Lately I’ve received an increasing number of new and ‘critical’ label tests and requirements that we need to deal with. As we all know, in terms of device cost percentage and engineering time and energy spent, the label is the proverbial flea on the dogs’ tail. Yet the label is still a critical component that needs to remain ‘affixed and legible’, as a well respected government agency puts it. To that end, more tests and standards have sprung up to make sure that those two simple criteria have been met. In this post I’d like to examine a select few of those tests, figure out if they actually provide any value as far as a label remaining affixed and legible on a terminally sterilized device package, and then make fun of a few that I think we all can agree border on the ridiculous.
In order to get a bit more ammo for the latter of the three blog post goals, I sent out an email to a few engineers and others that I’ve actually discussed the topic with over the years. I asked for, “….a bit more label testing stupidity than I have at hand so I can award an Olympics-like gold, silver, and bronze in the blog post. Plus, it will be kinda fun and a bit of needed catharsis for us all”. Not surprisingly, the response to this email could be described as a few levels above ‘brisk’. Both the number and promptness of the responses after I sent the email would lead me to believe that there is a bit of frustration among the package engineering community about the efforts needed to confirm ‘affixed and legible’. Some of these tests are not exclusively for medical device applications and a couple wander from pure label testing, but they are fun anyhow. Thanks to all who submitted them.
These test puzzlers fall into three broad categories. ”If X is good then 2X must be better’, ‘if it worked for the apples it must work for the oranges’, and ‘it must be 3.684 +/-.001’. A rub test for a paper label in a OTC environment had been set up at 200 rubs (Sutherland tester) with a 2# weight. Someone in the company decided that if that was good, a much better and more robust test would be 1000 rubs with the 4# weight. Would a package ever encounter that? Would it even be subjected to 200 rubs with two pounds of weight? The best part of this was that water was not enough; a second test of 1000 rubs at 4#’s with Coca Cola was performed as well. I wonder if the test would have been invalid had Pepsi or RC Cola been used? We will never know I guess. Finally, would a conditioning cycle like ASTM 4169 be enough to ensure that this package made it from the Acme Pharma plant to the store shelf? I’m going to go out on a limb and say chances were probably greater than 95% that it would. Yet five respondents to my little email note said that someone in their organization did not feel that the 24″ drop in ASTM 4169 was ‘robust’ enough, even with the 48″ bottom drop added on at the end of the sequence. Once again if X (24″) is not good enough, then 2X (48″ on all sides) must be better. One quick thought that was expressed regarding ASTM 4169; if the label passes that conditioning cycle, why bother to initiate more stand alone tests? That same ‘double x-ing’ goes for temps on accelerated aging tests and dosages of ETO or radiation in sterilization cycles. One respondent remarked that he loved the ‘triple gamma’ test. “You turn your plastics into dust but damnit, the label is still legible”.
The ‘if it’s OK for apples, it must be for oranges’ test that several people commented on was the venerable pouring of water on a paper label, letting it sit for one minute, and then wiping it off. That particular test is straight out of BS EN 45502 for active implantable devices. One quick observation: if you make ortho implants, stents, catheters, etc., why are you using the active implantable standard? The main subjective flaw that I’ve witnessed in this test is the line, “rubbed by hand, without undue pressure”. I recall my mother grabbing me, a dirty faced kid, and attempting to clean my face using the end of a handkerchief and ‘mom spit’. I’m sure she felt that the pressure was not undue but to a little kid it felt like she was using 80 grit wet/dry sandpaper as I tried to escape her headlock. My guess is that if my mom were performing this test nothing would pass. In my humble opinion the above mentioned Sutherland Rub test is much more objective and accurate if you really need to perform the test. The other issue is that you are pouring water on paper. It always leaves a watermark. If you peruse ISO 11607, the definitive guide to terminally sterilized package testing, you will find the following: 18.104.22.168 – Opened package samples shall be inspected for defects such as: d) the presence of unacceptable humidity, moisture, or staining. Soooo…..if you developed a label that survives the water test and it actually gets exposed to water and stained, would it pass this ISO 11607 criteria? I don’t have the answer to that. If you really feel you need to pass this test with flying colors, just switch from paper to white polypropylene, BOPP as it’s called. It will ratchet your cost up several levels but if this is indeed a critical part of your package evaluation, cost is no object, right?
