Bias in BioProcessing

“It’s not really a question of who is biased, but which bias is the correct bias with which to be biased.”

– Jonathan Sarfati


The Stainless People vs. the Polymer People? No, not a dystopian battle. Just a room full of engineers at a recent ASME-BPE meeting defending their biases on multi-use vs. single-use components. Whether male or female, baby boomer or millennial, end-user or supplier, everyone–especially those of us in the bioprocessing standards community–have inherent biases which we find all but impossible to check at the door. But, just as penicillin can stop an infection, facts fight biases. In order to uncover these facts, though, we need to collect them from all sides.

The fact that women are underrepresented in engineering is an important topic for another day, but, suffice it to say, standards are being written mostly by traditional industry men like myself who have been around the sun more times than we care to divulge. Destination conferences are heavily favored by us baby boomers. More progressive people (millennials), who also use these standards, don’t show up in large numbers. Perhaps that is because they have their own set of biases including in their communication skills; preferring interfacing over face-to-facing. History is repeating itself. Old-school pharma people were set in their ways before we biotech people came along and asked why, right? Today, this new crop of industry professionals brings a fresh perspective that needs to be heard and I’m willing to bet they don’t want to wait two years in between new versions of a standard. What my generation considers reasonable, this younger generation calls unacceptable. Why wait for the morning paper when you can get all the news on the phone in your hand? The world has changed. How do we balance our generational biases? On-line updates in real time? Maybe.

Suppliers and end-users bring their biases to the table, too. Suppliers are closely acquainted with the equipment they sell: the speed at which it runs, the best pressure for its use, limitations, etc., while end-users are most familiar with the process in which they use it and at what point issues might occur. But what happens when the application falls outside the envelope (stay tuned for my next blog on application ranges)? Do the suppliers test for “off-label” uses of their component like car manufacturers do? Consider this, Car Brand X performs best at speeds under 60 mph. It can exceed this speed occasionally, but only for a brief period of time. If I live in Germany and commute on the autobahn, where speed is a suggestion instead of a limit, I should avoid buying Car Brand X, right? In that case, the supplier has told me his car’s limitations and I realize I’d better buy that Porsche instead. But, in our field, the product doesn’t always consider the process, and vice versa.

Along with walking a mile in another person’s component, why not try running at their pressure and temperature, too. This would go a long way to busting biases. But how do we promote collaboration and exchange knowledge?

More often now associations’ meetings are bending toward commercialization and vendor exposure, and away from just discussing the issues and solving problems. For example, at a large upcoming industry conference, I wanted to display a poster explaining the surprising results of side-by-side component testing. The poster was accepted, but I was required to pay thousands of dollars to register for the event, first. I did not.

Industry magazines have merged with our conferences as well, making them more commercialized and costly to attend. We are losing the forums for inclusive discussions about the facts. Pay-to-play excludes those who can’t contribute because they can’t or won’t pay the price of admission. Meanwhile, when these events are increasingly run by conference event managers, who are not subject matter experts, the tendency is toward copy-paste for popular presenters and the dynamic connection between what’s happening and what needs to happen in order to make better drugs gets lost. We aren’t going to change it, but we should be aware of it nonetheless.

The alphabet soup of the bioprocessing community (USP, PDA, ISPE, et al) has its own set of biases. Sometimes a few cross over the chaos in reaction to events and needs, but some also perpetuate the chaos by not working together for a variety of reasons: proprietary content chief among them.

But some headway is being made. BPOG and BPSA have partnered in the recent past with some promising output. The Single-Use Newsletters (available on this website) were well-received and many have asked for it to be revived and to include a related LinkedIn Group (and no, not just the millennials). ISPE’s Communities of Practice has also seen success in these types of symbiotic activities.

Even outside these organization, some collaboration is occurring. In October, The BioProcess Institute co-hosted a free Working Summit on Particulate in Single-Use aimed at gathering information on what we are doing and what we should be doing from both sides of the industry: suppliers and end-users (there were no millennials there). The discussion was lively and important. It’s a small start, but one nonetheless. Understanding each other’s’ biases is the first step to overcoming them. Penicillin for partiality.

