Can Scope of Equipment Control be Tied to Risk?

[Mark Meer]

21 CFR 820.72 (Inspection, Measuring, and Test Equipment) says:

"Each manufacturer shall ensure that all inspection, measuring, and test equipment, including mechanical, automated, or electronic inspection and test equipment, is suitable for its intended purposes and is capable of producing valid results. Each manufacturer shall establish and maintain procedures to ensure that equipment is routinely calibrated, inspected, checked, and maintained."

I'm interpreting this as reading: "if a piece of equipment is used for any inspection, measurement, or test, it must be routinely calibrated, inspected, checked, and maintained."

Is my interpretation incorrect? If not, is there no way to justify not doing this on the basis of risk?

For example, we have calipers used in receiving inspections to confirm the distance between drilled screw-holes is correct prior to admitting into inventory. This check is in place for the sole purpose of avoiding discovering that the holes are mis-aligned later during assembly. It has NO EFFECT on final product. Our risk assessment is as follows:

1. Any non-conformities have 100% detectability during assembly (as you simply will not be able to sink the screw if the hole is not aligned).
2. EVEN IF it were possible, the screws are redundant - in other words, even if 1 or 2 screw holes are off, the other screws are sufficient to secure the part.
3. EVEN IF the part is not secure, this does not result in a hazardous situation (the part being secured is aesthetic in nature)

It seems reasonable that this would be acceptable justification for not requiring a calibration schedule for the calipers. Sure, they make an inspection measurement, but there is ultimately no product/consumer impact.

However, by my interpretation of the regulation, there is not allowance for such a justification.

Curious what others think...
MM.

 

[Ronen E]

Mark,

I think you read too much into this regulation.

You are required to ensure that the equipment is suitable for its intended purpose. You have done that, apparently.

You are also required to ensure that the equipmment is capable of producing valid results. The validity of results obtained is closely tied to the purpose of the test / inspection. In the situation you describe it seems that results validity is almost irrelevant from a product quality perspective. It’s more about economic efficiency (BTW, have you considered a false negative situation?).

Last, you are required to have procedures to ensure that equipment - in general - is <...>. It doesn’t say that the procedures must require each and every piece, in each and every role, to be calibrated. Rather, I would interpret it that the procedures must include guidelines / rules for deciding which piece requires what type of action. That’s where you could integrate risk analysis or some other way of rationalising.

Cheers,
Ronen.

 

[yodon]

I agree with Ronen that you're probably reading a bit much into the requirement; however, it does bring up a few interesting (to me at least) thoughts.

If I was auditing, my knee-jerk would be that since you're using the calipers to make quality decisions, why *wouldn't* you have it in the calibration program? Effectively, you're saying that the inspection step doesn't really matter. You do run the risk (figuratively) of rejecting good product / accepting bad product.

If you stick with not calibrating, I would be explicitly and overtly clear that those calipers are not calibrated ("for reference only"?) and that the inspection step is merely a sanity check and no real quality decisions are made (which may be tough to argue).

I know we don't want to do things "for" auditors but this might be something to document quite well.

 

[Mark Meer]

I agree with Ronen that you're probably reading a bit much into the requirement;...

Ya, we had a bad experience with an FDA auditor that also tended to "read a bit much into requirements", and ever since then we've been tending to interpret regulations in the most strict way we can think of, reassessing, and then tempering our strategy to what we can reasonably accommodate.

If I was auditing, my knee-jerk would be that since you're using the calipers to make quality decisions, why *wouldn't* you have it in the calibration program? Effectively, you're saying that the inspection step doesn't really matter. You do run the risk (figuratively) of rejecting good product / accepting bad product.

This raises an interesting discussion (I think) about the distinction between intent of a regulations versus those of quality system principles in general.

From a quality system perspective, I can see your point. Though I'd argue that, when all is considered - how often we receive, cost/benefit of performing, the likelihood of calipers being so far out of calibration to affect results... - the additional burden of a calibration schedule is not worth it, as we are confident of the results.

Regulatory bodies (like the FDA), on the other hand, should concern themselves only with final product, IMO, and should not burden a company with unnecessary controls, provided it can be demonstrated that the end product is not affected. Such is the case in this example.

This is not to say they are mutually exclusive. More like a Venn diagram, where the regulatory requirements are a subset of quality system principles:
Does it affect final product? - Then it is in the purview of both regulations and quality
Is it related to efficiency (but doesn't affect final product)? - Then it is in the purview of quality only

That's my viewpoint, anyway...