Efficient Packaging for Global Markets
Companies are challenged to maintain production efficiency as trends from new global markets to personalized medicine lead to smaller lot sizes. Against the back drop of a changing business model for pharmaceuticals and regulatory expectations, a Quality by Design approach by pharma and medical device firms can yield a deeper understanding of product needs and packaging requirements, leading to more options in materials and processes.
Speaking on the issue for this year’s Industry Outlook are the following members from PMP News’s Editorial Advisory Board: Jordan Montgomery, technical fellow, Medtronic CRDM Package Engineering; Dan Penny, director of packaging engineering, Cardinal Health; Laura Bix, assistant professor, Michigan State University School of Packaging; Rich Hollander, vice president, packaging services, Pfizer, Inc.; Michael Forehand, principal packaging engineer, global engineering and technology, AstraZeneca Pharmaceuticals; Michael H. Scholla, global director, regulatory and standards, Dupont Medical and Pharmaceutical Protection; D. Bruce Cohen, principal, PackTechPlus LLC; and Curtis L. Larsen, principal, Spartan Design Group. The roundtable is moderated by PMP News senior editor Dave Vaczek.
Q. Is low volume packaging comprising a larger part of companies’ product mix?
Larsen: Low volume packaging will always be there especially with new start up products. You are always starting with low volume and if marketing hits a home run you are up and running on high-speed form-fill-seal (FFS). But you have to have ways of handling these low volume, limited-distribution type products that we all deal with in a regulated industry.
Penny: There will always be a certain portion of the product line or the startup codes that will be low volume. We tend to do many more high volume products, but there will always be a need for low volume packaging.
Forehand: What we see is a lot more growth in some of the global emerging markets that are lower volumes. We want to be able to provide these markets with packs in their local language which meet local requirements. There are opportunities there that create challenges in the packaging space.
Larsen: Low volume packaging is all in the eyes of the beholder. If an OEM with a $100,000 defibrillator sells 100,000, that’s huge. If you are selling an experimental LVAD device that costs $1.4 million to put in and monitor-- 20 of those a year is pretty good.
Q. Low volume packaging could comprise new product for which demand is uncertain. You also have low volumes when drugs lose patent.
Hollander: That might not translate directly to small lot sizes; rather that might translate to a decrease in the frequency of runs. It will depend on a multitude of variables-- manufacturing lot size, inventory carrying costs, etc.
Scholla: As devices or pharmaceuticals come out you have the low volume part at the beginning but you also have companies that make high volume products. How do you make those material choices and packaging form choices? Do you start with pouches and eventually switch over to FFS? Do you use the same materials?
Penny: In some cases it depends on capital. If we have capital sitting on the floor in FFS machines and we have capacity, we may choose to purchase a die that will make that other size of package. If we don’t have capital on the floor it might be more likely we make a decision to go to a pre-formed package early on until we can justify the capital. Capital often helps drive that decision.
Scholla: And your choice of materials? When you are coming out with something new and you know it will be low volume, do you pick new packaging materials or do you go with the old standbys?
Penny: If time-to-market is urgent and we have aging on a material combination, that would receive a logical preference. We can use an existing aging study and not have to repeat that before launch. We have Tyvek pouches that we have accelerated real time on that would improve our time to market. If there is a requirement for the product that would preempt us from doing that we would be forced to look at a new material even if it was a pre-formed package.
Larsen: You hit the nail on the head. If you can use something that is pre-existing and you don’t have new device fragility requirements that you need to address you are going to go with what is proven, what works, and what you have specifications on--toxicity, biocompatibility, and stability data. Until something pushes you out of that area, why wouldn’t you use the data you already have.
Montgomery: We would do that whether it was going to be a high volume product or a low volume product. We focus on the requirements of the product first. From a production efficiency standpoint it is exactly the same thing. We would have our process capabilities and production efficiencies set for high volume or low volume products with the same criteria in mind.
