Importancia de la selección de una especificación de calidad adecuada en la aplicación del modelo Seis-Sigma en el laboratorio clínico

Abstract

IntroductionThe Six-Sigma model is a quality management tool based on the variability of a process in terms of defects per million (DPM). In the analytical phase, sigma value estimation is dependent of the allowable total error (TEa). Our goal is to demonstrate that the improper selection of quality specifications can lead to erroneous interpretations of the laboratory status. Material and methodsThe Sigma value was estimated using the equation (TEa - bias)/coefficient of variation. We used different quality specifications based on biological variability (BV), the consensus of the SEQC/AEFA/AEBM, and the American and German legislative specifications. The laboratory imprecision and bias were estimated from the internal/external control quality. ResultsSigma values are inversely proportional to the requirement for quality specification. Specifications based on German and American laws or the consensus of the Spanish experts were proportionally correlated with higher Sigma values. The optimal and desirable BV was associated with high rates of determinations that did not exceed the minimum Sigma requirement. ConclusionsThe selection of less demanding quality specifications involves elevations in Sigma values, which is not necessarily associated with an improvement in the detection of clinically significant errors.In order to avoid false elevations in performance reports and errors in the inter-laboratory comparison, we recommend indicating the sigma value and the TEa.