Sigma metrics application for validated and non-validated detecting systems performance assessment.

ABSTRACT

BACKGROUND: Sigma metrics provide an objective and quantitative methodology for analytical quality evaluation of clinical laboratory. This study investigated the testing performance of validated systems and non-validated systems based on sigma metrics, and explored the major parameters affecting the system performance. METHODS: Sigma metrics were evaluated by six biochemistry assays based on Beckman and Mindray validated and non-validated systems through crossing the reagents and analyzers. Imprecision and bias were assessed for all assays based on trueness programs organized by National Centre for Clinical Laboratory. Total error allowance obtained from the Chinese Ministry of Health Clinical Laboratory Centre Industry Standard (WS/T403-2012). RESULTS: The imprecision for all systems meets the quality specifications except TP assay (2.19%) detected by Mindray non-validated system, and the bias for four assays measured by non-validated systems cannot fulfill the criterion, including lactate dehydrogenase (LDH), total protein (TP), triglycerides (TG), and glucose (GLU). Higher biases were detected in six assays at different levels among non-validated and validated systems. Systems performed poorly or unacceptably for TP assay with sigma metrics lower than 3 except Mindray non-validated system. The sigma metrics for other assays with four systems were greater than 3 except the LDH evaluated on Mindray non-validated systems. CONCLUSION: Non-validated systems may introduce performance uncertainty compared with validated systems based on sigma metrics evaluation, and lower bias was provided by validated systems. The performance of non-validated systems should be evaluated thoroughly in the clinical laboratory before they were adopted for routine use.