Penny: The only thing that we would maybe do a bit differently is if we come out with a new product line that we know is going to be high volume. We may be looking at what kind of cost savings we might be able to get. If we have to put in a new FFS line, does the product have characteristics that would allow us to come right in with a less expensive material. Or maybe we go in with the standard material as you said and we are already working on something that is lower cost, because we know it is going to be a real high volume product.
Larsen: You are absolutely right. If you know that this is going to be a home run, and you have to tool up, start ordering FFS way ahead of time for your off shore plants.
Penny: Absolutely correct. If it is low volume we are probably mostly looking at time-to-market and something that will meet the requirements and not try to drive cost efficiency right up front.
Cohen: People have different definitions of low volume. If you are putting out a cancer drug the number of units you may sell would be classified as low volume but the dollars you generate in sales would not be and that may have an impact on the decision you make on how to put that pack out on the market. A lot of my clients are small-to-medium size with volumes you might consider low; they would not. In some cases they don’t have the wherewithal to do it so they are going outside to CMOs. That has an impact on their packaging decisions.
Forehand: It depends on your packaging process too. When we talk about low volume, we are talking about production order sizes where changeover time begins to creep into your conversion costs to a higher extent than you would like. That’s when we begin to think about new ways of working.
Hollander: That’s how Pfizer views it as well. A simple rule of thumb is when change over time exceeds your run time. And quite frankly we find that less in the US and more in emerging and our European markets.
Forehand: You might have a similar pack you make for the US market as you might make for an emerging market but if you try to run the emerging market’s volume on the same asset the changeover time between product strengths gets to be longer than your run time. For the US market you never really have that problem.
Hollander: What you are alluding to is our business has been evolving to require an increasingly higher proportion of lots that are low volume. This brings the equipment mix that we have in our sites into the spotlight. Do we have the right type of equipment for today’s demand versus the demand that we had in place when we originally purchased the equipment.
Penny: For international markets, the question for us is are we creating a separate international code because the product is different. For us in many cases it has been because of the labeling requirements. In some cases to minimize running the low volume codes we have gone to a global print where the same product we sell in the US has full global labeling on it. We have tried to minimize some of the lower volume runs with this approach. That is different than if it is a different product requirement for the international market.
Hollander: Most of us have similar strategies in place where we try to maximize a particular platform, for example, common blister card footprints or bottle sizes to minimize changeovers. We can further minimize the impact of small lot sizes by being clever about how we label product by looking for the opportunities to leverage common labeling—same package for multiple markets where we list all required languages on the same pack.
Forehand: That requires some standardization in your packaging as well. You need those (multiple) markets to take the same blister design or number of tabs in a bottle.
Penny: In some cases you run into geography issues where you don’t have enough room to do that. You then wind up with the lower volume codes due to the print requirement.
Hollander: We have been fairly successful as a company knowing where and how to leverage both shared and common packs, particularly in Europe and Latin America to help offset some of these smaller lot sizes in particular markets. We are doing what we can but we are concerned as to where things are headed with respect to serialization and the lack of uniformity in the coding standards. Technology does not allow you to put more than one serial number on a pack. If you have three different codes for product identification on the pack required by the health authority for the three respective markets we are sharing the pack with, and they also start requiring to tie a serial number to the product code in the three barcodes carrying the product code, we will need to revert back to individual packs and have even more smaller lot sizes.
Q. What is the opportunity to consolidate products on lines that may be partly idle?
Hollander: There is no easy answer to that. I think it is safe to presume that all companies are looking to optimize their asset base so if we have under-utilized equipment within a site or across our network, we consolidate volume to increase the asset’s utilization. This sometimes means moving products between lines, sometimes across sites, and sometimes if it makes sense replacing/augmenting the equipment we have to better match the demand profiles of our products. There are a lot of different ways to make more effective use of your assets.
