The role of functional equivalence analysis in the DuPont Tyvek Medical Packaging Transition Project
November 25, 2015
Package testing for the DuPont Tyvek Medical Packaging Transition Project (MPTP) is progressing on schedule, and all results published to date support the functional equivalence of the transitional material to the current Tyvek. While testing continues, DuPont expects that the U.S. FDA will affirm functional equivalence by the third quarter 2015.
Medical device manufacturers closely following the MPTP may wonder why DuPont is conducting functional equivalence analysis rather than statistical significance analysis, especially if the MDM is accustomed to running statistical significance testing.
In the recent webinar, “The Results Are In – But What Can I Do with Them?” members of the MPTP team defined functional equivalence testing and why they chose such analysis. Jennifer H. Van Mullekom, Ph.D., DuPont’s senior consulting statistician who has been working on the project since 2009, explained the statistical approach; Leslie Love, MDM Specialist, provided an update on MPTP timing and then moderated a discussion with Van Mullekom and Mike Scholla, Ph.D., Global Regulatory Director, to cover some of the frequently asked questions DuPont has received during the project.
Van Mullekom explained that the U.S. FDA’s official guidance was to use functional equivalence testing. “Equivalence testing was a technique that’s been used in the drug industry for the last 25 to 30 years,” she said. “One of the applications is in the area of ensuring equivalent performance for a brand name and a generic drug.”
To illustrate the difference between functional equivalence and statistical significance and what functional equivalence means for the MPTP, Van Mullekom compared the project to landing an airplane on a runway. “The airplane will serve as our Transition Tyvek protocol material in this analogy,” she said. “The centerline of the runway is like the current Tyvek material. The edges of the runway are like our functional equivalence bounds. We are trying to land our airplane on the centerline but realizing there may be some error so we need to think about getting our plane on the runway.” To see this analogy in action, please watch the on-demand Webcast at www.medicalpackagingcommunity.com.
“Let’s consider what would happen if we wrote our study design and required that the Transition Material have no significant difference from the current Tyvek,” she continued. “That would be like saying that the airplane which represents the Transition Material has to be perfectly lined up with the current Tyvek material every time. The airplane would have to land on the centerline every single time. That would create criteria that are virtually impossible to achieve in every single instance.
“Functional equivalence acknowledges the ‘noise factors’ that contribute to typical process variability,” Van Mullekom continued. “In the process of sealing a package these might be equipment maintenance cycles, film batches, ambient temperature, humidity, and even Tyvek batch to batch variability. Functional equivalence allows us to land our plane, the Transition Tyvek Material, on the runway but does not penalize us if we are right or left of center. We just have to stay on the runway.”
Functional equivalence testing employs the same statistical technique as statistical significance testing, but users just interpret the results differently, Van Mullekom said. “For statistical significance testing you are going to have a 95% confidence interval. For functional equivalence testing you have a 90% confidence interval.” Acknowledging that some MDMs may question whether “the 90% confidence interval maintains the Type I error rate at 5%,” she said that there are “quite a few statistical references” to support that claim. She did emphasize that statistical significance testing “cannot prove that two groups are the same.” It has “limitations,” she said, and “can only prove that two groups are different. Functional equivalence testing is the only technique that can prove that two groups are the same within a prespecified margin.”
After further explanation, she shared results from the MPTP study that show that “all are within the functional equivalence bounds.”
Scholla then examined how functional equivalence supports interchangeability.
“We’ve done collaborations throughout the value chain with MDMS, sterile packaging manufacturers, and regulatory organizations around the world and we’ve provided a wealth of data for you to use in your documentation,” said Scholla. “All this goes into this concept of functional equivalence and how it supports interchangeability.”
Scholla said that “interchangeability allows MDMs to use either current Tyvek or transition protocol material.” The concept is that if you have a process that is validated and you put the transition protocol material through that same process and get an acceptable answer, it doesn’t matter whether you are using control or transition protocol material, you are making a decent seal.
“There are some things that have to take place to get to that point,” Scholla continued. “You have to consider this in your change control risk management and regulatory submission processes . . . The specifications for the device package don’t restrict the use of either current Tyvek or protocol material.”
And there should be no changes to the package or sterilization process conditions, he said. “As long as you take the transition material and don’t change your process, you are fine.” And any regulatory submission should be structured to allow the use of either, he added
“If you structure it properly, you will be able to use the concept of interchangeability,” he said.
The advantage of interchangeability includes ensuring continuity of supply during the transition period, inventory doesn’t need to be segregated, and it allows for depletion of inventory that exist throughout the value chain, Scholla said.
Scholla then provided a regulatory update. He encouraged attendees to engage regulatory authorities now to understand expectations rather than wait until such entities are busy dealing with notifications next year.
Leslie Love then posed several questions DuPont has received during the project to Scholla and Van Mullekom and addressed several audience questions. Please view the on-demand Webcast to hear these questions and answers.
Love encouraged attendees to visit www.areyouready.tyvek.com to find a digital MPTP package test results selector tool that will enable MDMs to search for current MPTP testing results by selecting Tyvek style and package design. Additionally a new “My Tyvek” mobile application is available to help MDMs access DuPont data and documents from their smart phones and tablets.
About the Author
You May Also Like