Thursday 8 August 2013

Help on: Do I need Linearity, Accuracy and Precision for a Method Based on Area %?

MTS HELPDESK
Do you have any problems relating to analytical chemistry for pharmaceuticals or training? Send your questions to the MTS helpdesk using our contact form.

Question:
“I’m currently writing a validation protocol for an optical purity method which is used for the determination of a main component and 3 known impurities. The results are expressed in % relative (area/area). Since all the results will be determined in % relative, do I need to investigate linearity, accuracy and repeatability?”
“It is my opinion that you will need to consider linearity, accuracy and precision for this method, even though the method is based on an area percent measurement. My reasoning is as follows:
Although the results are expressed in % relative area, it is implied that these relative areas are representative of the amounts of the main component and the impurities in the samples that you are testing and that therefore there is a correlation between % area and %w/w.
Linearity - For both the main component and the impurities, you are assuming that the relationship between the size of the peak and the concentration of the analyte is a linear one. If you are using UV detection then you probably will have a linear relationship but it is usually necessary to demonstrate that it is linear over the range of the method. Additionally, by assigning a value of 100% to the total area, and then expressing the size of each peak as a proportion of this you are actually applying a single point calibration. Therefore the equation of the line will need to pass through the origin, i.e. the value of c in the equation, y = mx + c, will need to be insignificant.
Accuracy – If you have a reference standard for your main component then I would perform an accuracy determination since this is the only way that you can show that your method does actually produce the correct result and that the area percent approach is valid. For the impurities, you will need standards of known purity to perform accuracy and you may not have these.
Note that in ICH Q2(R1), page 10, section 4.2, it says “It should be clear how the individual or total impurities are to be determined e.g., weight/weight or area percent, in all cases with respect to the major analyte.
Precision – The result that you generate for a particular sample each time you run your method will be subject to errors and thus will be variable regardless of whether your results are expressed as %w/w or % area. These errors come from sample preparation, injection, integration, etc. An assessment of this variability is essential during validation, preferably by performing repeatability (performing the analysis 6 times on the same sample at the same time), and intermediate precision (performing the analysis on different instruments and by different analysts, etc.). The reason for this is so that you know how different the result may be when you follow the method as written.
The combination of linearity, accuracy and precision provides valuable information on the performance of an analytical method.”