In this country, states are allowed to pass and enforce laws separate from the federal government. Standards people often try to do the same, viewing a problem or solution only through the prisms of their own biases. But this isn’t the United States of BioProcessing. We should all be playing by the same rules and making improvements together. Drugs don’t have biases; they don’t care how they are made or within what they are made.  It is up to us–the Stainless People and the Polymer People, the suppliers and the end-users, the young and the non-young–to find the best way to make them safe and effective for the patients who need them. Period.

A Bone of Contention

Did you hear about the guy who went in for a hip replacement and the surgeon swapped out all 206 of his bones just to be on the safe side? Me neither, because that would never happen. Right? Then why does it happen every day in the bioprocessing industry?

In our industry, a single process can contain hundreds or even thousands of components working in harmony to manufacture a pharmaceutical product. But, although they work together, these components should not be treated as a set. In order to avoid excessive maintenance, it is important to follow these three simple rules: 1) know the process; 2) think like an individual component; and 3) apply sound risk management when establishing maintenance cycles.

When you understand the process, you are better able to look for trouble in the right place and not waste time focusing on the majority of the areas where the risk is low. Each and every component’s application range must be considered individually in order to determine its appropriate maintenance cycle. Just as in the practice of medicine, replacements and adjustments should always be made based on actual data. This helps reduce risk and cost at the same time because you are not opening up a system with a worst-case-scenario mindset; unnecessarily exposing it to contamination and replacing costly gaskets and diaphragms that are still able provide many more cycles of service life.

When the proverbial squeaky wheel is telling you something important, listen to it. But when the other wheels aren’t squeaking at all you need to hear that, too. If you think like each component, you will be less likely to treat them as if they are a set that needs to be changed out all at once; thus avoiding unintentional risk caused by exposing an entire system to change-outs which are only required in a small area. For example, a WFI pump outlet is always under much higher pressure than the rest of the system and has more extreme ups and downs (see chart). Gaskets and diaphragms located there naturally experience added wear and tear and thus the adjacent piping usually has more rouge. As a result, the pump requires more frequent maintenance cycles than the rest of the system. Similarly, in a Clean-in-Place (CIP) system, a valve may turn on and off ten times more often than every other valve in the process so it understandably needs more attention. This is also true for other parts like spray balls.

Understanding the cause of a failure can save headaches and dollars. A sound risk management approach uses testing to balance the real data collected before adjusting a maintenance cycle. Sometimes a component fails because it, alone, is operating outside of its design specifications. The point is to not overreact when failures happen. I know of a major player in the industry who changed their global policy on diaphragm replacement after a single diaphragm failure. In the desire by many to get to the root case, sometimes the kneejerk reaction is to design to the worst upset condition or the worst location. This response is even more stressful when it involves a nonconformance or CAPA. Avoid reacting too hastily and broad brushing implementation beyond specific areas of the process. Certainly, there are isolated cases when drastic measures are unavoidable, but, if you understand the process, think like each component, and apply sound risk management to address issues, global replacement will become the exception instead of the rule.

Every day at The BioProcess Institute we help clients navigate the maintenance cycles of their process lines by analyzing real data so that the process can be modified to treat each component or segment individually. We also analyze used components and failures and compare them to what we have already seen over the past 30 years of failure investigations and analyses.

Like the best physicians, our experience has shown us that asking the right question can lead to the root of the problem before drastic measures are taken. A surgeon would never open up a patient in a hundred different locations to look for problems and make replacements at each incision. The doctor would first perform tests followed by thoughtful analysis of the resultant data. They call it a “surgical strike” for a reason. The same goes for maintenance of a process: perform testing, analyze the data, and only then come up with a strategy.

At The BioProcess Institute, process is literally our middle name and we will work our fingers to the bone on your process issues.

Misapplications, Misuse & Misunderstandings

Have you ever tried to assemble a gas grill, a toy on Christmas Eve, or anything from IKEA? If you have, you might understand how that kind of confusion, pressure and frustration can cause component failure in bioprocess lines.

