Validation of Hemolysis Index Thresholds Optimizes Detection of Clinically Significant Hemolysis

• Published in: American journal of clinical pathology Vol. 143; no. 4; pp. 579 - 583
• Main Authors: Goyal, Tanu, Schmotzer, Christine L.
• Format: Journal Article
• Language: English
• Published: Oxford, UK Oxford University Press 01.04.2015
• Subjects: Adult; Aspartate Aminotransferases - blood; Blood Chemical Analysis - instrumentation; Blood Chemical Analysis - standards; Hemoglobins - analysis; Hemolysis; Humans; L-Lactate Dehydrogenase - blood; Magnesium - blood; Potassium - blood; Reference Values; Hemolysis; Preanalytic error; Specimen integrity; Automated chemistry
• ISSN: 0002-9173; 1943-7722
• DOI: 10.1309/AJCPDUDE1HRA0YMR

Objectives: Automated hemolysis index (HI) measurement has standardized the identification and gradation of sample hemolysis. Methods: This study evaluates whether clinically significant changes in the concentration of intracellular analytes occur at manufacturer-recommended automated HI thresholds (HI ≥3, >25 mg/dL hemoglobin). Results: Adult outpatient results for serum potassium (K+), magnesium (Mg), lactate dehydrogenase (LDH), and aspartate aminotransferase (AST) were analyzed. Mean ± SD analyte concentration and distribution within the reference interval (RI) were calculated for each HI group (1–7). Potassium results with an HI of 4 or more demonstrated clinically significant differences (≥0.5 mmol/L) in mean K+ concentration and RI classification compared with nonhemolyzed samples (HI = 1). LDH and AST showed clinically significant differences (+20%) for an HI of 3 or more. For Mg, only the group with an HI of 7 demonstrated a clinically significant difference (>25%); however, the number was low. Conclusions: Mean measured potassium concentrations are not clinically significantly affected by hemolysis at the manufacturer-recommended HI threshold, while AST and LDH are. Aligning reporting of sample hemolysis with clinically significant changes provides clinically meaningful alerts regarding this common preanalytic error.