Should written procedures be established for physical inventory/handling discrepancies?
Hello all, here's a question that I didn't see addressed on the forums anywhere. We do periodic physical inventories such as cycle counts (usually we do cycle counts weekly). This includes finished goods as well as components and in-process goods, cycling through different items so ideally we cycle through everything eventually--hence the term "cycle"--I don't know if this is typical, but it is how we do it. We also do year-end inventory of everything we have in inventory; we do this at our fiscal year-end.
My question is this: What are your thoughts on whether it is necessary to establish written procedures describing methods and responsibilities for performing cycle counts and other counts and how to handle (and whether to investigate) inventory discrepancies? Without written procedures, I think a regulator or auditor can reasonably question why we don’t investigate discrepancies, how we know theft/food fraud are not occurring, and whether our traceability/inventory controls are adequate. It’s troubling to me that seemingly random quantities of finished goods appear and disappear with associated uncertainty regarding whether they are acceptable in food safety, authenticity, and quality.
I don't expect inventory to be perfect, and I don't think you can ever achieve perfection. I am also unaware of regulations requiring such written procedures. However, it just seems to me that without such written procedures, there are too many questions relating to food safety and quality without some written method of whether/how to investigate that something has occurred to adversely affect food safety, authenticity, or quality. Currently, we seem to view inventory discrepancies as merely minor financial losses or gains, and this does not seem right to me, for the reasons explained.
We are regulated in the United States under 21 CFR Part 111 and 117 and we have SQF and NSF GMP certifications.
Thank you,
Matthew
I think it's worth a procedure but I'd get the people who do it to write it. Certainly large discrepancies should be investigated and CAPA done. But that requires a process and a definition of what "large" is. I don't think it needs to be complex though.
Thanks GMO -
I should also explain that the most common reason given by coworkers for finished goods being more or less than expected are that we accidentally pulled from the wrong lot--thus, the lot accidentally pulled from will have too few bottles, and the lot we intended to pull from will have too many bottles, by the same quantity. We even have a term for this discrepancy, calling it an 'over under' error, because one lot is over on inventory, and the other lot is under, by the same quantity. This has been a topic of conversation for a long time, and we have shadowed workers who pick/ship and written at least one CAPA to try to uncover the root cause for such discrepancies and have not been able to stop them from happening.
You could decide though that for some you might just record them and then do some more RCA on a trend. It's absolutely possible to get rid of the kind of error you describe given time and money. But if you know how many times you're having to adjust your inventory because of this and the risk to product trace, time people spend on it etc, it might make the benefit / effort swing your way.
Call me jaded or a perfectionist, but....... If you're indicating that you've got inventory discrepancies at annual cycle counts between incorrectly handled lots, then I dare say you've got traceability issues that absolutely open the door to all of the fraud, adulterated product, and other associated recall issues that could easily be brought up by an auditor or worse a regulatory body.
Say you do have a recall, but with this internal error that happens so much you've given it a common name in your site, you thought you recalled all of the affected bottles just to find out you swapped the recalled bottles on a wrong order to whom you didn't notify of a recall. That feels problematic.