Comparison of the sigma metrics using the total error allowable algorithm with variation of bias source

Abstract

Sigma Metrics, as a quality indicator, have been widely applied in clinical laboratories to assess the performance of analytical methods. Described in the document Clinical and Laboratory Standards Institute (CLSI) EP15- A3, the use of target values can be sourced from certified reference standards, survey materials from the Proficiency Testing (PT)/External Quality Assessment (EQA), materials used in inter-laboratory quality control programs and internal quality control materials with predetermined targets. This research aims to determine whether there is a difference in the sigma metrics between the bias derived from the manufacturer's target value and those from the peer group source in the External Quality Assurance Services (EQAS) program. The research methodology employed is descriptive comparative analysis, utilizing the results of material inspection data for 15 internal quality control parameters of Clinical Chemistry over a span of 2 years at the Pramita Laboratory in Bandung. The calculation of the sigma metrics commences with computing the coefficient of variation (CV), and the appropriate Total Error aalowable (Tea) sources for each parameter are determined beforehand using the TEa algorithm. The research findings indicate a difference between the sigma metrics derived from the manufacturer's target value and those from the EQAS-peer group target value, accounting for 33% or 10 parameters out of the total parameters with 2 levels of inspection are calculated on the sigma scale. However, in 67% or 20 parameters out of the total parameters, no such difference is observed. Bias associated with the target value from the manufacturer and the EQAS peer group shows no significant difference, suggesting that the laboratory can utilize pre-existing target values confidently.

Keywords

• Sigma metrics
• TEa algorithm
• Variation of bias sources