Forehand: Our equipment manufacturers for blister machines for solid oral dose packaging have done a really good job in improving the design of their equipment to support faster and faster changeovers, and I am sure this holds true in other types of packaging equipment as well. But what we see today is a lot of times equipment changeover becomes faster than the actual procedural changeover. It doesn’t take an hour-and-a-half any more to clean a blister machine feeder and change the tooling. They have automated the process to make it simpler and faster. Now the challenge for us becomes to change our procedural ways of working to match how quickly we can mechanically change lines. We need to continuously improve how we should operate with low volume.
Q. Can existing equipment be upgraded for more efficient for low volume production?
Forehand: There is a certain level of improvement you can make to equipment to make it changeover faster. That is something you are going to do regardless of what your order size is. Improving overall equipment effectiveness is constant across the entire volume spectrum; you will work on improving changeover as much as you can. The point here is that even our best efforts haven’t got it into the range where we are satisfied with how changeover time impacts our conversion costs for low volume.
Penny: I would agree with the procedural issue. We took a look across all of our plants where we had FFS lines and did a time mapping from shut down to start back up of the next order. We did this because we were looking to lower our order quantities to reduce inventory levels as part of our lean manufacturing efforts. Certainly there is more opportunity with the new technologies for faster changeover of the machine, but I would agree that we found that there was a lot more procedural downtime from shut down of one order to start up of the next than there was actual mechanical change over. By using a pit crew-type of approach where people were doing things concurrently--clearing paper, changing ink colors--we were able in some cases to reduce our downtime by 60 to 70% just by changing the way we worked, without changing any capital.
Q Does the trend to lower volume make outsourcing more attractive?
Cohen: The majority of the small-to-medium size companies I deal with do very little of their own manufacturing and packaging. For one client with a new product the CMO did everything for them—manufactured it, packed and labeled it, and shipped it out. They had an emerging market order where they wanted 500 units for an annual usage in that market. It was a problem getting those made without losing your shirt because they couldn’t take 500 units out of their US stock and re-label it. Do you do a separate order, do you go to bright stock and have someone else do the packaging and labeling. Many of these new products for emerging markets tend to be very small orders for the first couple of years until you establish yourself and actually have a product line with increased sales. It is not a Pfizer world or Merck world where with some products out of the box you have hundreds of thousands or millions of units you are selling.
Hollander: That is not really the case these days any more Bruce. We don’t have many blockbusters these days. In terms of the direction of big pharma, we see an increase in precision medicine targeting populations with certain genetic profiles for which we our products will be most successful. The lot sizes tend to be very small and the overall volumes tend to be very small as well. We are learning how to manage the complexity that comes from these types of products as the volumes are comparatively smaller and for all markets, US included.
Cohen: I’m dealing with volumes now where less than 25,000 units is a production lot.
Hollander: Yes, we are talking about the same thing, as many lot sizes are well less than 3000 packs. A point I would like to make is we need more of our OEMs to start understanding that the business model for pharma has changed. I think there are a lot of technologies they need to embrace faster that will help us with managing these small lots, such as making use of our digital assets when it comes to printing, quick change machines, and machines that will allow us to postpone production as late in the process as possible and allow us to get it to the intended market/patient quickly. It is not just that there are small lot sizes; ideally what we would like is to delay the packaging, not just the labeling, until we actually have a demand for the product from a particular market. If we package product even as bright stock we are tying up money in inventory; we are only buying ourselves flexibility in which market we will target the filled package. When dealing with such small volumes and expensive products, it is optimal to fill/label/pack to a specific market demand if at all possible.
Q. Digital printing becomes more cost effective when dealing with lower volumes?
Hollander: There are a lot of technologies in the area of digital printing. We have been leveraging digital print technology for 15 years on lid stock and 20 years for shipping case labels! Over those years technologies have evolved in this space very well so now we can digitally print tiny bar codes as well as six-point font on knurled blister packs and have it readable. When we move to cartons, are we talking about digitally printing cartons on line, at/near line, or having the carton supplier digitally print in their own site to match our small lot size demands. The same thought process applies to leaflets. There are a lot of different ways to skin it and each one of them with different potential technologies to support it and different points where one method makes more sense financially from a total cost perspective than another. Each company needs to define how it wants to leverage the technology based on their demand profiles, printed component type/sizes, equipment base, and capital plans.
