USP system suitability criteria

[gtma]

As per USP, "data from six replicate injections are used if the relative standard deviation requirement is more than 2.0%".
If I use 2 standards and inject them 3 times (total of 6 injections) upfront, calculate the response factor (e.g. area/wt) and then calc the overall RSD for the 6 std injections, doesn't this satify the USP system suitability criterion? Do I have to inject 6 injections of standard 1 & 2 (total of 12 injections) to meet USP criterion? I believe 2 standards injected 3 times is equivalent to the USP approach and therefore, I recommend USP to change the description in chapter <621>....."replicate injections of a standard preparation.....". Often, the system suitability standards are used for the sample calculation so the 2 standards injected 3 times approach is frequently used instead of a single standard injected 6 times approach.

Your response is appreciated.

[Gerhard Kratz]

There are many pros and cons on statistic in HPLC results, but, to be honest, argueing with USP on that is waste of time. Just do your 6 injections of one standard and be happy, if you want to get your validation done in time.

[Blazer]

Often, the system suitability standards are used for the sample calculation so the 2 standards injected 3 times approach is frequently used instead of a single standard injected 6 times approach.

I can honestly say that I have never seen the "2 standards/3 injections each" approach in any method I've seen at any company. I'm curious, is there some advantage to this approach that I'm not seeing?

[gtma]

The only advantage is you can see early into your run if your standards are likely to pass the "agreement between the standards" criterion........unless you run the other stardard as a "prime" prior to the 6 injections of a single standard.

[krickos]

The only advantage is you can see early into your run if your standards are likely to pass the "agreement between the standards" criterion........unless you run the other stardard as a "prime" prior to the 6 injections of a single standard.

..and thats the difference between system suitability and "concordance, agreement between the standards, etcetera".

What I mean regardless if you like what USP/EP has set up, so their SSTs focus on the system. As soon as you toss in an "agreement between the standards" criterion, the chromatographic system is only one part, accuracy of analyst work and the procedure with regard to how to prepare your standards also comes into play. Consequently you do not check the same thing.

This discussion is similar to recent linearity thread ie difference between detector linearity and calibration.

[DSP007]

Hi
"Two standarts" for "RSD six inject" have nothing to do.
"6 shots", need for check the correct operation of the device, primarily dosing AutoSampler. However, It should be remembered, that at very low concentrations of impurities its RSD can be much worse than 2% (summed up his own instrument mistakes -noise and emissions), but it's not critical in situation if "no impurities in substance".
Two standards - you check that both the standard solution you are prepared (made) properly and equally. And the first standart not gives the result of analysis is ten times less than the second. Keep in mind that they should not be compared "to the peak areas" but peaks area must be converted into a single coordinate system - for example the concentration / response.

[Peter Apps]

If you extrapoate the argument from one standard six injections, past two standards three injections each, you get to six standards one injection each, which is either a calibration curve or a repeatability of standard preparation excercise depending on whether the concentrations are the same or different.

How many angels can dance on the head of a pin ?

Peter

[adeerre]

two standars indicate the suitability of method perparation of the samples and test solutions (reproducibility), but not the suitability of de chromatography system, that must be evaluate with a invariable solution, in other way is incorrect.

[carls]

I'm definately not a USP expert so I'm just curious as to why more replicate injections are required when the %RSD tolerance is higher, i.e. higher variability allowed? Seems backward.

[adeerre]

One question.
in the system suitability test, the %RSD is calculated with area value of the 5 replicates of standar?or with area/weigh ratio?who is more correct?

[Consumer Products Guy]

One question.
in the system suitability test, the %RSD is calculated with area value of the 5 replicates of standar?or with area/weigh ratio?who is more correct?

We use the area counts.

Besides, if the weight is a constant, wouldn't the RSD values be the same?

[adeerre]

ok, what you say is logical.

What about Tr (retention times)? be necessary to establish an appropriate range of Tr, for example + / - 5% respect the Tr of the "preparation standard" to identificate the peak on the "Sample preparation" ?

Thank you in advance