Sending same samples to multiple labs
Question regarding a reference or a guide that specifies the practice of sending same sample to multiple labs.
For example:
Testing a sample for Protein % (higher protein is desirable)
Seller sends the same sample to 2 different labs for Protein testing on a routine basis. Depending on the situation, and without knowing who the potential customer is going to be, both COAs are kept and only one is presented when the customer is determined.
Background: some customers (based on Quality Agreements) specifically requests COAs be tested with their approved lab. But they would want to know the Protein % before committing to the purchase. The sample for micro & heavy metals etc. is then tested based on the customer's selected lab in order to have QA release on seller's side.
What are your thoughts on this practice?
Is there something in Good Laboratory Practices, ISO 17025, or other standards and guidelines that says this is not a good practice?
If this practice is to be continued as a business decision, is this justified based on Quality Agreements?
All accredited labs use the same testing methods, approved by Health Canada - so, the results should be pretty much the same. We also sometimes send the same samples to different labs, for double-verification. Also, some of our customers has their own preferences RE: labs. Those are common practices, and I do not recall I've read something in ISO 17025 restricting these practices. Let's see what other professionals would suggest.
I misinterpreted the title. I assumed, incorrectly, that the seller was only using the results with preferable data.
The only issue I can see is misleading customers like companies who find a problem, take additional samples, and keep testing until it comes back "clean".
That is not what is happening here, but I will look for the post that referenced shady repeat testing. I think it had references that I cannot remember off the top of my head.
I wouldn't be surprised if Charles is able to find the post I'm thinking of before me. He would be a good resource for this. :shades:
Question regarding a reference or a guide that specifies the practice of sending same sample to multiple labs.
For example:
Testing a sample for Protein % (higher protein is desirable)
Seller sends the same sample to 2 different labs for Protein testing on a routine basis. Depending on the situation, and without knowing who the potential customer is going to be, both COAs are kept and only one is presented when the customer is determined.
Background: some customers (based on Quality Agreements) specifically requests COAs be tested with their approved lab. But they would want to know the Protein % before committing to the purchase. The sample for micro & heavy metals etc. is then tested based on the customer's selected lab in order to have QA release on seller's side.
What are your thoughts on this practice?
Is there something in Good Laboratory Practices, ISO 17025, or other standards and guidelines that says this is not a good practice?
If this practice is to be continued as a business decision, is this justified based on Quality Agreements?
Hi jtang,
Sorry but I don't entirely understand the OP.
What is yr objection to the buyer's requirement ? Not trust the lab ? Not trust the sample ?
TBH, since you, the seller, are presumably paying for the analysis, IMEX the choice of Lab should be negotiable if you are not happy with the buyer's suggestion.
IMEX, the seller/buyer typically agree on a minimum % protein for a lot, (usually a sample-based average). As long as the lab. result complies with minimum specification, the remuneration is independent of the specific value.
(Furthermore, IMEX buyers usually sample/test the Lot quality on reception. Failure to meet contracted minimum > claim).
IMEX no two labs will give an identical numerical result on a "same" sample. ( = sampling/analytical error)
Arguments can occur when one lab gives a result just above a minimum specification and another gives a result just below, ie the true value is likely close to the contracted requirement. This IMEX can lead to a third referee analysis.
Or are you being paid depending on the absolute protein level ?. This can involve determining whether a particular lab procedure has bias (+ or, particularly from sellers POV, - ).
@ ABFC - thks compliment but memory sadly fails in this instance. However the event(s) you refer could certainly occur hence this rather cryptic phrase which typically, somewhere, appears on most, independent, 3rd party Lab reports - "The results stated apply only to the sample as received". :smile:
This is exactly what I am getting at, and I will elaborate more.
Example:
Internal specifications 80-85%
Lab A results: 82%
Lab B results: 84%
Currently, both results are kept as valid internally (although I disagree with this practice). Customers may prefer Lab B (then Lab B report would be presented); some customers may prefer Lab A (then lab A report would be presented). Sales also wouldn’t know which customers this product can be sold to until results come back from the labs (ie. Some customers may specifically say they will only purchase products 84% or above).
What would be the best practice here?
-Stick with one lab and not send to the second lab, for consistency?
-OR, once you decide to send sample to another lab, always take the latest/newest result?
I misinterpreted the title. I assumed, incorrectly, that the seller was only using the results with preferable data.
The only issue I can see is misleading customers like companies who find a problem, take additional samples, and keep testing until it comes back "clean".
That is not what is happening here, but I will look for the post that referenced shady repeat testing. I think it had references that I cannot remember off the top of my head.
I wouldn't be surprised if Charles is able to find the post I'm thinking of before me. He would be a good resource for this. :shades:
You are spending an absolute fortune and getting nothing but headaches in return
Pick a lab, stick with it, your protein value is going to move from lot to lot, that's normal and communicate that to your customers. They should not be dictating this level of minutia in your business
If ACME customer prefers lab B, then let them pay for the resample.
Lab results will vary even when following the exact same procedure and the same sample! So to my point, pick a lab and wash your hands of this
It was my post previously, but it was in reference to Cannabis and the absolutely atrocious boss I had who lacked a conscious
Thank you and I totally agree with you. I wish it was that easy to say no. I was hoping that I could have some guidelines/reference for me to support the "saying no" and "sticking to one lab". It is difficult when alot of it is customer driven.
You are spending an absolute fortune and getting nothing but headaches in return
Pick a lab, stick with it, your protein value is going to move from lot to lot, that's normal and communicate that to your customers. They should not be dictating this level of minutia in your business
If ACME customer prefers lab B, then let them pay for the resample.
Lab results will vary even when following the exact same procedure and the same sample! So to my point, pick a lab and wash your hands of this
It was my post previously, but it was in reference to Cannabis and the absolutely atrocious boss I had who lacked a conscious
Thank you and I totally agree with you. I wish it was that easy to say no. I was hoping that I could have some guidelines/reference for me to support the "saying no" and "sticking to one lab". It is difficult when alot of it is customer driven.
Hi jtang,
I deduce that you are apparently unaware of what yr customers actual specifications are before submitting a result to them. Strange business.
Are you saying that some customers want, for example, minimum 80%, others want, for example, minimum 84 % ?
IMEX most declared laboratory results are an average. Some product types are more difficult to analyse than others.
If you simply don't know the customer's requirement, it's obvious which Certificate to select assuming you wish to maximise yr chance to do any business ?. Of course, if selected lab happens to have a significant positive bias, then eventually you may get a claim after delivery which you can then blame on the lab. Good luck with that !.
Similar comment to yr quoted case where customer has minimum 84%.LabA Certificate > No potential business, LabB > potential business.
But does B actually have a positive bias ?. I usually submit duplicate samples to Labs and often include a dummy sample of known true value.
IMEX the customer "specification" in yr post 5 would possibly be minimum 80% protein so either Certificate would suffice. No prizes for guessing which Certificate a customer would "prefer" but that is irrelevant from a contractual POV.
I suggest you -
(a) be aware of a customer's quantitative requirement.
(b) use a lab which gives reliable results.
Option (b) may generate less business than using one with a positive bias but should also generate less claims.