Accuracy for external low QCs bad but not for high levels...

Basic questions from students; resources for projects and reports.

12 posts Page 1 of 1
Hey!

I am doing a validation of a HPLC based bioanalytical method.
My problem is: if I am using self prepared quality controll samples (stock solution is the same as for the calibration) I always obtain good results. Every time I am using external quality control samples I am also getting good results for high concentration levels (about 95%) but not for low (only 75%) ...

I have already tried to do weighted regression analysis with 1/x as a weighting factor but this makes the results even worse (only for the external low quality control samples).

Does anyone can help me? What is the reason for that? I don't have the foggiest idea :(

Thanks!
Have you tested the preparation accuracy of the stock solutions?

Unless your preparation replication is good it can badly affect you extraxted preparations.

If using the same stock for STDs and QCs you get good results which then deteriorate when you use a second stock for QC prep I would suspect poor stock prep replication
Good judgment comes from bad experience, and a lot of that comes from bad judgment.
I assume the QC mix is matrix spiked with your stock solution? Some times coeluting compunds can make peak shapes bad and quantitation poor. As you talk aboud low recovery, I assume there is a preperative step or two along the way. Could there be something in the matrix that would tie up some of the analyte? Recovery issues often show up at low analyte concentration - and the presence of matrix often does not help :( .
The difference is I prepare spiked quality control samples. Stock solutions are certified and I use plasma to dilute them. The external QCs which I bought are already in plasma. Maybe this is the reason but I still don't know how to fix this because I have to use spiked calibration standards. :(
JGK wrote:
Have you tested the preparation accuracy of the stock solutions?

Unless your preparation replication is good it can badly affect you extraxted preparations.



Actually no. DO you mean testing stock solution diluted with water against stock solution diluted with plasma?

But if this is the problem I don't know how to fix it :?:
seemeto wrote:
JGK wrote:
Have you tested the preparation accuracy of the stock solutions?

Unless your preparation replication is good it can badly affect you extraxted preparations.



Actually no. DO you mean testing stock solution diluted with water against stock solution diluted with plasma?

But if this is the problem I don't know how to fix it :?:


No, I mean making a dilutiuon of each prepared stock which can be analysed on the system and comparing the unit response (area/concentration value) of the analyte in each. Many laboratories have a criteria for acceptance on replicate preparations. when I worked in Pharma assays we had a 2% limit, i.e the ratio of unti concetration values had to be in the range 0.98-1.02.
Good judgment comes from bad experience, and a lot of that comes from bad judgment.
Are the result from individual vials (from the supplier), one per result - or do you have several results from a single vial? How do you ensure samples are homogenized before you draw your samples? Is there sediment in the vial? If you take several samples from a vial, do you take them all at once or do you leave hours or days between - and how do you store the sample if you have periods between taking samples?
there are always 6 vials for each standard. I do not inject several times. There is also no sediment in the vial.

I checked for Freeze and thaw stability and my samples are only stable for only 1 cycle. The calibration standards I used had 2 cycles. Tomorrow I will check for that. Maybe this is the problem...
Multiple injections from the same vial may help tell what is going on. If you can make two or three injections from a vial and the results from the injections are close to each other - it tells you that the problem is not something occuring at or with the instrument -- at least not in the length of time it took to complete the several injections. And that points to variability in the samples.

On the other hand if you make several injections, one after the other from the same vial, and the results scatter badly, you know that there is a problem that occurs at the instrument - and that can be sample stability or a chromatographic issue. (Do the results just scatter or is there a trend across several injections from the same vial?)

Combining these if you can run a copule of injections and the results are close and then set the sampels asside for a while and run them again, if the results change, there is somethign that is not stable over the longer period of time. A sample that is known to be stable needs to be run as a control through the experiment to ensure the instrument remains stable. With a stable instrument and changes in results - the suspect becomes the samples and the manner of storage.
I injected from 2 vials with different concentration levels several times and the results are the same. Maybe the storage conditions are the reason... My calibration standards are thawed 2nd time and my quality controll samples the first time before they are further processed. This is the main difference in storage conditions for the samples.

But I have to check if the results remain the same if I use calibration standards which are only thawed once.
Sounds like you have some good information -- keep us posted.
I checked freeze and thaw stability for samples frozen once and twice against freshly prepared calibration samples. I have got the same results :( : external low qc samples have an accuracy at 70%. This time I used 3 external qc levels. Highest QC level had the best results while the lowest QC (almost at the LLOQ) had the worst results.

Maybe it is a problem of linearity?

What I did next: I used the external QC levels as calibration standards and checked my own calibration levels as QC standards. The concentration range for the external QCs is only half as of the calibration standards... The results were good (95% or higher) except for the LLOQ where there was an accuracy of 85%.

but what I do not understand is why my own prepared QC levels near the LLOQ have got an accuracy of 90 - 95% (using weighted regression with 1/x or 1/x2).


Does anyone have any ideas?
12 posts Page 1 of 1

Who is online

In total there is 1 user online :: 0 registered, 0 hidden and 1 guest (based on users active over the past 5 minutes)
Most users ever online was 1117 on Mon Jan 31, 2022 2:50 pm

Users browsing this forum: No registered users and 1 guest

Latest Blog Posts from Separation Science

Separation Science offers free learning from the experts covering methods, applications, webinars, eSeminars, videos, tutorials for users of liquid chromatography, gas chromatography, mass spectrometry, sample preparation and related analytical techniques.

Subscribe to our eNewsletter with daily, weekly or monthly updates: Food & Beverage, Environmental, (Bio)Pharmaceutical, Bioclinical, Liquid Chromatography, Gas Chromatography and Mass Spectrometry.

Liquid Chromatography

Gas Chromatography

Mass Spectrometry