Members of PMP News's Editorial Advisory Board focus on trends in package quality assurance.
Inspection technology and supply chain controls are helping to ensure packaging quality and product protection, as pharmaceutical and medical device manufacturers face demands for user-friendly packaging. Speaking on the issue for this year’s Industry Outlook are the following members from PMP News’s Editorial Advisory Board: John Bitner, director of packaging development, Watson Pharmaceuticals; D. Bruce Cohen, principal,
PackTechPlus LLC; Michael L. Forehand, principal packaging engineer, global engineering and technology,
AstraZeneca; Curtis L. Larsen, principal, Spartan Design Group; Michael H. Scholla, global director, regulatory and standards, DuPont Medical and Pharmaceutical Protection; Laura Bix, assistant professor, Michigan State University School of Packaging; and Nick Fotis, director, R&D–surgical drapes, gowns, & custom sterile kits, Packaging Technology Center, Cardinal Health. Karen Polkinghorne, packaging engineer and MDM specialist, DuPont Medical Packaging, joined the discussion at the invitation of Dr. Scholla. In this roundtable with PMP News senior editor David Vaczek, the panel discusses challenges of managing quality and cost.
Investment in advanced inspection technology has grown considerably. What impact is automated inspection having on quality?
Forehand: Traditionally, automated inspection systems were put in place in response to GMP requirements, such as for 100% inspection of labels to prevent mislabeling. However, we are expanding the use of vision systems in other ways to catch problems earlier on in the manufacturing process and correct them at the source. The emerging enhanced speed and capabilities of today’s inspection systems are enabling this. We are actively engaging our suppliers in increasing their use of automated vision inspection to support defect-free supply. We prefer to have automated inspection as far upstream with suppliers as you can get it because defects there find their way down to our lines and impact our efficiency.
Cohen: I’ve heard that FDA is looking to implement 100% inspection on a number of processes. Their thinking at this point is that it would eliminate rogue events on a line that you wouldn’t catch with periodic in-
process inspections. I don’t think it fits every single thing you do, but there are places for it where it is beneficial.
Bitner: It is component specific. There are a number of areas such as Bruce said where it is very relevant. We have sequentially numbered labels where the roll can be taken off the floor, put back in inventory, and returned to the floor very quickly by checking the last label number. So we have complete reconciliation of how many labels were used, how much product throughput was achieved, and how much waste is involved in the process. Visual inspection systems can detect a 2-mm difference in the height of a bottle closure. This capability allows 100% nondestructive in-line inspection. Off-line manual inspection is an albatross from the past. Results were always subject to variance based on inspector fatigue regardless of tightly controlled inspector rotation.
Fotis: I have been on both sides of the medical device supply chain, and when I worked for a medical films manufacturer, we looked at automatic detection of films for gels and cinders. The technology was available, but expensive more than 10 years ago. The biggest issue was whether the customers would understand the level of “defects” that would be acceptable, or would want to dial down the instrument to detect the smallest possible gel in the film. These small gels would never pose a functional problem, or even a cosmetic issue, but the concern was that many audits take place when medical device manufacturers send their quality folks out to assess the film supplier’s capabilities. Some of these folks don’t always understand that there may be such a thing as “a threshold defect”—one that has no bearing on functionality or cosmetics.
Scholla: There are some things automated inspection is perfectly applicable for and can be used with relative ease, but others where it becomes much more difficult. It is pretty easy to detect coating skip in the application of coatings. [But] automated inspection of material supplies can become very complicated in that sometimes the machinery gets more sensitive than you really want it to be. It can’t figure out what is normal and what is not.
Larsen: People are looking at this as a way of saving money. You have to take the time to understand what it can and cannot do, and properly validate it.
Scholla: When somebody tells you to do it on a process that it really isn’t suitable for, it only creates problems and there will be no cost savings, only a cost increase. You are dealing with inspection reports on a process where the automated inspection is not really appropriate.
Do increased rejects become an issue?
Bitner: This goes back to the definition of a defect. We have leak detection systems now that can detect to five microns but that may not be of value based on stability of a given product. Tolerances are calibrated at a level compatible with protecting the safety and efficacy of the product. Mass extraction with micro-flow technology enables more blisters to be inspected without related cost of product/blister destruction.
Cohen: You can’t have one test for all things. You cause yourself a lot of undue work. The pass/fail requirements won’t be the same for a PVC blister with a product that is not moisture sensitive as with a cold form foil blister.
Larsen: Now that we can find more anomalies, we have to ask whether those anomalies are really a failure. Are the pass/fail requirements changing because of the sensitivity of the equipment? We see films that are widely used that fail quite often in testing. Are you going to tell people they can’t use the film? There may be a flaw in how they are testing or setting their pass/fail requirements.
Is the scope of quality requirements on suppliers increasing?
