from an ISO five to an ISO seven does not change, we still target a 409 filter, we just don't have the air changes per hour, or the unit direct unidirectional factor involved from a five to a seven to an eight. In general, these are just general thresholds, an ISO eight would have 20 air changes per hour, ISO seven is going to have 30. And an ISO five is unidirectional, we really don't care about their changes per hour anymore, we actually want enough unidirectional air to protect that product. And then the other factor between a five SEVEN and an EIGHT is going to be pressure differential. So generally, in a non hazardous application, we're going to target point 02 inches of water column every time we go across that classification. The other aspect, we had briefly touched on vials right? We're going to go from 110 and 100. We also have the non viable or particle counts where we can exceed 30 520.05 micron particles in an ISO five environment. And then right it keeps jumping, we're allowed 352,000 particles of point 05 microns. And then when we jump up to an ISO eight, we're allowed three and a half million particles of point 05 microns. So yes, there's a bunch of domino effects. And as that classification changes, all of those utilities change, or all those aspects of the H fac change with it.

Aaron Harmon:

The case of this facility with these moldy HEPA filters, why were they so reluctant to change?

Unknown:

Yeah, it's cost. You know, one of the ways I was when I read the article, one of the ways that they would have had it very early notification was doing a viable sampling sequence that was either dynamic or more samples within the room. Because when we do viable air sampling, we actually take an air pump, if you would, that pulls a known quantity of air across the media, and basically embeds in the media, and then you grow it. So whether you're using TSA or SDA agar plates, you easily would have been able to capture that because if there is almost any discoloration on the HEPA filters, with an adequate sampling plan for variables, you would have caught it, you, you would have seen it way ahead. The other aspect that I come back to and, you know, I know the FDA is kind of a little bit reluctant to, I call it push it. But I feel part of the checks and balances within the industry could be using third party services as ourselves. There's several Quality Certification companies across the country, and I get that it's always it comes back to the dollar, it comes back, they didn't want to change those HEPA filters, because they could easily be 600 to $1,000 apiece, then you get them installed, then you do your HEPA filter leak test, you easily now went from a $600 filter costs almost a $1,200 total project cost per filter. So it has a huge domino effect and cascade effect that way. But I really want to come back to that. It's not just one test. Now you have viable non viable HEPA filter, leak testing, air changes per hour, and pressure differential. And then for the other aspect that's really catching on to as far as testing goes, is doing a smoke analysis or an airflow visualization test. So you actually add smoke into a room or into a PAC, and you capture it with the video. And what you can see is right when a person walks into a room, you're actually going to have a physical visual aid of how that airflow is going to interact with people. And that's huge. It's just huge. There's just there's so many more things to it. Getting back specifically to the Baxter deal would be humidity control and humidity monitoring. If you had a cold damp situation where you're actually able to grow without mold, something should have been flagged in there humidity control system, something whether they weren't monitoring it enough, whether they weren't capturing it, there should have been another red flag. Well before the discoloration on the filters, though, that that's what I get out of it when I read the article is there should have been other red flags.

Aaron Harmon:

I don't know if they did this at this facility, but I've heard of putting like MERV filters upstream.

Unknown:

Yeah, so So The Merv filters there are several different ratings of MERV filters, just like there are several different ratings of HEPA filters, and The Merv filters allow you to really pre clean the air before it hits the HEPA filter. And what that does or what it buys you is what we call reserve filter capacity. So because that MERV filter, let's say that MERV filters 80% efficiency rating, it's going to capture 80% of those particles passing through it. So then only 20% of those are hitting that HEPA filter. And that has a direct relation on that reserve filter capacity. That really comes back into effect when you're designing the facility. Do you have enough HEPA filters in the room, all of those play a huge huge factor into it,

Aaron Harmon:

and is the M for medium efficiency, like I don't know much about these filters, but I'm assuming that H and HEPA filters for high efficiency is that medium efficiency.

Unknown:

So I don't I don't deal in the MERV market, but HEPA filters high efficiency, particulate air. And again, it just there are several several different kinds of HEPA filters. Just like there are several different kinds of MERV filters, I cannot stress that enough. When we're talking a four nine filter, you're actually talking glass, silica fibers, that if you touch them, you just put a hole in them. They are that fragile, but they work good for filtering, bacteria, proteins, viruses, all of that. So

Aaron Harmon:

so you don't want to be poking at it while you're cleaning or working in a room like that.

