A persistently important issue mentioned when ABC staff discusses advocacy with the members has been the quality control (QC) burden associated with labeling and distributing leukoreduced (LR) apheresis platelets. FDA has determined that we must demonstrate statistically, with validation and QC testing, that our processes attain and maintain 95 percent confidence that 95 percent of our units contain <5 x 106 residual white blood cells (WBCs). This level of performance is clinically appropriate in my clinician’s brain (spare me the epithets please). If we believe that LR is clinically important, we should be stringent about what we distribute. The problem comes when the QC burden is greater than the QC benefit. Many of us think that with our robust plateletpheresis platforms and processes, we are well past that threshold.
The first question is not how good our products are or should beper se, but how do we know how good they are, and how many ways there are to know. FDA accepts very standard statistical estimates using binomial and hypergeometric distributions. ABC members insist the volume of testing required is unreasonable, and for many of our smaller members this approach requires testing all or most of the products. The agency responds, “make an alternative proposal.”
We are trying to do so, but the effort requires a lot of work by ABC members, and it is not clear we are willing to do the lifting needed. In MCN 15-067, we described the concept of “parametric release” wherein if process parameters are within specification, QC testing is not needed. Parametric release has worked in venues like device sterilization, where documentation that appropriate sterilizer conditions obviate the need for spore strips to prove sterility with biological indicators. This approach requires that we show the agency data proving that when our process is correct, we have the requisite statistical certainty of compliance and need not perform QC, in this case counting residual WBCs. Depending on the method used, volume of testing, and your contract, each residual WBC count can cost as much as $30 – which under parametric release, is required only when there is a process issue suggesting the possibility of failure (and during periodic revalidations).
The MCN asks ABC members to answer a survey and asks volunteers to compile a significant amount of data for their calendar 2014 plateletpheresis collections. The response to date is underwhelming. Go back, look at the MCN, and get in touch with me. If it’s really important, we need to do the work. Commit to filling in the survey by Oct. 8 and the data collection spreadsheet before Dec. 1, and ABC will commit to help move the issue forward within the FDA, armed with the facts and prepared to advocate on your behalf for a more cost effective and robust QC program.
Louis Katz, MD; Chief Medical Officer; email@example.com