Evaluation of glucose testing performance in clinical laboratory by single-rule and Westgard multi-rules

Abstract

Background : Blood glucose testing is one of the most common laboratory requests for diabetes. Generally, internal quality controlled (IQC) using single-rule is used to monitor its precision while the accuracy is assured by an external quality assurance scheme (EQAS); and both results are then evaluated separately. Thus, the problems of imprecision and inaccuracy are not simultaneously solved. Therefore, the Westgard multi-rule scheme was later developed and used to evaluate imprecision and inaccuracy together, making it a useful QC.

Objective : The goal of this study is to evaluate laboratory performance in blood glucose testing by using the single-rule and Westgard multi-rule scheme while aiming to optimize the QC rules for the parameters.

Methods : Analysis of the data from glucose IQC and EQAS performed on Hitachi 917 (Roche Diagnostics GmbH, Mannheim, Germany) was used from September 2006 to May 2007 for retrospective review, and from June to August 2007 for prospective evaluation.

Results : Our IQC means of Precinorm U (PNU) and Precipath U (PPU) were in designed ranges of the manufacturer; meanwhile our standard deviations (SDs) were lower than the designed SDs. Our glucose testing performance from prospective study are in acceptable limits, i.e., between 3.01-4.65 sigma, while that of the retrospective review are 0.91-8.66 sigma. Although the single-rule is easier to monitor, it is not appropriate for accuracy.

Conclusion : Our study has demonstrated that the multi-rule scheme provides useful data for improving the performance in terms of precision and accuracy for glucose testing QC making it more appropriate than the single-rule since it could be used to evaluate imprecision and inaccuracy simultaneously. The multi-rule scheme could be an optimized QC tool for errors detection and false rejection.