Packaging problems

Here at the happy medical device factory we've run into a couple problems, but I think we've worked through them in not ideal ways, but the best we can hope for with limited time (sterilization date is quickly approaching) and money. As part of packaging validation we have decided to run dye penetration, burst and peel tests on our standard Tyvek/film pouches, once after sterilization and once after both sterilization and a shipping simulation (ASTM D4169). To hedge our bets we also sent some samples off for a preliminary shipping simulation and test before sterilization. This is all pretty standard, we're being a bit more careful than I think you have to, you could probably get away with not testing after sterilization, but we like to have the extra information in case of failure.

Our first snag was the burst test (ASTM F1140) results on one of the pouches, the burst test basically consists of filling the pouch with air and measuring the bursting point. We sent off 25 samples and the results were not wonderful. The samples were put through a shipping simulation, then we sent them on to a testing lab for burst strength tests. Two of the samples had rather low burst strengths, 2.5 and 4 psi, while the average was 13.5 psi and the standard deviation was 6 psi (all numbers ballparked). The good news was none of the seals that broke were our seals (we seal one side, vendor seals 3), and as expected the lowest numbers came from the peel away seal. The bad news is obviously no consistency and a couple very low results. As far as we can tell, neither the FDA or the standards set any pass/fail criteria on this test.

How did we solve this problem? After various discussions with QA and the testing lab we decided that since the burst test and the peel test are basically measuring the same thing (Tyvek to film seal bond), that we'd focus on the peel test. Like I said, not an ideal solution. The peel test is more reproducible and reliable, and we easily meet the generally accepted standard (greater than 1 lb per inch).

The second problem we ran into was more difficult and multifaceted. The companion disposable to the previously mentioned product first failed a carton (box) compression test (ASTM D642) during the shipping simulation, even after I lowered it to schedule III test. This alone is not that bad, we are using a stock carton and will design a custom box in the first year, the custom box we will obviously design to pass the standard. For now we are sticking in a reinforcing box liner that will hopefully give us good strength. Like I said, by itself this is not a huge problem, the FDA is not concerned about the outer box- it is more of a marketing issue, only the product packaging must remain intact. The rest of the shipping simulation we passed without incident.

However, the real problem came up when we got the sample back from the shipping simulation, remember where I said that the important thing is that the product packaging remain intact? Well on examination I discovered small wrinkles in the film of our bag seal on half of the pouches. When tested with the dye penetration test (ASTM F1929), which consists of timing the dye's travel through the seal, sure enough some of these samples failed at the wrinkle. Now I am 95% certain that these wrinkles were caused during our initial sealing of the bag, and not during the shipping simulation. I'm thinking that the film wasn't stretched completely tight during the heat sealing process. None of the vendor seals have the wrinkling.

The problem, the seals did not undergo inspection before the shipping simulation. Our company had a QA who would never let anything go out without his final inspection and approval. Unfortunately this QA no longer works for the company. Now, we have a QC in charge of inspections who sometimes doesn't inspect test samples. Even worse was that I actually sent three of the samples back for rework before I sent them because I saw at a glance that they had other unrelated problems, so I knew the samples had at least some problems before I shipped them. I should have inspected them better and insisted that they undergo a full QC check. I've definitely learned something from this.

Now I'm stuck with a couple not really great choices, redo the whole test ($1000+) with newly completely checked samples, or go ahead without the complete preliminary test results and hope everything is good enough to pass. I had thought to get manufacturing to seal a couple pouches and see if the wrinkle showed up in the new ones he sealed, but the manager has already told the guy about the problem, so hopefully its unlikely to happen again. No other seal on the pouch failed the dye penetration test. I'll probably retest, but its not making me happy tonight!