Manufacturers do their best to provide users with what they consider to be the proper set of assembly instructions for their product(s).  But, whether those instructions are for a two-story doll house, a trundle bed, or a bioprocess component, at the end of the day, if the instructions are confusing, or if the end user misunderstands what they are supposed to do with the parts, failure is inevitable: the tiny doorbell won’t work, the trundle bed will fall off its track, and the process line will leak and allow bacteria in.

That’s not to say that manufacturers are the lone culprits. End users can be to blame as well. For example, last year I led a customized training course on bioprocess component use, assembly and installation at a nationally renowned pharmaceutical manufacturing facility. One of the technicians in that class brought with him a set of instructions he downloaded from the internet on how to install one manufacturer’s components. That wasn’t a bad idea in and of itself. The problem was that the footer on the PDF that he brought was ©2004. That would be like using Ford’s blueprints for 1967 Mustang to assemble a 2017 Fusion Hybrid.

To complicate matters further, industry terminology can be confusing. What is the difference between percent compression and tightening torque? Can they be used interchangeably? Should they be? The answer to the last two is no, but that’s a discussion for another day. Likewise, in bioprocessing, we are dealing with components whose material has specific limitations, through no fault of their own. Misapplying these components can lead to failure just as surely as misusing them can. Do instructions always tell us this? Hairdryer tags warn us against showering with them because some bozo did and that. Should diaphragm valves have tags warning about the possibility of PTFE fraying. Probably. But how is one to know that this is a possibility without testing first?

The best solution to prevent misunderstanding and misuse of components is to test and verify, or consult an expert to do it for you. This goes for manufacturers and end users. If we do not understand the product and the individual parts inside and out, we are setting ourselves up for failure; failure that can be inconvenient at best and expensive, in terms of dollars and lives, at worst.

How many gas grills do you think the average person assembles in their lifetime? None, if they are smart, but in reality maybe one. Does that make them an expert? Probably not. But the employee at the hardware store, whose only job is to assemble gas grills every single day all summer long, might earn the title expert if he or she does it well, because they become intimately familiar with the product and the most effective and efficient steps in the assembly process.

At The BioProcess Institute, we test, we train, and we educate every day, all day long.

First, the performance and exposure testing we do brings us in tune with all types and manufacturers of components. We assemble, expose, disassemble and analyze bioprocess components constantly. We take manufacturers’ advice into consideration, but we don’t rely on it completely. If we find a more effective way to tighten their diaphragm valves, we tell them. If we believe, through meticulous testing and analysis, that their tightening torque recommendations are off, we tell them that, too. And we tell the client who entrusted us to find out.

Secondly, we train based on our testing and consulting experiences and this is what sets us apart from everyone else. In a nutshell, when something goes wrong, they call us. Time after time we answer the call; when something breaks, when there’s a deviation, for CAPAs, etc. Once we right the ship, we use each individual experience as weapon in our arsenal to help prevent similar occurrences in the future for other clients. But we don’t just educate like a pamphlet or video tries to. We train to instill proficiency. We’ve done the work and, by passing the fruits of that experience along, you reap the benefits.

We are completely devoted to helping our industry make better, safer drugs. Our consulting experience has led us to become the industry leader and we pass that expertise along to you. So leave the grill assembly to the guy at the hardware store and save yourself some grief. And leave the testing, training and education on bioprocess components to The BioProcess institute for the same reason.



Testing on Orphans

Now that I have your attention, let me explain. I have long considered gaskets and diaphragms to be the orphans of the bioprocessing industry. My analogy isn’t perfect, but bear with me. When measured in square inches, they collectively make up a tiny fraction of a process line and are nearly forgotten, but processes cannot function without them. Neglected and underestimated, they try to manage the task of creating perfect seals to prevent contamination and leakage. When something inevitably goes wrong, they are the prime suspects. It doesn’t matter if a system is comprised mostly of stainless steel or of single-use parts, when there is a failure, all fingers point to the tiny orphans that are just trying to do their jobs. The problem is, they may have been forced into a situation for which they were not intended.

The materials used to manufacture these orphans (elastomers and thermoplastics) will not change in the foreseeable future. It is important to acknowledge that they behave in different ways and therefore the applications for which they are used need to be suited to their inherent qualities. Most gaskets and diaphragms work well. If they fail, it is usually because of misapplication. Either they are asked to do something they weren’t meant to do, they were installed wrong, or they were used past their suggested service life.

