Detection of drift effects before calculating the standard deviation as a measure of analytical imprecision.

The sample standard deviation is commonly used as a measure of analytical imprecision, calculated from a series of n data obtained from one sample split for n assays. Drift effects cause an overestimation and consequently a misinterpretation of the standard deviation in clinical chemistry. The r-ratio test is recommended as a simple procedure for detecting drift effects. It has been found necessary to eliminate outliers before drift effects can be recognized.