Larsen: Last year during the HealthPack 2012 conference we visited Ethicon Endo Surgery in Albuquerque where they have print-on-demand set up for all their instructions for use. It was simply amazing—a truly versatile system. They ran what they needed for each job individually. You can make changes on the fly for each lot and you don’t have a lot of waste.
Q. Are companies adopting a Quality by Design approach for new products?
Montgomery: Very much so for our organization. We are designing for liability of products and designing for manufacturability of products. QbD is being pulled into product development at the earliest levels.
Larsen: This has always been a part of any good engineer’s process--coming up with the most effective, safe, and functional package. It is getting a lot of focus now and that is a good thing, but if you were an engineer worth your salt you were doing this 30 years ago.
Penny: I agree with you Curt. We’ve always done it. In some cases, we have formalized doing FMEAs (failure mode and effects analyses) as a part of any packaging project. So we have put in more formal techniques but we always would be looking up front at potential risks and building in the controls for those.
Larsen: We have formalized tools of measurement and that is all good. But I think it has always been a part of industry.
Hollander: We design our container/closure system to be fit for use; namely to protect our products for the environment, be able to deliver the right dose and not interact with the product. In pharmaceuticals, most companies are now working to better understand the specific protection needs of a product so they can optimize what type of container closure system is to be used and leverage that product understanding to use good judgment in pursuing post approval packaging changes to minimize unnecessary stability programs.
Forehand: Our thinking is right in line with that. We are looking for the opportunity to understand the product better and that will allow us to optimize our packaging. As a conservative industry, pharma has probably over packaged in a lot of cases.
We are trying to think creatively about how to best protect the product keeping in mind the product characteristics and risks. This allows us to advance objectives like sustainability, reducing pack materials by weight, and reducing the amount of material we buy to provide cost savings.
Bix: I had the opportunity to attend a conference earlier this summer where Dr. (Janet) Woodcock spoke and touched briefly on QbD. The thing that struck me that was maybe different from traditional approaches is they are starting to engage the end-user behavior more than they have historically. Where it has always been about product protection, product efficacy, product safety; they are talking about the user being a part of the risks and the user being part of the benefits, and how can the system deliver not only from a product protection standpoint, but how does the system interact with the people to deliver what it is supposed to do. My spin on that is that FDA is moving toward a more inclusive model in terms of what the system needs to deliver. How will people use it, how will they interact with it, and will it deliver from that end of the spectrum as well as what we have traditionally focused on.
Larsen: Is this the human factors part of the equation that is certainly getting more and more focus, with the new human factors standard that has been written? This is really important.
Bix: Yes, I think they are rolling it up into QbD and saying safety and efficacy is not just about the interaction of the product and the package. But the package also has to be used in the way it was intended to be used. You are looking more at total system risk and total system benefit and testing or evidence that will be more robust than it has been in the past.
Montgomery: Laura, what you have said regarding the user very much aligns with FDA’s whole design control philosophy in Part 820. You need to establish your users’ needs and the intended uses of the product and that has to flow down through each level of design via design inputs and design outputs. It is very clear that is the expectation of FDA and it is how you conduct design validation. You need to show that your final product including packaging is mapping back to your users’ uses of the product.
Q. As you derive a better understanding of risk, or the edge of failure with QbD, this creates more options for packaging and processes?
Larsen: Yes. What is the fragility level of this new artificial brain? What do I need to do to package it most efficiently so it will get there but won’t cost me an arm and a leg. We often times don’t know what that fragility level is. As a result, we’ll over package it. So we’re not doing our very best as an engineer in optimizing the cost and maximizing the protection because we don’t know what it takes, how much protection the package contents need.
