Using the confidence interval of the mean to detect systematic errors in one analytical run

Background. The commonly used Westgard control rules to detect systematic errors in one analytical run are far from perfect. Mean rules are more effective, but are in little use. We propose using a variant of mean rules, where we compare the confidence interval of the mean directly with the allowable systematic error. Methods. Probabilities of false rejection and error detection were calculated using standard Normal distribution functions. Results and conclusion. Confidence interval of the mean rules are more flexible and are more easy to practise than the traditional mean rules. They outperform the commonly used Westgard rules if more than one control specimen is analysed in each run.