A study to examine the accuracy of potassium measurements in clinical laboratories across Canada

Objective: To assess the accuracy of potassium measurements in clinical lab oratories across Canada. Design and method: The flame atomic emission spectrophotometry reference me thod for the determination of potassium was established at the Canadian Ref erence Laboratory by using National Institute of Science and Technology sta ndard reference materials. The method was subsequently used to assign targe t values for potassium to Canadian Reference Laboratory's External Quality Assessment human-serum-based testing material. A total of 503 laboratories participated and 9,279 individual External Quality Assessment test results were included in the study. Bias was determined by using difference plots. Results: Clinically significant bias (>1.6%) was observed in 45.9% of the l aboratories. Bias ranged from 0.34 mmol/L to -0.54 mmol/L. At low concentra tions (<3.5 mmol/L) a positive bias was most frequently observed (14.7% of analytical systems). At high potassium concentrations (>5.1 mmol/L) a negat ive bias was most frequently observed (31.4% of analytical systems). Conclusion: Inaccuracy in potassium results can contribute to test redundan cy and mismanagement of patients, while prohibiting the merger of laborator y data from disparate testing sites for the purpose of trending and consoli dation within a "universal health record." Inaccurate test results and the lack of standardization among laboratories adversely impact our ability to establish common reference intervals and critical limits. This inability ha s an adverse effect on medical decisions and patient care. Copyright (C) 20 00 The Canadian Society of Clinical Chemists.