Retained samples from each batch of spinal implant medical screws that we make

[Sandi A]

Hi all
We are an ISO 13485 certified manufacturer in South Africa, manufacturing Spinal Implants. We have just had our surveillance audit and received a Minor NC for not retaining samples from each batch of screws that we make. Up until know we have retained the bar end of each batch of material used, but not an actual implant. We also retain Production control cards recording all relevant information, material batch, machine used, measuring equipment, operator quantity made, issued etc as well as Inspection records. Is it necessary to retain samples of manufactured product, and what would the reason be behind it? I know that the standards change and become more demanding but we have been certified for 17 years and have not done this before. The auditor refers to clause 7.1, but I am struggling to see any indication that the records need to include samples. Thanks

 

[Tagin]

I don't see any explicit requirement for physical samples in 7.1. The closest would be 7.1d ("records needed to provide evidence..."), but its a stretch to interpret 'records' as meaning physical samples.

 

[Sandi A]

Thanks Tagin. How do we justify not doing it though? If it is a minor this year and we get the same auditor next year he may escalate it to a Major. Would there be any benefit to doing it? The machining process to make the screws does not alter the titanium in any way (we have had this tested) and out fatigue testing is done on completed screws. If a screw broke in situ, we would have the bar end of the material to test to see if there are any problems with the material. I cant figure out if I am missing something. I have no problem doing it if it adds value to our system, but I dont want to do unnecessary work either. Would love any inputs. Is this an industry standard perhaps?

 

[rob73]

Ask the auditor why they think clause 7.1 requires you to keep samples. Maybe they feel that your records are not enough to satisfy 7.1 (d).

 

[Jean_B]

Thanks Tagin. How do we justify not doing it though? If it is a minor this year and we get the same auditor next year he may escalate it to a Major. Would there be any benefit to doing it? The machining process to make the screws does not alter the titanium in any way (we have had this tested) and out fatigue testing is done on completed screws. If a screw broke in situ, we would have the bar end of the material to test to see if there are any problems with the material. I cant figure out if I am missing something. I have no problem doing it if it adds value to our system, but I dont want to do unnecessary work either. Would love any inputs. Is this an industry standard perhaps?

The requirement to retain samples of production batches could feasibly come from:

1. a standard generic to your industry,
2. a standard specific to your product type (in which case he should have written you up to that standard specifically, or at least in combination for not identifying and applying that standard ),
3. a regulation from a market to which you distribute,
4. a customer specific requirement (i.e. you deliver to an industry which puts this expectation on you as a supplier)
5. your own risk management file, which identifies this as a risk mitigation to aid in timely and adequate targeting of field safety corrective actions.
   1. Note: the auditor could have awareness that your risk management should contain it if it doesn't. This could come from field safety notices and actions of similar companies you should have been (made) aware of.

As stated by the others: if it is a non-conformity the auditor needs to be specific about the non-conformity, otherwise it is merely a best practice.

As for whether a sample can count as a record? Yes, by leaping through the definitions of "record" to "document" of ISO 9001:2015, which is normative to ISO 13485:2016. It's note 1 on document explicitly includes "master sample", and the composition of the material (especially after processing, watch out for adjuvants under the EU MDR for example) could be argued to be a result achieved which is pertinent to maintain. The auditor might argue that you've shown the titanium isn't altered under 'happy path/everything goes as planned' conditions, so your validation needs to cover boundary conditions and your monitoring on approaching but not exceeding those.. Readable can be argued to be analyzable.
But that in itself does not make the sample a mandatory record to keep unless one of the earlier mentioned points is determined to apply.

 

[Tagin]

Thanks Tagin. How do we justify not doing it though? ....

The onus is on the auditor to justify the nonconformance and to explain specifically where (i.e., specific sentence, not just a general clause#) in the standard such a requirement is stated.

As for whether a sample can count as a record? Yes, by leaping through the definitions of "record" to "document" of ISO 9001:2015 9000:2015, which is normative to ISO 13485:2016. It's note 1 on document explicitly includes "master sample"

Yes, but it is not required that the record must be a master sample; rather, the note 1 for the definition of 'document' provides a number of possible forms for a document (or record) - master sample is just one option:

Note 1 to entry: The medium can be paper, magnetic, electronic or optical computer disc, photograph or master sample, or combination thereof.

 

[Sidney Vianna]

Is this an industry standard perhaps?

As previously mentioned, the onus is on the auditor to clearly identify the source of the requirement s/he claims you are not in compliance with. Any audit must have well established scope and criteria. If the auditor unilaterally invokes criteria on to the audit, without your awareness and consent s/he is not following proper audit protocols.

If the auditor finds evidence that you are not following a voluntary industry best practice, a finding, properly categorized as an observation, opportunity for improvement, comment or the like could be offered on the audit report. What they can't do is to report a finding categorized as a nonconformity for a non-existent requirement.

Further, the business impact of this expectation is massive. This material is very expensive and collecting and maintaining such samples would be a significant cost burden on to the organization.

Fight back, by appealing this NC and good luck. It is NEVER a good strategy to acquiesce and roll over, for misguided auditors. You have to maintain and improve your system, on a daily basis. They show up once a year. The system is yours, defend and protect it from tortious interference.

 

[yodon]

I certainly agree with everyone above that, as written, it seems a bit sketchy.

The post @Jean_B made got me thinking. I don't know if the OP is distributing in the EU but I got to looking at the MDR and there is wording that might lead one to needing to retain samples. In Article 93, they talk about "...provide the necessary samples of devices or access to devices free of charge..." For implants, if you distribute 100% and, in theory, all get implanted, you wouldn't exactly be able to provide access to samples.

Again, no argument with any previous posts and not even asserting the above is applicable (or if I'm understanding correctly). I always think that even if they're blowing smoke, there may be some fire.