too high recovery

[hocka]

When I make accuracy examination, the mean of my results 101-101,5 %.
I have an ion-pair HPLC method. I solve my tabletts in water, they very easy dissolve in it., then I shake the solution (15 min 400 1/min), then I spin the solution 4000 1/min (10min), then I dillution my sample with water, this sample I measure with HPLC. My standard and sample solution have same concentration.
I don't know, what cause 1-2% shift.
Please, if you have some idea, write me!
Thank you!

[Tim]

How many preparations of your samples have you made? Are they always coming out with these values, or are they sometimes higher/lower.

Are you preparing and diluting your standard and samples in the same way. If not, you could be seeing a dilution error introduced by a pipette or flask.

Have you considered that your tablets do actually contain 101-101.5% active material? Did you make the tablets yourself and know exactly how much active is in them? Were the tablets and standards made up with exactly the same lot of active material - if not, have you accounted for the purity of the actives used in the standard and samples?

One other possibility is your calculation - are you certain all the parameters in the equations used to calculate the result are the right way round, i.e. could you have got two values the wrong way round and your actual recovery is 98.5-99%.

A few things to consider. Let us know how you get on.

[hocka]

Dear Tim!
In the accuracy examination I prepare 2-2 sample in 5 concentration point. The standards and the samples have the same dilution, I calculate the standard purity. I developed two independent HPLC method (one of them is ion-pair, the other is ion-exchange chromatography) to measure the active gradinet's quantity, both of them give same recovery, in all concentration point 100,5-102% values.

[Tim]

One other thing that comes to mind - have you injected samples that contain no active material (placebos) to confirm that you have not got a peak from the excipients in the tablets coming out at the same time as your active?

[hocka]

The placebo not give peak in the active gradient's retention time.

[Tim]

Hmm. Are you certain that the purity of the reference standard you are using is correct? Could it have degraded? Do you have another lot of reference standard that you could try?

Also are you certain of the purity of the active that was used in the samples. Can you get hold of some of that active material and make it up as you would the reference standard, then see what recovery you get for that.

Is it possible that the active material has combined with one of the excipients and that is causing the higher recovery? Do you have access to a diode array detector, or Mass Spec which would enable you to check that the peak is just your active and not combined with something else.

[hocka]

I have got certificate of standard, which prove its purity,I don't know what the active material's purity, I persume its purity is 100%. I have a dioda array detector, but tomorrow I examin that placebo could cause spektrum shift. This explain the anomaly.

[Tim]

More things keep coming to mind.

You haven't said if you made the tablets yourself (or made by whoever you work for). Are they all from the same lot? It could actually be that there is more than 100% active in the tablets - that's why limits on tests are often 95-105% label amount. Can you get hold of the batch documentation which shows how much material went into the tablets?

Have you tried to make up some spiked tablets? That is, take the placebo tablets you have, add reference standard and take them through the sample preparation. You know exactly how much active you added, so can work out if your recovery is correct. You should do this across a range of spiked amounts (e.g. 50-150% of label amount spiked into the same amount of placebo each time), to ensure you are getting a linear response from your samples.

Final thing I can think of at the moment - although you have a COA for the reference standard, has it been stored correctly, how old is it - if it has degraded at all (and if your method doesn't show the degradadates, you may not know this), its actual purity will be lower, which will cause your sample results to be higher.

[hocka]

I have measured more than 10 lot of tabletts, all of them have more than 100% of active gradinet. I always get certificate with them, which prove they have 100% content of active agent. The standard' purity is measured in every two mounth with HPLC method.

[Basil]

Hi hocka,

In order to control standard for HPLC assay you should try to do it with a non HPLC method. Could it be controled with chemical assay (NaOH or HClO4),... ...

This is the only way that you have to control purity in a period of time, what if the standard is higroscopic and now have 1 % of water (you wouldn't see by HPLC)....

[AA]

I ahve been thinking about this one. If I had to guess, this is an issue with the standard. Purchased standards are seldomly 100% pure, and once opened have a tenancy to change with time (degrade, absorb water whatever) as was mentioned in an earlier post. I would suggest (if possible) using an alternate detection method to help track some of the possibilites down.

[hocka]

The standard don't absorb water it was measured by thermogravimeter. I examined yesterday with spectrophotometer that placebo could change the active material's relative absorbance (Abs/concentration). The shift is approximately 0.7% (between average values), but I worked too high RSD, I repeat it with HPLC this work more accurate instrument.

[fatBrain]

You never know the true assay result for your tablet sample. You should test spike/recovery first before jumping on the conclusion that the result was high.

[hocka]

I tried spiked recovery test that gave me same result: 101-101,5% recovery.

[fatBrain]

So I think slightly high potency may be due to system error of the test method. In my opinion, 101-101.5% recovery is good enough.

I dont know how you prepare your sample. If your tablet is big and you use small volumetric flasks, the result could go high since tablet will take some volume. If this is the case, you may try bigger volumetric flasks if you dont want to use internal standard.

Good luck.