ISO13485 Incoming Receiving Parts Verification

C

cnrestina

Hello all,

I need help understanding what inspection results need to be retained as records. A company uses an AQL for inspecting incoming bar stock. They will "inspect" the AQL level but only record the high and low on the receiving inspection report and throw the other results out. Is this a violation of the standard? Is this a violation of 21CFR part820? I am getting conflicting information and would like to run it by the knowledgable people of the forum. Thank you for your time and consideration!
 

somashekar

Leader
Admin
Is this a violation of the standard?
NO
Is this a violation of 21CFR part820?
NO
If that is the way you have set your inspection record to be, and it has been giving you necessary effectiveness, keep maintaining it.
So you are not using the other results for any statistical analysis. You may not need it from the PASS / FAIL identification viewpoint.
May I ask what are the parameters the "inspect" contains, and what you mean by high and low.
 
Last edited:

Marcelo

Inactive Registered Visitor
In cases like this, my suggestion always is:

- Perform an analysis of the information: do you need the information to show compliance to anything? Do you need it for traceability purposes?

- If the answer is that you need, you will need to keep the results.

_ If the the answer is no, don?t keep the results, but record the analysis and decision, because probably someone will ask.
 
C

cnrestina

NO

NO
If that is the way you have set your inspection record to be, and it has been giving you necessary effectiveness, keep maintaining it.
So you are not using the other results for any statistical analysis. You may not need if from the PASS / FAIL identification viewpoint.
May I ask what are the parameters the "inspect" contains, and what you mean by high and low.


Thank you for reply! No results are used in a statistical analysis.
The parameters are dimensions of the bar stock width/thickness/height with +/- according to drawing. They use the highest result and the lowest result and record those results onto the inspection report. My debate is using an AQL, taking the samples inspected, and only recording the highest and lowest results on report. Why use the AQL if you are only recording 2 results? How do we know that the inspector actually looked at 13 pieces from a 50 piece lot (AQL=1.0)?
 
C

cnrestina

In cases like this, my suggestion always is:

- Perform an analysis of the information: do you need the information to show compliance to anything? Do you need it for traceability purposes?

- If the answer is that you need, you will need to keep the results.

_ If the the answer is no, don?t keep the results, but record the analysis and decision, because probably someone will ask.

Thank you for the reply. The information gained and written on the incoming inspection report ensures purchased product meets specified purchase requirements.
 

Ronen E

Problem Solver
Moderator
Hi,

Pretty strange to me. AQL decision making is based on numbers of "pass'es" and "fail's". If only min and max are recorded, there's no way of challenging or verifying the inspection retrospectively. This basically means this process is hardly auditable and hence lacks a feedback loop, which makes it hard to improve or at least maintain at an acceptable level.

It all comes down to how important this process is in the overall picture.

Cheers,
Ronen.
 

somashekar

Leader
Admin
Thank you for reply! No results are used in a statistical analysis.
The parameters are dimensions of the bar stock width/thickness/height with +/- according to drawing. They use the highest result and the lowest result and record those results onto the inspection report. My debate is using an AQL, taking the samples inspected, and only recording the highest and lowest results on report. Why use the AQL if you are only recording 2 results? How do we know that the inspector actually looked at 13 pieces from a 50 piece lot (AQL=1.0)?
Your AQL inspection record will mention that the random samples of 13 # were verified from a 50 # lot.
The record will also have the lot decision status. PASS or FAIL
Now it is entirely your requirement to record the individual 13 sample values, or just the High and Low or nothing at all.
As long as the verification activity ensures that the purchased product meets specified requirement, you are good.
What comes into the record is your risk based decision.
 
U

uhohraggy

I'm surprised at most of these responses. If your AQL said you should measure 13 and you only recorded the measurement of the high and low, you have no evidence that you measured the other 11 or that they were in fact in between the high and low recorded. All results should be documented as the golden rule applies (If it isn't documented it didn't happen)
 

Marcelo

Inactive Registered Visitor
golden rule applies (If it isn't documented it didn't happen)

The "golden rule" works well if you don?t know what you are doing or if you want to be on the very high end of the safer side. Unfortunately, this creates bloated systems and a lot of paperwork that serves nothing.

From a quality management standpoint, everything should first be analyzed for need and then defined as really required.

The trend on a risk-based approach is really this, first analyze stuff, then do whatever you need after the analysis. Hopefully this will help create streamlined systems.
 
Last edited:
U

uhohraggy

You are correct since this is an ISO board. I deal with US marketed products only. I can tell you the golden rule applies to the FDA and they would find this deficient without question.
 
Top Bottom