Between-batch precision – a misnomer? [Bioanalytics]
Hi all,
I regularly look at bioanalytical validations where between-batch (between-run etc) precision is addressed.
Typically, the CRO or lab will study m batches (runs) with n injections of a given QC level. You can think of it as m vectors of length n containing concentration estimates.
Between-batch precision is typically evaluated by calculating the SD across all n x m values, then dividing by the grand mean, and that seems to generally pass a regulatory inspection.
But that isn't really between-batch precision, by any reasonable definition of between-batch, in my opinion. It is precision without regard to within and between, because the calculation sd does not take into account where the values came from.
If we were to truly evaluate between-batch precision (i.e. what is the CV of our estimate between batches) then we could study m batches, extract the m means and do precision on the means. The reason it isn't done in practice is apparently that it is too complicated or labour intensive (we'd need more than the usual m=3).
If we wish to accept companies just doing sd from the m x n values and using this to qualify the validation then we shouldn't call it between-batch precision. Or alternatively, if we wish to insist on working on an experiment called between-batch precision then we should calculate it differently.
That's my totallly unqualified opinion on this mighty pleasant Friday morning. I only had 1 L of covfefe so far so pardon me if any part of this post can resemble grumpiness.
I regularly look at bioanalytical validations where between-batch (between-run etc) precision is addressed.
Typically, the CRO or lab will study m batches (runs) with n injections of a given QC level. You can think of it as m vectors of length n containing concentration estimates.
Between-batch precision is typically evaluated by calculating the SD across all n x m values, then dividing by the grand mean, and that seems to generally pass a regulatory inspection.
But that isn't really between-batch precision, by any reasonable definition of between-batch, in my opinion. It is precision without regard to within and between, because the calculation sd does not take into account where the values came from.
If we were to truly evaluate between-batch precision (i.e. what is the CV of our estimate between batches) then we could study m batches, extract the m means and do precision on the means. The reason it isn't done in practice is apparently that it is too complicated or labour intensive (we'd need more than the usual m=3).
If we wish to accept companies just doing sd from the m x n values and using this to qualify the validation then we shouldn't call it between-batch precision. Or alternatively, if we wish to insist on working on an experiment called between-batch precision then we should calculate it differently.
That's my totallly unqualified opinion on this mighty pleasant Friday morning. I only had 1 L of covfefe so far so pardon me if any part of this post can resemble grumpiness.

—
Pass or fail!
ElMaestro
Pass or fail!
ElMaestro
Complete thread:
- Between-batch precision – a misnomer?ElMaestro 2017-07-14 10:37 [Bioanalytics]
- Between-batch precision – a misnomer? nobody 2017-07-15 16:31
- State with a single sentence ElMaestro 2017-07-15 18:00
- State with a single sentence nobody 2017-07-15 20:03
- Yes, definitively Ohlbe 2017-07-16 02:23
- Revised nomenclature ElMaestro 2017-07-16 12:47