Evaluation of Analytical Methods Used for Regulation

Open Access
Publisher Website
Google Scholar
Abstract
Regulatory analysis requires methods for analysis for 3purposes: surveillance, to detect problems; monitoring, to follow trends; and compliance, to enforce specifications (standards, action levels, tolerances). Each purpose requires a different weight to be assigned to the relative importance of the scientific characteristics of methods of analysis: systematic error, repeatability, reproducibility, limit of determination, and specificity. Evaluation consists of balancing the level of scientific requirements against the practical considerations of cost, time, and level of (raining required. In trace analysis, FDA utilizes recovery limits of 80-110% at ≥0.1 ppm and 60-100% below. The criterion for acceptable reproducibility (total of between- and within-laboratories) of methods of analysis can be represented by a relationship between coefficient of variation and concentration. The results of collaborative studies show that 10% outliers must be tolerated, but above 20% is excessive. Other new, important criteria for evaluation of methods at trace levels are % false positives, false negatives, and the magnitude of the blank.

This publication has 0 references indexed in Scilit: