Bente Flatland1, Liesl C Breickner, Michael M Fry. 1. Department of Biomedical and Diagnostic Sciences, College of Veterinary Medicine, University of Tennessee, Knoxville, TN, USA.
Abstract
BACKGROUND: The Heska Dri-Chem 4000 uses dry slide technology to evaluate serum or plasma. No previous independent performance evaluation is published to the authors' knowledge. OBJECTIVES: The objectives were to (1) characterize analytical performance of a Dri-Chem 4000 by measuring precision and bias, (2) compare analytical performance of that Dri-Chem 4000 unit with a predetermined quality requirement, and (3) determine whether statistical QC of the Dri-Chem 4000 is possible using the 13s control rule. METHODS: Sixteen analytes were measured using plasma from dogs, cats, and horses. Coefficient of variation (CV), bias, and observed total error (TEobs ) were calculated. TEobs was compared with allowable total error (TEa ). Sigma metric and quality goal index were calculated where relevant. QC validation was performed. RESULTS: Bias and TEobs calculated using quality control material (QCM) data were smaller than those calculated using method comparison data. Using TEobs calculated from species-specific CV and QCM-based bias, 100% of analytes in each species met ASVCP-recommended TEa . Desired error detection and false rejection rates were achievable using the 13s control rule and ASVCP-recommended TEa values for 9/16 (56%) of analytes in dogs, 9/14 (64%) of analytes in cats, and 8/13 (62%) of analytes in horses. CONCLUSIONS: Analytical performance of the Dri-Chem 4000 is comparable to that reported by other authors for other small benchtop biochemistry analyzers. Statistical QC using a simple control rule is possible for most analytes in dogs, cats, and horses.
BACKGROUND: The Heska Dri-Chem 4000 uses dry slide technology to evaluate serum or plasma. No previous independent performance evaluation is published to the authors' knowledge. OBJECTIVES: The objectives were to (1) characterize analytical performance of a Dri-Chem 4000 by measuring precision and bias, (2) compare analytical performance of that Dri-Chem 4000 unit with a predetermined quality requirement, and (3) determine whether statistical QC of the Dri-Chem 4000 is possible using the 13s control rule. METHODS: Sixteen analytes were measured using plasma from dogs, cats, and horses. Coefficient of variation (CV), bias, and observed total error (TEobs ) were calculated. TEobs was compared with allowable total error (TEa ). Sigma metric and quality goal index were calculated where relevant. QC validation was performed. RESULTS: Bias and TEobs calculated using quality control material (QCM) data were smaller than those calculated using method comparison data. Using TEobs calculated from species-specific CV and QCM-based bias, 100% of analytes in each species met ASVCP-recommended TEa . Desired error detection and false rejection rates were achievable using the 13s control rule and ASVCP-recommended TEa values for 9/16 (56%) of analytes in dogs, 9/14 (64%) of analytes in cats, and 8/13 (62%) of analytes in horses. CONCLUSIONS: Analytical performance of the Dri-Chem 4000 is comparable to that reported by other authors for other small benchtop biochemistry analyzers. Statistical QC using a simple control rule is possible for most analytes in dogs, cats, and horses.