Table 6. Position of curative maintenance.

Recommended practice	The laboratory must verify by a method of its choice the absence of drift before failure: re-testing of patient samples, use of mean values of results, or any other means that must be justified. In case of re-testing of samples, the number depends on the size of the series (N) (square root of N, 10%, “rolling back the series” to determine when the system malfunction had an impact on the results, etc.). The impact study is recorded and corrective measures are implemented.
Acceptable practice	Review of all patient samples (subject to measurand stability) since the last valid IQC (or other relevant practice), but no reflection or strategy.
Unacceptable practice	No impact study and therefore potential release of erroneous results.
IQC – internal quality control.