Polkinghorne: A common issue with medical device manufacturers and their suppliers is misalignment of specifications. The end customer would like zero defects, but the supplier’s process produces a certain percentage of defects. Once an MDM and supplier reach agreement or alignment of specification values, both parties are better served. A best practice is for MDMs to use a collaborative approach with
suppliers when writing or revising packaging material specifications.
Cohen: I agree. Your wants and needs as a manufacturer of products doesn’t necessarily translate as the same needs of the supplier. Some time ago, I had a head of manufacturing who wanted no splices on label rolls coming to the packaging line. We said we can do that but it will change the price of the labels and you will never have a roll of labels the same size. We agreed on a specification of no more than three splices. That gave everybody buy in as to what would happen on the production line. There has to be a conversation to make sure that the supplier can produce what you need at a reasonable cost to keep you in production and keep your products on the market. Certainly, we as manufacturers have been pushing our suppliers to be more proactive in understanding their processes and control of what they manufacture in order to give us as close to zero defects as possible.
Larsen: Understanding the capabilities of the supplier’s process is extremely important, but also you both have to be singing off the same page on how you test for an attribute. Are you both doing the testing the same way using the same methodologies? If one uses an in-house method, the other an ASTM test method, the two will never meet. There is a lot of built-in noise with many of these tests. We see it all the time with ASTM F88 seal-strength tensile testing There are three different methods within the standard that you can use. You have to collaborate when putting together the specs and setting the pass/fail requirements to avoid miscommunication.
Cohen: The testing is a good point. I’ve gone through that with flow-through testing on aerosol actuators. The manufacturer was doing it one way and we were doing it a different way on incoming inspection. And never the two shall meet.
Scholla: Another consideration is regional differences and expectations. There are four very distinct expectations around packaging suppliers in the United States, Europe, Japan, and China. In facilities that supply packaging in China today, almost every one of their pouching rooms is a cleanroom environment with everybody looking like they are working in a controlled environment. You won’t find that very often in the United States except with pharma suppliers.
Fotis: I think that the expectation from medical device manufacturers has always been very high for medical packaging suppliers. The amount of added value that is put into the package mandates that packaging reliability be very high. The increasing expectation is that suppliers are not only running a validated production system, but that they have a handle on their variability and are taking active steps to reduce it. One example would be variability in film thickness. Tolerances that were acceptable to the industry several years ago are no longer OK. The more suppliers fine tune their processes and reduce their variability, the less our packaging equipment and process windows need to compensate for what might come in the door.
Forehand: Getting defect-free materials is critical to a lean manufacturing system. The more variability we receive in supplied materials, the less efficient our operations will be. The use of “AQLs” to “allow” a certain level of defects is seen as counter-productive. All defects should be corrected immediately at the source.
What is the progress in defining an allowable minimum hole size for maintaining packaging sterility?
Bix: The testing that we have done suggests that microbes can penetrate extremely small holes, sometimes even in the absence of pressure but certainly pressure, as you would expect, has an effect, as do myriad other factors. When reading anything that we put out regarding Ondrea Kassarjian’s or Jane Severin’s work [on hole studies], you have to put it in the framework of the testing that we did. We used extreme conditions. Another presumably worst-case aspect of the testing was the use of non-porous lid stock so the pressure differential has to be entirely relieved through the defect. We were trying to answer the question: Can microbes penetrate, not, do they penetrate. You have to view and interpret the findings in that context.
Polkinghorne: How did you quantify how dirty to make that environment? Is it based on a hospital central service or distribution environments?
Bix: It would be really valuable if we could characterize what is realistic in terms of the distribution channels and the hospital environments. We relied on other people’s literature in microbial challenge, and used an extreme load of microbes as our starting burden.
Scholla: A lot of work has been done on microbial concentrations in various environments but it is all published mostly in places you wouldn’t go to look for it, like building and cleaning magazines. A recent published study looked at the efficiency of vacuum cleaner bags and the impact of microbial burden in the air when you run a vacuum cleaner. OSHA has established that acceptable in-door air quality has to be below 1000 cfu per cubic meter. There is information out there, but many of these studies go back years and are hard to dig up. Recent literature has looked at the effect of new construction or renovation in a hospital on microbial load in the air in the rest of the hospital.
Bix: This work could eventually go in the direction of modeling probabilities to understand how factors impact a probability.
As Overall Equipment Effectiveness (OEE) is more often used by pharma, quality is a component of the evaluation. Is OEE being driven at the plant or corporate level?
Cohen: When you have multiple plant sites vying for more work, plants want to be seen as the most-efficient and most-cost-effective production center in the group, otherwise they may be in jeopardy. Generally, OEE is a plant-level decision as plant teams are trying to attract the new product work and keep their plant work from dropping off.
