MA QC Alarm InvestigationsOf the 20

MA QC Alarm InvestigationsOf the 202 MA QC alarms, in 131 cases a fully executed and documented MA QC work-up was available for evaluation. An overview of the MA QC alarms generated per test and their evaluation is presented in Table 3. For 34 (26.0%) of the evaluated alarms, there was a significant difference between the two analyzer systems that exceeded the total allowable bias criterion. For 18 (13.7%) of the evaluated alarms, the detected error exceeded the total allowable error criterion of the assay. Furthermore,in at least 12 (9.2%) cases, the detected error enabled an immediate technical correction without a patient-result recall procedure. In 8 of these 12 cases, the error originated from the sodium electrode which was corrected by calibration. In one case, a sodium MA QC alarm was associated with detection of a clinically relevant post-calibration error (5.23%) that was not detected by QC measurement and required correction of some recently reported patient results (Fig. 1). Furthermore, one MCHC alarm triggered an urgent recalibration notification to the supplier. Shortly thereafter, a second MCHC alarm led to rapid technical failure detectionthat resulted in discontinuation of diagnostic testing on one of the hematological analyzers (Fig. 2). Also,in 12 cases (9.2%) an MA QC alarm was linked to a single patient with (one or more) extreme results (Table 3). For the remaining 85 alarms (64.9%) no cause was identified.For 4 MA QC procedures, minor changes in the control limits were made to achieve a manageable number of alarms. Generally, this followed several false MA QC alarms without any apparent analytical, preanalytical or clinical cause.