Unknown:

Correct. I still get a kick out of I was at a workshop. And the maintenance guy goes lit, we literally we took out our vacuum cleaner, and we're cleaning the HEPA filters in the room. And I just said, Well, did you replace them when you're done? Well, no, we just cleaned them. So there's no cleaning to a four nine HEPA filter. If it doesn't pass HEPA filter integrity, or if it's discolored, it needs to be replaced.

Aaron Harmon:

So you said a four nine HEPA filter. What's the other kind of filter? You know, you had mentioned a different one earlier?

Unknown:

Well, there's really there's two different rating systems that are available commercially available, most of your building facilities are going to go by a MERV rating. And I honestly don't remember all of those terms as far as what the acronym is for and their rating schedule. I'm more familiar with the HEPA filtration on the life sciences side. But they actually have all the way to a six nine filter that's made with Teflon even now, to make it more durable and to capture more particles. But then that Teflon HEPA filter now is you're starting to talk a couple $1,000 filter, and those some of them are actually cleanable. I have yet to come across that and 21 years though, I am not saying I'm a skeptic, I'm just saying I believe it when I see it.

Aaron Harmon:

Now we'll take a quick break to hear from one of our sponsors.

Unknown:

Today's startups become smiles growth engines. In South Dakota, we're entering a new stage of expansion for a biotech industry, and you'll want to be part of it. Hi, I'm Tony Johnson, Executive Director of South Dakota biotech, where the state affiliate of the International bio organization and we're proud to be leading a state that's driving innovation to feed, fuel and heal the world. South Dakota biotech is here to inform, to connect, and to advocate for our critical industry. Whether you're directly involved in biotechnology, or looking to learn more about it, we want to hear from you find us at www that SD bio.org. Now back to the show.

Diane Cox:

I was just gonna ask him, I think it's kind of taken us a little bit back into the conversation. So in the Baxter case, I think Aaron mentioned that there were five HEPA filters left behind with discoloration, so to speak, if this had not been escalated, what are some of the potential issues that might have been caused with the product itself? In this case, it was What did you say sterile saline bags? What are some of the issues that could have occurred with that product?

Unknown:

It's a cascade of things, unfortunately, so the first question I have when you when you read that article, though, was the discoloration actually in the pack? Or was it in the room, right? And they don't tell you everything I've read on about this, they do not tell us if it was in the buffer space, or if it wasn't in the primary engineer control. If it's truly in the buffer space, you could actually argue that the system was designed right in the bull's eye approach that that ISO five classification space or that critical area still had integrity, right. So it still comes back to risk management. Not that I'm downplaying this at whatsoever. I'm just stating that. That's why when you design a facility, you give those classifications because every time you change that classification, you're adding additional protection from the outside environment to that ISO five device, the FDA is very crystal clear that if you have an ISO five device with any discoloration on the HEPA filter, it must be replaced. But again, they don't tell us in the article, if it's in the ISO five location, ISO seven or, or ate or, or whatever, I would assume those are pretty tight details that they're going to keep under their cuff there.

Aaron Harmon:

In that case, there was no reports of the adverse issues with patients either,

Unknown:

it makes me believe that it was actually in the buffer space. So a little bit of speculation, right, we're gonna pull open our crystal ball here a little bit, but it actually goes back to still, the people making the product are still the best, they have to have good aseptic technique. Because if they're walking through a buffer space that has mold particles floating around in this, who it before they go into that ISO five device, they need to do a proper aseptic technique, wiping down their hand spraying their hands, and before they go in and make that product. So again, the bull's eye approach has been taught for many years, and I know they're moving away from it. But it still gives you that level of protection every time you go into that classified space. And the people are important, you know, following the aseptic technique, and all of that aspects to creating product that isn't contaminated, because you can have the perfect classified space and still have bad product with bad technique.

Diane Cox:

So Tony, you mentioned the overall cost for changing a HEPA filter would be roughly 600 to 1000 plus some testing costs in addition to that, but what did you say Aaron? Was the amount that in millions, a million? Yeah. What a big difference.

Aaron Harmon:

Something was short sighted there,

Unknown:

there should have been other flags at that Baxter facility, that should have triggered way before the HEPA filters, whether it was humidity control. There are other aspects, there's other tests that are required to be done, that really should have captured that I would love to see their viable sampling plan. Because that viable sampling plan, if they're taking one viable sample in a 1000, square foot cleanroom. That's just not adequate. But does it conform to the standard? Technically it does. And that's where this whole cost savings versus management versus best practice really comes into play. I'm a little old school on it, you know, we use the square root of meter squared, that's like our minimum sampling size for a room. On top of that, if it's smaller than that, you know, at a bare minimum, you should take three samples in a room. It gives you a an adequate dist, description of what's happening in the air, maybe you're going to do two impaction plates and a contact plate, having enough statistical data so that you know what's happening, I suppose maybe cost an extra $150 every six months for an FDA facility per room. So yes, 18 million compared to some of that initial testing costs. It is peanuts, but it does come back to having an adequate sampling plan. humidity control is really, really important. The location of a clean room is important. Yeah, there's a lot of different factors. It'd be really nice to find out just more information about the article and see see what other factors were involved.

