M
MIREGMGR
Naturally you have to record what units were tested in any testing program. How else would the test results be able to be provably associated with your development/production product stream?
What do you mean by "trap"? Do you have information that they are unlikely to find out except by using a "trap"?
What do you mean by "trap"? Do you have information that they are unlikely to find out except by using a "trap"?
J
Jeff Frost
Lot numbers are part of identification and traceability requirements for the product produced. See Section 7.5.3 of ISO 13485 as an example.
Performance test lab reports and biocompatibility test reports, are type test reports that are taken up on production sample to assure that they meet the design specification. A lot number assures that the test samples come from a specific production lot, which has all the relevant DHR information. Unlike routine tests, these are not performed on regular production lots.
Therefore the validity of these reports at any future time can be assessed based on the study and comparison of the ongoing lot number DHR. The reviewer, as well as the company can assess the validity or decide to retest based on significance of the changes made. These changes reflect into the DHR.
Michael Malis
Quite Involved in Discussions
As stated above by 3 individuals, lot numbers are used for product Material Traceability. And based on the product, some of the tests ARE performed on regular lots. It is important to remember, that biocompatibility can be affected by raw material. For example, if we have questionable results on the product (made from Polycarbonate, for example) that was perfectly acceptable before, we can evaluate raw material lot to check for the presence of contamination.
R
regfocus
Thank everybody's help. I understand ID/Lot/Batch is necessary in the testing report. My question is if FDA concern the tested samples came from single batch/lot or multiple lots.
I saw a FDA guidance "Guidance For Conducting Stability Testing To Support An Expiration Date Labeling Claim For Medical Gloves Draft"
It said,
3. Preparation and Sampling
3.1 An appropriate number of boxes of finished product (i.e., after final packaging and, if applicable, after
sterilization) should be randomly selected from each sampling lot on the day of manufacture.
Does that mean if only single lot was tested will not be accepted by FDA for 510k submission?
I saw a FDA guidance "Guidance For Conducting Stability Testing To Support An Expiration Date Labeling Claim For Medical Gloves Draft"
It said,
3. Preparation and Sampling
3.1 An appropriate number of boxes of finished product (i.e., after final packaging and, if applicable, after
sterilization) should be randomly selected from each sampling lot on the day of manufacture.
Does that mean if only single lot was tested will not be accepted by FDA for 510k submission?
M
MIREGMGR
My understanding of the guidance in question is that FDA wants the testing process to fairly represent the production variability of a day's production.
So, if your testing is representative to that extent, they probably will accept your explanation.
If on the other hand it appears to them that your testing essentially was a "snapshot" of production, and there is no evidence that it was fully representative of production variability, they might not accept it.
So, if your testing is representative to that extent, they probably will accept your explanation.
If on the other hand it appears to them that your testing essentially was a "snapshot" of production, and there is no evidence that it was fully representative of production variability, they might not accept it.
R
regfocus
1) Does that mean samples from multiple lots is better than from single lot from the FDA expectation perspective?
2) If FDA prefer samples from multiple lots then is there any guidance about how many lots are appropriate?
3) For example, I have samples from 3 lots (Lot#1, Lot#2, Lot#3), and I have 3 product properties (property#1, property#2,property#3) need to be verified via lab tests. How should I arrange the samples to have better representative of production variability? Which of the below protocol is the best?
Protocol 1:
property#1 (Lot#1, Lot#2, Lot#3)
property#2 (Lot#1, Lot#2, Lot#3)
property#3 (Lot#1, Lot#2, Lot#3)
Protocol 2:
property#1 (Lot#1, Lot#1, Lot#1)
property#2 (Lot#2, Lot#2, Lot#2)
property#3 (Lot#3, Lot#3, Lot#3)
Thanks