Forehand: At AstraZeneca, company-wide, lean manufacturing objectives are driven at the corporate level. Our company chooses to use OEE as one measure of lean plant practices and line efficiency. OEE is measured and monitored at the plant level. We have OEE targets for different types of packaging lines across the AstraZeneca network. We feel suppliers should also be using the same lean programs we use to put out a quality product, including OEE evaluations and enhanced automated inspection on their production equipment. As an example, one of our suppliers determined that if they doubled the frequency of preventive maintenance recommended by the equipment supplier, they ended up with less machine downtime and a higher quality product. These kinds of improvements have a big impact downstream. Automated inspection may initially decrease OEE at the supplier level, but in the longer term OEE should improve as problems are identified and corrected earlier.
FDA is focusing on packaging and labeling guidance to reduce medical errors. How can package redesign help minimize end-user confusion?
Cohen: Avoiding confusing labeling and packaging sounds good with one exception you don’t control. People don’t read labels.
Bitner: Exactly. I think there is an inordinate amount of information in the literature that is neither pertinent or relevant to the patient. The drug manufacturer is mandated to provide a road map–sized amount of literature folded multiple times and wadded up into a one-inch square. What begins as a massive insert becomes nothing more than a single sheet of paper when delivered to the patient at the pharmacy.
Cohen: Labels have to be clear and concise. You certainly have to be able to tell one strength of product from another. However, it requires people to read. For a long time, the ophthalmic industry had color coding of caps to denote what type product you have. Color coding of caps can be ineffective if another manufacturer is using a similar color. It has happened many times when people standardize on labels where you wind up looking at me-too’s on the shelf and people grab the wrong one. Also in the operating room, you want things clearly marked so dose strength is clearly visible and easy to read in a quick moment when someone has to grab something.
Forehand: Standardization of packaging components across our packaging network is seen as an efficiency enabler for us and our suppliers. However, you have to be careful if you are standardizing packaging across the pharma industry too much as you are increasing the ability for counterfeiting to take place. Also, you want packaging and labeling that differentiates products from each other so pharmacists won’t confuse them. Years ago, we had a drive at one of our sites for different color labels across all products and strengths as a way to help operators distinguish one product from another. The problem we ran into was we ran out of truly distinct colors. Today, we use different colors to differentiate between strengths of a product, but we use product name logos with differing fonts to distinguish between different products.
Bix: I think there’s a huge opportunity for standardization of labeling. It helps in that people could look for the information they need that’s most
germane to them in the same place. We are using an information processing model from the discipline of cognitive psychology to guide tests which we can use to actually measure perceptual and cognitive behaviors. Our goal is the creation of designs that are based on science so that we can enable people to easily encode and comprehend things, even in very chaotic environments. I think there are a lot of opportunities for quantified decisions in this space and hope the standards that emerge will be based on data.
Cohen: In a lengthy study some years ago with another university, we submitted an exhaustive amount of samples to people to see what they could read. We came up with a conclusion as to what colors worked best on the packaging and changed one whole marketing division’s graphics. Then new marketing people came in with a whole different concept as to what they wanted their packs to look like. That standard went right out the window.
Larsen: I’m working with subcommittee F04.22.28 “Formatting and Labeling” of ASTM International Committee F04 Medical and Surgical Materials and Devices. They are attempting to standardize a master label template for arthroplasty implant products. It’s a heroic effort chaired by an orthopedic surgeon from Wisconsin. You can’t argue with the premise: If you’re standing on one side of the room and you know you need a 12-cm hip stem, for example, the nurse can pick it out on the other side right away. The trouble is different companies feel their format is the perfect label and they know what their customers want. There is conflict, and different companies are hesitant in making changes or agreeing to a standard way if it doesn’t match what they are doing. This would be a voluntary standard, so how do you get people to agree? Be careful of voluntary standards. There have been occasions where, over time, they have “morphed” into requirements.
Fotis: Medical packaging departments have been validating the functionality of their packaging for many years. There are seal-strength tests, accelerated aging, simulated distribution, and visual inspection. A next step for the industry is validating the label copy of the package. By obtaining test data via voice of the customer research, a company can determine whether the “open here” arrows are quick to find and easy to understand. Surveys of simulated or actual users could determine how long a nurse takes to preinspect a package before opening it. How long does it take to properly orient it, to find the expiration date, and to confirm the product is usable, and then how long to actually access the device? Cardinal Health, like several other medical device manufacturers, has taken candidate packaging out to the field in facilitated focus groups to determine how actual users not only react to the functionality, size, and shape of the package, but also the labeling.
Scholla: FDA in a recent meeting and AAMI in an upcoming conference are focusing on human factors in medical devices that are intended to be reprocessed in the hospital. FDA is clearly concerned about instructions around reusable medical devices: how to disinfect them, how to take them apart and clean them, how to put them back together again, how to sterilize them. If I look down the road, I think that we will start seeing application across additional areas that FDA is concerned about related to the human factor element.