Aaron Harmon:

My guess is the underlying issue was a culture problem. Like you said, there should have been enough signs. But the number of people that were that were terminated after everything had settled that there was probably a much more widespread problem at that specific plant. And they probably had to do a reset.

Diane Cox:

Yeah, very well. Could have been some some other kind of preliminary flags that were also on the hush hush. Like you said, Aaron due to culture and management direction.

Aaron Harmon:

Tony, if someone's listening, and they have HEPA filter questions and want to reach out to you and contact your business, what would be the best way to do that?

Unknown:

The best way is via email because I am still a field technician. So I do travel the five state area. My email is Tony at ATA. I can see that Bz. You're welcome to either email me direct, you can go to through our website and hit a service request form. And that all comes we're a small business, we get to wear many hats. But we do have a great staff and we try to help people out as much as we can.

Aaron Harmon:

No good and I have worked with you and your business for quite a long time.

Unknown:

I remember meeting you fresh out of biomed school Aaron

Aaron Harmon:

and we just had Oh Dan was out just doing some work on a or testing on one of our HEPA filtration systems that we use right now. So

Unknown:

yeah, that's actually I'm gonna, you know, you got my brain going a mile a minute now and you know, one of the other aspects that I want to bring up for those that are listening is what is HEPA filter integrity testing? And one of the caveat that that I could say we're arguing with even with our own colleagues in the in the CETA is do you test just the filter media? Or do you test the entire housing have that HEPA filter. And you know, if you look at IES T, and there's all these other different standards, and we're a firm believer that when you do a HEPA filter leak test, it's not just the media, it is the entire housing that holds that HEPA filter. So if there's bad craftsmanship in the ductwork, and you actually can have leaks coming into that cleanroom that the areas and even passing through the HEPA filter. So I really want to, you know, just maybe highlight that a little bit of one of the tests that that is important that again, it just seems like there was way too many, it's not just one thing that probably failed at that site, and HEPA filter leak testing, you know, we take a known mount of particles, put it upstream of the HEPA filter, and then we scan on the downstream side of the HEPA filter. So any of those particles that we're injecting, we have a threshold of point zero 1%. That's it. If there's a leak, we can usually find it. So

Aaron Harmon:

I think at ours, we had leaks around the house, where it was probably like coming out a cocked up.

Unknown:

And that's a very common scenario, because especially like on a gel seal, HEPA filter housing or something of that nature. You know, over time, air passing across any type of a medium usually dries it out. And then it's no longer making that sealing edge that it needs.

Diane Cox:

Yeah. Before you set up, have a filtration system. How important is it that you have your other infrastructure kind of up to par before you do that installation? Can you just simply install an H back system? Even if it's a crappy infrastructure? And everything is going to be fine? Or do you need to kind of take some steps to mitigate?

Unknown:

Why would you put a critical area in a subpar facility? So I would come back if if an ownership structure was saying hey, this is what we want to do. It really is up to that team that is designing that facility to state Hey, timeout. Is this the best location for this? We call it a site assessment. clean rooms are expensive, you know, you could have anywhere from 400 to $600 a square foot for cost of that clean room. Why would you put it in a subpar location or subpar facility? It just doesn't make sense.

Diane Cox:

You've done a fantastic job Tony explaining all of this technical stuff, stuff that I've never had to deal with personally. So I appreciate your simple explanations, and it's very helpful for us to kind of put the story in the context.

Unknown:

Thanks for having me.

Aaron Harmon:

Yeah, thanks for being on the show.

Diane Cox:

We hope you enjoyed this episode. This was brought to you thanks to South Dakota biotech Association. If you have a story you'd like us to explore and share, let us know by visiting www.sd bio.org.

Aaron Harmon:

Other resources for quality include the University of South Dakota's biomedical engineering department where you can find courses on quality systems, regulatory affairs, and medical product development. Also, if you live in the Sioux Falls area, check out quinnbet A local Quality Assurance Professionals Network. You can find out more about wibit by clicking on the link on our website to Diane and I would like to thank several people but a few who stand out are Nate pepple for his support with audio mixing Barbara Durrell, Christian or support with graphics design and web. And lastly the support from South Dakota bio