Method Validation - Accuracy

[michaelcarolus]

Hi All

When doing accuracy in method validation should your spiked soln be different to the std you are using? e.g


Stock std soln: 100 mg to 100 ml

Std: 10 ml of stock std soln diluted to 100 ml with solvent.
100% accuracy sample: 10 ml of stock std soln + placebo diluted to 100 ml with solvent.

Surely this procedure is incorrect because you are using the stock soln as your std as well as your spiked amount. Shouldn't your std and spiked solns be different?

Any comments will be appreciated.
Thanks
Mike

[aceto_81]

Why should it be different?
My point of view is that you are trying to find that there is no influence of your sample matrix/preparation on your assay.
So if you spike with a known amount, you will know what the outcome should be and therefore you can calculate the % recovery.
If you spike with another, newly made reference, I think you are introducing and extra error, wich can be and should be avoided.

Again, this is my point of view, there are maybe others.

Ace

[danko]

Ideally the calibrator (the standard) and the unknown should be different from each other.
If the same stuff is utilized, it looks (to me) more like an investigation of the precision. You may find 100 % recovery of what you expect, but it is only relative to the utilized standard. So, if the standard value is 80 % true (without knowing it), the accuracy of the determined sample is 80 %, even though the apparent result reads 100 %.
If the goal is to investigate the effect of the sample matrix, I suppose it’s OK to use the same stuff, but in my mind it looks more like specificity, rather than accuracy.
I my opinion the most trustworthy documentation of the accuracy would be to use 2 different batches of the analyte where, at least, one of them (the standard) is investigated more thoroughly – using orthogonal method/s of analysis etc.

Best Regards

[scottythree]

I've always spiked at 80, 100, and 120% and calculated the recovery for accuracy to show the placebo has no effect when various levels of active are used.

[michaelcarolus]

Hi All

Thanks for your comments. My aim is to show that the placebo does not interfer with the active. Therefore I guess it does not matter if the spiked soln and the std are the same.
Thanks
Mike

[michaelcarolus]

Hi All

Thanks for your comments. My aim is to show that the placebo does not interfer with the active. Therefore I guess it does not matter if the spiked soln and the std are the same.
Thanks
Mike