ISO 19227 Validation Cost

[Chrisx]

I recently reviewed ISO 19227 and I'm a bit surprised at the extent of testing. It would appear to require testing with statistically valid sample sizes for: visual, LAL, TOC, THC, Inorganic contaminants, and cytotoxicity. I believe we could waive bioburden and particulate, since we don't perform the final sterile packaging of the device. When I consider 2-runs for an OQ and 3 for a PQ, times 6-tests, times a valid sample size, then the numbers rapidly grow. Does anyone have any experience with the cost to validate to this standard? It is probably the right thing to do, but the cost would appear to be outside what most manufacturers are likely to absorb. I suspect compliance to this standard is very low. Do any of you have experience justifying a very low sample size? In my experience, many manufacturers use a single unit for many biocompatibility tests. I'm not suggesting that's right, but I'm not aware of the ISO 10993 series providing any guidance on sample size.

 

[planB]

Chrisx,

ISO 19227:2018 is not prescriptive, but extensively points out in section 4.2. that the design, validation and monitoring of the cleaning process, including chosen acceptance criteria and test methods should be based on risk management activities. In this context, the standard lists the criteria/endpoints you are mentioning, but frequently starts with wording similar to "If the purpose is to demonstrate cleanliness regarding [criterion], ...", so it is up to your assessment which criteria you regard as significant to achieve your defined level of cleanliness.

Furthermore, the standard is rather vague on sampling, section 4.5 "When appropriate (for example, when establishing the reproducibility of the process), sampling plans can be based upon statistically valid rationale for number of test specimens"

In case you are talking about a terminal cleaning process, then the relevant endpoints and acceptance criteria are often linked to ISO 10993-1:2018. While this standard is indeed not very explicit about sample size, it implies that biocompatibility-related tests can be regarded as "type tests" (n=1). This is explicitly spelled out in ISO 18562-1:2017, Biocompatibility evaluation of breathing gas pathways in healthcare applications section - screenshot from the table of contents:



Number of OQ runs: "NOTE 3 In some cases the relevant limits of the process might be both the lower and upper limits, while in other cases only one of these limits might be relevant."

PQ runs: it is up to your risk assessment whether you require re-testing all your OQ criteria to demonstrate cleaning process reproducibility and consistency.

Hope this helps,

 

[Chrisx]

Thanks for the additional information. I had not previously thought of biocompatiblity as a type test, however I agree that a sample size of n=1 is common.