Balancing costs and quality can be challenging for any manufacturer. But for those involved in health care, it’s especially difficult because tools used in medical procedures should never be compromised by efforts to cut expenses.
But cutting costs and ensuring maximum quality are not mutually exclusive. For example, Boston Scientific—one of the world’s leading developers of medical devices and technologies—recently saved time and resources and ensured product reliability with an innovative approach to process validation.
When an outside company transferred production of a diagnostic catheter to their facility in Heredia, Costa Rica, Boston Scientific needed to prove that catheters produced at the new location would be just as good. But the project team handling the transfer had just four weeks to complete test method requirements needed to secure approval from regulators—a deadline that didn’t allow for lengthy test method validations.
So with the help of Minitab Statistical Software, the team used data analysis to verify that the new catheters’ pull wire strength continued to meet specifications using the process settings already in place. The team’s results permitted Boston Scientific to avoid performing a costly process validation, and enabled the company to proceed with its scheduled manufacturing plans.
The team at Boston Scientific’s Costa Rica facility, above, used Minitab Statistical Software to prove that adapting the catheter tension testing process would not impact the strength results of the pull wire.
The diagnostic catheter was designed for the coronary sinus, a collection of veins in the heart. Doctors use it during early intervention procedures for heart-attack patients. The handle of this tube-like device controls the pull wire, which manipulates a flat wire that activates the electrode tip responsible for the tool’s curvature. This arch controls the catheter’s stability during a procedure, making the “pull wire” critical to its success.
To guarantee the pull wire’s performance, the solder securing it to the handle is examined for strength by applying tension during a tensile test. Boston Scientific needed to show that catheters made at the Heredia facility could match the performance of the original manufacturer’s catheters in the tensile test. While both companies’ settings produced catheters that met the 5-pound specification for pull-wire strength, there were still inconsistencies in each company’s tensile test settings.
A universal testing machine measures tensile strength by applying increasing amounts of pressure to a catheter held between two grips. The distance between the two grips, or gauge length, differed between the two companies’ test methods, since Boston Scientific’s gage length aligned with ASTM’s (American Society for Testing and Materials) standard for tensile strength. In addition, the original catheter was designed and approved in an era when test method validation of specific test method standards wasn’t required. The team needed to verify that the difference in gauge length did not compromise the effectiveness of the tests, and to demonstrate that both the new and original catheters were safe and reliable. A time-consuming and costly test method validation process might be required.
To avoid incurring unnecessary cost, the project team set out to compare the results of the manufacturer’s test settings with those produced by Boston Scientific’s more rigorous standards for tension testing, and to demonstrate conclusively that adapting the catheter tension testing process would not impact the strength of the pull wire. If this could be shown to the regulators’ satisfaction, it would reduce both the amount of time and the number of catheters needed to complete test method validation by leveraging its existing ASTM tension testing standard.
How Minitab Helped
The project team decided to compare the mean pull-wire strength of each company’s test settings using a 2-sample t-test. To ensure reliable test results, they used the Power and Sample Size tools in Minitab and determined they would need to test 23 catheters with the tension test settings from each facility.
The team gathered strength test data from a random sample of 23 catheters tested with Boston Scientific’s more rigorous ASTM standards for tension testing and 23 catheters tested with the tension test settings used by the original manufacturer.
Now the team needed to verify the data met two assumptions in order to produce reliable results. First, according to general guidelines, each sample should fit a normal distribution—a criterion that was successfully proven using Boston Scientific’s statistical procedures. Second, because the variance of a sample influences its mean, the samples needed to have statistically equivalent variance.
Minitab’s Statistics menu provides easy-to-use tools to evaluate data for both of these conditions. The team used the 2 Variances test to quickly determine that their data met the requirements for a reliable 2-sample t-test.
The chart above displays the variance of each data set and illustrates that the difference between them is not statistically significant.
The individual value plot above shows that although Boston Scientific’s production settings have a wider range of results, the two means for tensile strength are nearly identical.
Minitab’s 2-sample t-test showed that the tensile strengths were not significantly different. While performing the 2-sample t-test, the project team found that Minitab’s Assistant feature—which provides on-demand explanations and recommendations in addition to step-by-step guidance—gave them more confidence in their analysis and results. “From the background provided in the Assistant, I learned that meeting the assumption of normality for a 2-sample t-test is not critical when you have at least 15 units for testing,” says German Vindas, manufacturing engineer at Boston Scientific. “It confirmed that our 23-unit sample size was more than adequate, which gave us great confidence in our results.”
The team’s 2-sample t-test demonstrated that both companies’ test settings for gauge length produced equal tensile strengths, and the data they collected showed that the pull wires in both sets of catheters met the 5-pound minimum. Based on the results, regulators agreed that Boston Scientific could adapt the catheter test method validation by leveraging its ASTM standard with absolute confidence that the new pull-wire tension strength would continue to meet the standards set by the original manufacturer.
Vindas estimates that using the statistical tools in Minitab saved the team about a month on this phase of transferring catheter production. The reduced time spent on this phase slashed testing costs while allowing Boston Scientific to meet project deadlines and move on to the next phase of Costa Rica’s regulatory approvals.
“With Minitab, the process was very straight forward,” says Vindas. “It saved us lots of time and money, and the project was able to continue on schedule.”