3458 E 33rd Place Yuma, Arizona 85365

928.783.2050 Phone

888.905.3434 Toll-Free

928.782.1887 Fax

202 E Ross Avenue El Centro, CA 92244 760.352.2082 Phone 760.352.9899 Fax

A division of

tHe coNcePtS PreSeNted here pertain to determining the variability of a method and setting realistic limits on that variability. Assay method validation is a required cornerstone of any testing program. Without a well-conceived validation plan, a substandard method could pass if acceptance limits are too liberal, or a good method could be rejected if limits are too strict. This article presents an outline of method validation procedures using a set of acceptance criteria and refers to s as the standard deviation estimate. Standard deviation is defined as the positive square root of the variance within the sample.

Outlined below are eight criteria for method validation. Not all eight criteria will be applicable to every method. An example would be a qualitative test would not have to have a defined range of linearity. But the majority of the criteria are applicable to all methods. The criteria are outlined below.

Accuracy measures the exactness of the method. This is reflected by the agreement between a measured value and an accepted reference sample. The method should have an s less than or equal to 2%.

Precision is measured by the repeatability of results. Method precision should not exceed 90% of s.

Specificity is the ability to accurately measure an analyte of interest in the presence of other components that may be expected to be present. Typically these might include proteins, nucleic acids, impurities, degradants, matrix and buffers.

Limit of Detection (LOD) is defined as the lowest analyte concentration that can be detected. The LOD is not necessarily the lowest amount that can be quantitated to an exact value.

Limit of Quantitation (LOQ) is defined as the lowest analyte concentration that can be quantitatively determined with suitable precision and accuracy.

Linearity is the ability to elicit test results that are directly
proportional to analyte concentration. The objective is to find
the range of linearity. A minimum of five standards should be
used to generate a standard curve. Do not use unreasonable
concentrations when determining linearity. Unduly wide ranges
lead to misleadingly high correlation coefficients and can mask
fluctuations within the actual linear range. There are several
techniques that are used to diagnose and correct linearity
failures. The main cause usually is that the range of interest
is beyond the limits of linearity. This means that the assay is
incapable of delivering a linear response within the range of
interest. There are several potential corrective solutions:
• Adjust the sample preparation dilution to target another
portion of the response curve that demonstrates
linearity.

• Adopt a multi-point calibration procedure.

• Check equipment operational status.

• Tighten in-house specifications to reduce the effects of bias.