Are there advantages for manufacturers with the proposed One-Document approach where elements such as the package insert, Med Guide, and the pharmacies’ printed material would be combined?
Bitner: I’m not so sure about that. If all that information is in a single document, how do we print it, how do we assemble it to the package, how do we assure ourselves that the patient gets that wad of literature, and how can we as the drug manufacturer ensure that they are going to read and comprehend everything provided. The right people are more likely to get the information intended for them with separate pieces of literature. It is way too cumbersome to squeeze all the copy, graphs, and drawings required into a single document. My view is combining all the literature into a single document is heading in the wrong direction.
Cohen: I worked with a PhRMA task force for over ten years with FDA trying to get them to approve an electronic package insert, to get it off the package, get the paper out of the way, and get it electronically available where it needs to be in the dispensing pharmacies. There is something currently in the works at FDA to prepare a proposed rule change to allow that to happen for manufacturers. At the same time, there is another industry group working with FDA to place the Med Guides, the wallet cards, all the laymen’s information that’s given to the patient when they receive the script off the package and into the same software at the pharmacy that prints the label. That’s all well meaning, but if somebody doesn’t have the conversation with the patient, there is a disconnect. The interaction with the pharmacist who is giving you that bag is the last chance you have. Sticking it on the vial doesn’t mean it will be read. As John has said, we used to have to put two or three Med Guides attached to a leaflet attached to the bottle, and I can almost guarantee you that none of the patients got any of the information at the pharmacy. I have yet to receive a Med Guide in unit-of-dispense packaging from mail order. There are two glue spots where they ripped off the leaflet. It is replaced by a printed version that is not exactly the same level of information.
Bix: Studies have suggested that when people are inundated by information it can become overwhelming. As such, I would support the idea of concise and standardized information that enables providers and patients to access information that is critical to their needs. That said, I think it is important to carefully consider design decisions. For instance, we recently did a study that investigated the effect of color on the ability of patients to notice the prescription warning labels placed on vials at the pharmacies’ discretion. An eye-tracking device was used to gauge patients’ ability to notice these colored strips. We compared the behavior of a panel of college-age students with a 50-plus age group. The differences in their behaviors were quite startling. The older population just sat on the white pharma label, where the younger people very actively rotated the vials and looked at the warning labels. Possibly the Internet generation is less linear in their processing. I think that it is very important to understand the needs and behaviors of the patient in order to make informed decisions.
Cohen: Back in the 1980s, the Giant Food chain put in a new—and costly—labeling scheme for their OTC product lines with pictorial information on the labels and signs on the shelves. It drew attention to the things they needed to find on the label, and helped non-
English-speaking people. Unfortunately, it didn’t get picked up by anybody else.
How is downstream quality being addressed, such as when manufacturers’ qualified packaging is cast off in the “last mile”?
Larsen: What makes me cringe as a medical device packaging engineer is there are times when we do not know what happens to the product once it gets into a large distribution warehouse. They may be grabbing sterile packaged units from a multipack box and combining them in a distribution box with other product for a specific point-of-use. We need to understand when the original packaging system is no longer in place and if or how package protection might be
Scholla: One of the questions I always ask at our seminars is how many of our packaging engineers have actually gone to a hospital and watched their product arrive. Usually, one or two out of 40 or 50 say they have done this. And I ask: What did you learn? No cardboard goes upstairs. People design shelf packs for shelves, but in many instances the shelf pack never makes it to the shelf. It is dumped into a plastic bin that fits perfectly on the hospital’s organization shelves. While we wouldn’t like that to happen, there is another viewpoint. If this is what normally happens, you have to design your sterile barrier system to meet those challenges. Also, you will have one-off situations. A box is left on a loading dock in Spain over the weekend, and it rains. By Tuesday morning it has mold spots on it. The label says sterile unless opened or damaged. Is it still OK?
Larsen: Often times we do not understand our distribution cycle. We will do a particular ASTM D4169 Distribution Cycle stress test on our packaging system to judge its design performance. In reality, it may not represent what it will be exposed to once it goes out the door. There is not nearly enough thought, money, and time devoted to address these unknown downstream issues, and it is frustrating because it is a significant concern; they can and do cause damage in the field. There is very little control of the packaging system once it leaves your finished goods.
Package coding requirements to support package traceability are in place in France and Turkey and coming in the United States. How active are companies in deploying in-line coding at this point?
Cohen: When somebody comes forth with a set of standards that the industry can live with on both the manufacturing side and consumption side, this thing will happen. As the rest of the world leaves the United States behind, it is going to be a form of catch up and then you will have more than one type of system being used globally. Hopefully, we will not have to live with that. It would be nice to have a standard system that all of us could use. Standards development is taking way too long.