OUP user menu

Performance Characteristics of Six Homocysteine Assays

Sonia L. La’ulu, Mindy L. Rawlins, Christine M. Pfeiffer PhD, Mindy Zhang MD, William L. Roberts MD, PhD
DOI: http://dx.doi.org/10.1309/AJCP64BJIPNPSQDJ 969-975 First published online: 1 December 2008

Abstract

Elevated concentrations of homocysteine (Hcy) are associated with a range of disorders. Linearity, imprecision, interference, method comparison, and accuracy were evaluated on the ADVIA Centaur (Siemens Healthcare Diagnostics, Deerfield, IL), ARCHITECT i2000SR (Abbott Diagnostics, Abbott Park, IL), AxSYM (Abbott Diagnostics), and IMMULITE 2000 (Siemens Healthcare Diagnostics) methods and analyzers and the Catch (Equal Diagnostics, Exton, PA) and Diazyme (Diazyme Laboratories, San Diego, CA) methods, both on the Modular P analyzer (Roche Diagnostics, Indianapolis, IN). All methods were linear with maximum deviations from target recoveries of less than 10%. Total coefficients of variation ranged from 1.7% to 9.4%. The effects of hemolysis, icterus, and lipemia were assessed. Method comparisons were performed using high-performance liquid chromatography as the comparison method. Correlation coefficients were 0.95 to 0.99. Bland-Altman plots demonstrated percentage bias of –29.3% (IMMULITE) to 7.2% (Centaur). Accuracy using the National Institute of Standards and Technology Standard Reference Material 1955 showed varying results with only 1 method within the certified range for all 3 levels. All methods demonstrated acceptable performance except the IMMULITE, which is less precise and accurate. Standardization of most methods seems acceptable, although continuing efforts are warranted.

Key Words:
  • Homocysteine
  • Method comparison
  • Imprecision