An inter-laboratory comparison on one lot, tabulated — what changed since — setting out what I have, and where I think it stops being reliable.
What changes if the standard account of inter-laboratory comparison is wrong? I ask because I have been treating it as settled and I noticed this week that I could not say why.
Working through the consequences rather than the evidence, since others here are better placed on the evidence.
Why two laboratories may disagree: after validating the same method, they may still report different purity on the same sample due to integration differences, column age differences, subtle differences in mobile phase pH or temperature. This is normal and not a sign that one is wrong.
A qualification I should have led with rather than closed on.
The arithmetic in post #3 is right; the assumption feeding it is the part to check.
My position on inter-laboratory comparison is current rather than settled. I have revised it once already and I expect to again, so treat it accordingly.
Validation is compound-specific and matrix-specific. A method validated for one peptide is a starting point for another and not a validated method for it.
That is the version I would defend. It is not the version I started with.
Narrowing post #7, because the general version has more than one answer.
I read the earlier replies on inter-laboratory comparison twice before writing this, because I had assumed the opposite and wanted to be sure I was disagreeing with what was said rather than what I expected.
I had written a reply contradicting post #7 and deleted it. Here is what survived.
Inter-laboratory comparison has a well-known answer and a correct answer, and the interesting work is establishing that they are the same. Nobody has done that here yet.
Confirming post #7 from a second method, which matters more than confirming it from a second person.
The most useful single question about a method: what would it fail to detect? Every method has an answer and few documents state it.
Reading back through the inter-laboratory comparison threads from last year, the same three questions come up every time and only one of them has ever been answered properly. That seems like a documentation gap rather than a knowledge gap.
Confirming post #12 from a second method, which matters more than confirming it from a second person.
Where a pharmacopoeial monograph exists, a method that follows it inherits a great deal of assurance. Almost nothing discussed here has one.
The answer changed when I changed how I was measuring, which was informative.
I had written a reply contradicting post #10 and deleted it. Here is what survived.
The limit of quantitation determines what the impurity table can honestly contain. Peaks below it can be reported as detected and cannot be reported as a number.
That is one dataset and I would not build a rule on it.
Marking my uncertainty on inter-laboratory comparison explicitly. I am confident about the direction, much less confident about the size, and not confident at all that it generalises past the case in the first post.
Stability-indicating method: one that can separate a compound from its degradation products. Critical for assay methods that claim to measure actual degradation (as opposed to purity, which is orthogonal).
Where I part company with post #14, and it is a narrow parting.
An independent laboratory's method being different from the supplier's is not a discrepancy. It becomes one only when the results differ by more than both methods' demonstrated precision.
That is the version I use. It may not be the version that is correct.
Narrowing post #16, because the general version has more than one answer.
Transfer between laboratories: a method can be transferred from one lab to another, but the receiving lab needs to demonstrate that they can achieve the same performance. This requires comparative testing and sometimes small method refinements.
I would put the burden of proof on the interesting explanation, not the dull one.
Post #19 and I disagree about the size of the effect, not about the direction.
A stability-indicating method is one demonstrated to resolve the analyte from its degradation products, usually through forced degradation. Calling a method stability-indicating without that work is a claim rather than a property.
System suitability is the ongoing evidence that a validated method is still performing. Validation is done once; suitability is done every run, and it is the one that appears on a certificate.
Noting that I have skin in this question and have tried to discount for it.
Adding a small correction to the inter-laboratory comparison summary above rather than a disagreement with it. The substance holds; one of the figures is out by a factor that matters.
Narrowing post #23, because the general version has more than one answer.
Inter-laboratory comparison is a good example of a question where the honest answer is boring and the interesting answers are unsupported. I would go with boring.
Worth separating two things that post #23 runs together.
A method that reports the same number for every lot is worth a second look. Real processes vary, and a total absence of variation is a statement about the measurement rather than the process.
The most useful single question about a method: what would it fail to detect? Every method has an answer and few documents state it.
I am reporting what happened, not recommending it.
The limit of quantitation determines what the impurity table can honestly contain. Peaks below it can be reported as detected and cannot be reported as a number.
Take it as a starting point and not as a specification.
| Topic | Participants | Replies | Views | Activity |
|---|---|---|---|---|
|
About the Method validation category
Specificity, linearity, precision, accuracy, robustness, system suitability, and why two labs disagree. This post is a community wiki: any member at trust level 3 or above can edit it, and every edit is…
|
+1 | 5 | 468 | 5mo |
|
Sample handling between despatch and injection
Sample handling between despatch and injection Writing it up because I had to work it out twice and would rather nobody else did. Trying to work out what would count as evidence on sample handling between…
|
+63 | 68 | 2.4k | 3mo |
|
LOD and LOQ: definitions people get wrong
LOD and LOQ: definitions people get wrong — setting out what I have, and where I think it stops being reliable. What changes if the standard account of LOD and LOQ is wrong? I ask because I have been treating…
|
2 | 43k | 8mo | |
|
Revisiting: Sample handling between despatch and injection
Posting this under the heading it deserves: Revisiting: Sample handling between despatch and injection Everything below is what sits behind that. Reading back through what has been written here about Sample…
|
+5 | 9 | 1.1k | 6mo |
|
Validating a method you did not develop
On the subject in the title: Validating a method you did not develop Working notes rather than a conclusion. What changes if the standard account of validating a method is wrong? I ask because I have been…
|
+10 | 14 | 15k | 3mo |
| Topic | Participants | Replies | Views | Activity |
|---|---|---|---|---|
|
Validating a method you did not develop
On the subject in the title: Validating a method you did not develop Working notes rather than a conclusion. What changes if the standard account of validating a method is wrong? I ask because I have been…
|
+10 | 14 | 15k | 3mo |
|
Shoulders, tangent skims, and defensible integration — the long version
Shoulders, tangent skims, and defensible integration — the long version Writing it up because I had to work it out twice and would rather nobody else did. Setting out an integration problem with the actual…
|
+16 | 20 | 27k | 3mo |
|
What consumer-accessible tests can and cannot determine — a second dataset
What consumer-accessible tests can and cannot determine — a second dataset I have a specific reason for asking rather than idle curiosity, and the context is below. Posting a small dataset on…
|
+117 | 131 | 45k | 10h |
|
Relative retention time as a better inter-laboratory comparator
On the subject in the title: Relative retention time as a better inter-laboratory comparator Working notes rather than a conclusion. What changes if the standard account of relative retention time is wrong? I…
|
+7 | 11 | 28k | 14mo |
|
Coming back to: Where the integrator drew the baseline, and how much it moved the number
On the subject in the title: Where the integrator drew the baseline, and how much it moved the number Working notes rather than a conclusion. Posting the method first, because I know what the first three…
|
+20 | 24 | 2.7k | 10mo |