ISSN:
1573-904X
Keywords:
accuracy
;
method validation
;
systematic error
;
statistical diagnostic tests
;
corrigible error correction technique
Source:
Springer Online Journal Archives 1860-2000
Topics:
Chemistry and Pharmacology
Notes:
Abstract A validation procedure is presented that satisfies the FDA requirements of accuracy (including precision repeatability), sensitivity, linearity, dynamic range, and homoscedasticity, all with a single set of data. The procedure utilizes the corrigible error correction (CEC) technique comprised of three response curves—standard, Youden one-sample, and method of standard additions (MOSA) plots, from a total of 15 to 18 X, Y data pairs. For the bias component of accuracy, the systematic bias error of the method is quantitatively separated into its constant and proportional error components. The overall constant systematic error is further separated into the system (blank) and analyte-matrix (sample) components. The CEC data also provide an internal, i.e., in situ corrected assay for the sample for comparison with alternative method data. Statistical diagnostic tests are used for the final evaluation of the method acceptability, specifically in deciding whether or not the systematic error indicated requires a root source search for its removal or is simply a calibration constant of the method.
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1023/A:1015876718419
Permalink