Finally the ‘it must be 3.684 +/-.001 type of test. In the good old days we used to get prints generated by mechanical engineers with the same tolerances as a piece of machined steel. They did not know our process capability and, frankly, neither did we. The saving grace was that they really didn’t measure it after the FAI anyway or didn’t use a high tech, calibrated tool anyhow. Things have gotten a lot more precise and controlled since then, but in many cases involving package testing it is much more sensible to have the criteria set up as pass/fail rather than attached to a certain precise value. One personal experience with this issue was not a label but the Tyvek lid that it was affixed to. We produce Tyvek lids, printed and unprinted, at Advanced Web and have since the late ’70’s. There was a merger/acquisition and suddenly one of our customers was a small cog in a big machine that had much more ‘robust’ quality standards. The large company had a maximum porosity reading of 130 seconds in their inspection procedure for Tyvek lids. Absolute max,c=0, period. All of a sudden, after a perfect quality 100% record for three years, we started getting lots of Tyvek lids, perfectly functional Tyvek lids, rejected and held hostage in quality hold. The problem was the c=0 thing rather than the average of several readings, either on one lid or on the sample. If you look at the image of a Tyvek lid (above) you can see the thickness difference, the ‘ropes and veins’ of the varying caliper of the Tyvek. I would like to bet anyone a case of microbrewed beer right now, that if you stuck that Gurley meter right on the dark spot in the top center of the image, the reading would be in excess of 130 seconds. Summit Extra Pale Ale, please. Of course it’s a useless requirement because the ETO gas will flow to the lighter, thinner spots and sterilize the device. If the entire lid is not porus enough, our guess was that the seal would blow. A noted Dupont technical person, since retired, agreed with our assessment and after a process that was more painful and drawn out than a presidential election campaign, the inspection criteria at MegaMedical was changed to a more realistic and useful process.
I have one more test, a test that defies categorization in any of the three I’ve outlined. I will have to give it the Gold Medal. This story also requires you to paint a mental picture of this test in progress. A group of company managers and directors are standing in a conference room and throwing pouched trays holding 100cm Guide Catheter against a wall to see if the snap fit on the tray worked to hold the device. I’m sure it was not described and codified in any official company test protocol and I would hope the label held up under the wall impact. My guess is that no ASTM standard will be balloted any time soon for that test. A similar test was for a device packaged in a polypropylene box that failed the “sit in the rear window of the sales guys locked car in Arizona in June” testing protocol. We actually addressed that one and found a suitable adhesive but that’s another story.
That is indeed the bottom line after all. If you have a test protocol and share it with us, we will do our damndest to come up with a label that works. Adhere to the nose tiles of the space shuttle, survive re-entry, and be easily removable upon landing? Let us give it a shot. However sensible testing and qualification protocols, based on peer reviewed and industry approved standards such as ASTM, ISTA, TAPPI, and ISO standard to name a few, are really the way to go. Dr. Mike Scholla of DuPont Protection Technologies pointed out at an ISO Technical Committee meeting a couple years back that “it is OK to use non-standardized test methods, but if you do you may have to show how it compares to the standard method. And since you have to do that work anyway, why not use a standardized test method if you can?” You don’t need to go rabbit hunting with an elephant gun and the word on the street is that cost reduction is the mantra in many device companies these days. There are a number of ladies and gentlemen in the tough environment of Barcelona Spain, as you read this, toiling away diligently on developing and fine tuning various ASTM standards. As well as some possible wine testing. As I pointed out in a previous post, it’s time to come up with a unified guidance on testing the labeling for terminally sterilized devices. We need to have a series of tests that will reliably predict that labels will remain ‘affixed and legible’ to meet FDA standards. Email me please; let’s see if we can get something moving forward on this continuing issue. I get a real sense that we have the resources; we just need to pull them together.