Hollander: If we can deepen our understanding of our product packaging needs we can further optimize our packaging to use less packaging--lightweight bottles, down gauge film thicknesses, etc. One thing we are progressing internally is a modeling tool which helps us predict how our product will react at various temperature and humidity conditions over time with any known container/closure system we have qualified. The tool leverages data generated from experiments where we look to determine how long the product can withstand extremely aggressive temperature and humidity conditions for relatively short timeframes. With this knowledge, we can accurately predict what container closure systems we need to use in order to obtain a target expiration date for a specific ICH climactic zone.
Scholla: QbD is not a new concept, it has been around for ever, right? For me this comes down to two questions: How much more time do we spend documenting for appropriate regulatory agencies around the world that we design quality in? And, though we may be over packaging as has been discussed, is it really worth financially redesigning packages that then under go additional scrutiny because they are new and different?
Forehand: In understanding your product needs you are gaining flexibility for post approval changes, where you are lowering the cost and complexity of those changes.
Q. Pack-to-order is an approach to the small lot size trend for global markets. What is the outlook for medium-speed, high-flexibility machines?
Hollander: The need is definitely there and the equipment manufacturer base has certainly stepped up their offerings in recent years. I would say they are still all in the low single digit serial numbers in terms of number of pieces of equipment delivered to customers at this point. We are at the very beginning of the life cycle for this type of equipment.
Penny: We are seeing the need for some of the newer lower capacity lines and we have in some cases invested in Multivac FFS machines for more flexible production.
Forehand: One of the challenges is getting the equipment manufacturers to extend the reach of their technical support around the world, especially in the Asia and the South America regions.
Q. Are new markets prompting evaluation of new primary packaging materials?
Hollander: ICH Climatic Zones IV A and B have certainly required pharmaceutical companies to increase barrier protection for some products. These increased barriers are expensive and do impact cost of goods. Pfizer would like to see a new generation of high barrier films come to market with a lower cost.
Forehand: As with the equipment tech support, we would like to have more global supply sites from the pack material vendors.
Penny: There has been a move in that direction with some of the major global packaging suppliers who are more active in Asia. But I would agree; it is not enough and it needs to continue. We are looking at regional suppliers as opposed to counting on the global suppliers for Asia markets.
Q. Are extended cold chains promoting reevaluation of tertiary packaging and supply chain logistics?
Cohen: It is a question partially of economics. For domestic cold chain shipping we are usually working 2 to 5 days maximum. Depending on what you are packing, you can sometimes get two-day delivery and it usually works very well. If you are going outside the borders, you really have very little control of the pack after its leaves your door. Even if you use a special carrier, customs can hold you for weeks or longer. If you have a non-mechanical environment (cool packs) and you get stuck in a customs location, there goes your product. There is always an economic decision to be made: do you spend the money for a self-contained pallet load container that in worst case can be plugged in if there is a delay or that runs on its own fuel or do you cut that price and go to something that you build around a pallet in the hope that it makes a certain time delivery. Where do you justify spending $10 grand for an active unit or something under $5 grand for a pack around a pallet. The passive works well, until it doesn’t, then you wind up losing that shipment and what does that cost you?
Scholla: You bring up a very good point. You qualify your packaging based upon your distribution. The moment you move outside the US and Europe it is pretty hard to define that distribution cycle. There are so many idiosyncrasies that can occur. So it is very difficult to go to something cheaper because you don’t know what’s going to happen when its gets to Penang, Malaysia or Shanghai.
Cohen: One of my clients is shipping a pallet load of refrigerated product that is upwards of $800 to $1,000 a vial. I’ve got significant dollars invested in that product. Most of the time an active container will be reliable even if you are delayed. That is more insurance than a passive solution packed for a specific time period. Where I used to work we had a truck that crossed the Canadian border and sat in traffic because of a storm, froze up totally, and we lost the whole truck.
Scholla: The whole issue ends up being you are qualifying the packaging based on a distribution cycle that is not definable. There are way too many variables.