Introduction and Justification

Explain what the analyte is and why it should be tested. Include the testing technique and the benefits of the technique compared to other techniques in determining the analyte. Describe how the method was optimized. For example, why these operating conditions, why these reagents, or why this column was chosen.

Linearity

The first step in method validation is to assure that in a non-interfering matrix (such as reagent water) a concentration proportional response can be established. For example, a linear curve of concentration versus signal, or an increasing volume of titrant with increasing concentration. If the method uses an instrument, a calibration curve is established. Linearity is the ability of the method to elicit test results that are directly, or by a well-defined mathematical transformation, proportional to analyte concentration within a given range. The validation report must state how linearity is established and how a user determines whether the system is calibrated correctly (acceptance criteria). For the study, three to five repeats of the calibration are made to demonstrate that it is repeatable. For qualitative tests there is no linearity established. In addition, for expected EPA approval of the method, include a minimum of 5 points if the calibration is linear and 6 or more points for second order fits. 

Interferences

Using knowledge of the technique, or by literature search determine or estimate expected interferences. Test the method by adding known concentrations of the suspected interference to the interference free blank matrix and to the interference free blank matrix containing the analyte at the LOQ. Measure the effect of, or absence of interference up to the expected concentration of the interference in average samples, or up to the point that the interference significantly affects results. Record the interference in the method and include potential mitigation. If the interference can be mitigated, include test with and without mitigation. 

Limit of Detection and Limit of Quantitation

Seven replicates of the lowest calibration standard or seven replicates of a blank interference free matrix (usually reagent water) can be used to estimate the lower limit of detection (LOD). If the analyte is not found in blanks, you do not need to run blanks, however you do need data showing no analyte in the blanks. Multiplication of the standard deviation of the seven replicates by 3.14 estimates the detection limit. Multiplying the detection limit by 3.18 (standard deviation of the seven replicates times 10) estimate the limit of quantitation (LOQ). Other statistical methods may be used to estimate the LOD.

Repeatability and recovery in representative matrices

Select three to six matrices similar to the matrices to be named in the applicability section of the method, for example surface water, ground water, tap water, wastewater effluent, and wastewater influent. Measure the analyte in each matrix using the same method conditions that have been established in steps 1 – 5. If analyte is present, perform replicate analyses to establish repeatability and spike at concentrations to approximate the concentrations tested in step 3. If there is no, or very little analyte in the samples, spike in triplicate at the same concentrations used in step 3. Compare recovery and repeatability. Alternatively, plot the expected concentration versus found concentration of the results from step 3 and step 6. Compare visually. The lines should be nearly identical with the approximate same slope. If one matrix has a significantly different slope and all calculations are correct, there is an interference. Either find the interference and repeat steps 4 – 6 for that matrix, remove the matrix from the method applicability, or caution users of the method of the potential interference in that matrix. his is the text area for this paragraph. To change it, simply click here and type

Collaborative Test 

Use the same matrices tested in step 6. In this test, different laboratories use the revised or new test method to analyze the matrices to determine the method’s bias and reproducibility as would occur in normal practice. Laboratories receive the completed method and are expected to follow the method as written. Since step 6 demonstrated no interferences across the range of the methods, select matrices each with one concentration, but concentrations that bracket the range. For example, Matrix 1 = 1 ppm, Matrix 2 = 2 ppm and so forth. Involve at least three, and preferably more laboratories, and analyze at least two replicates of each matrix per lab.