How much trouble could a misapplied gasket possibly cause? Let’s ask the astronauts who took the last trip on the Challenger space shuttle. Well, we would if we could. But the fact is that the 306 ton behemoth they were riding in was brought down by an O-ring operating in a process condition outside of its specifications. The O-ring was not intended for exposure to freezing conditions and a rare Florida frost damaged it with catastrophic consequences.

In that case, the component and the process were not a match.

The process is king. End users and manufacturers must match process knowledge with component knowledge in order to ensure that the components do their jobs and the process lets them. Component manufacturers often don’t understand process stresses and data. Before installation, we must ask does this component meet the requirements of this process? Sounds simple, but how do we know when the answer is yes? How can we compare specifications against capabilities? Simple. Through experience or testing.

Experience is crucial for collecting knowledge about anything. However, in the bioprocess industry, the days of pilot plants are long gone and regulation resultant from experience doesn’t always serve a purpose. Harvesting the collective experience of the industry would appear to be the most beneficial, but there are forces working against this such as competition. Successful companies don’t want their competitors borrowing their process line playbooks.

Is it the end user’s responsibility to find the best gasket for their process? Or is it the supplier’s responsibility to understand how their product should be used and make the end user aware of its strengths and limitations. The answer is both, but how would such information sharing take place?

The natural evolution of experience takes us from guidelines to standards and eventually to regulations, but even these are not fully comprehensive and can lead to confusion. This leaves end users to figure it out for themselves by pitting their own budgetary restrictions against their suppliers’ marketing claims. Most component manufacturers don’t understand process stresses and data, and most end users don’t understand the mechanical design aspects for withstanding certain process conditions. As Steve Jobs once said, ‘Design is not just what it looks like and feels like. Design is how it works.’

The main job of a gasket or a diaphragm is to keep fluid in and bacteria and particulate out. As with terrorism, you can set up security protocols to try to keep the bad guys from crossing your border. At the airport, we take off our shoes, throw away liquids, and empty our pockets. In bioprocessing, we use clean rooms, gloves, quality control and sterilization. In both cases, precautions help us feel safe, but do they really make us safe? In order to prevent an airport breech, you have to think like a terrorist. In order to prevent a process breech, you have to think like a bacterium. You have to know your process inside and out to understand where problems can exist. This is most important with gaskets and other seals. A safe border equals a safer process.

If experience isn’t the answer, then testing is. As I mentioned earlier, misapplication is the biggest problem faced by end users. The best way to avoid misapplication of a component is to understand it better by acquiring reproducible data and test methods through performance and exposure testing. But unless the testing is robust and identical across the board, it isn’t valuable. To date, a standard method for testing does not exist, and if everyone is using a homemade yard stick you might as well not measure at all.

Certain industry consortia have emerged for one side of the equation or the other to address the issue of testing by borrowing tactics from other industries. But what they are asking of the bioprocessing industry does not necessarily marry process knowledge and component knowledge with reproducible data and testing methods because they are making rules from only one viewpoint instead of both.

Processes must be designed while keeping in mind how they are wed to the components they use. Qualification and validation are crucial steps in the design of your process, but proper testing can provide a foundation to minimize those steps and give you the rationale and understanding for the best design of your process.

Airports come up with rules after something goes wrong. We never removed our shoes at security checkpoints before a terrorist attempted to blow up a plane with explosives hidden in the heel of his shoe. Performance testing helps you understand your process and component limitations before an event occurs. Airports perform drills to think like a terrorist. We must perform testing in order to think like bacteria.

At The BioProcess Institute, we like to force failures in simulated processes to study the causes and effects. In our case, catastrophic gasket failure usually results in some great video footage, a few paragraphs in our BioProcess Performance Report, and a mop and bucket. In a pharmaceutical company, this same failure could jeopardize revenue, jobs, and, in the worst case, patients’ lives.

We help clients on both sides of the equation: process and component. Our testing provides data and our expertise interprets that data to produce understanding. If you want to know which gasket is ideal for your process, we can help you. If you want to know why, we’d love to tell you that, too.