Magnani B, Fenton T, Lapp C, Brugnara C
Center for Experimental Therapeutics and Reperfusion Injury, Department of Anesthesia, Brigham & Women's Hospital, Harvard Medical School, Boston, Massachusetts 02115, USA.
Am J Clin Pathol. 1997 May;107(5):527-33. doi: 10.1093/ajcp/107.5.527.
The purpose of this study was to assess the interchangeability of the nephelometric (immunoassay), von Clauss, and optical (derived) fibrinogen assays. When the nephelometric assay was compared with either of the two functional assays (199 samples) and when the von Clauss and optical methods were compared (879 samples), Pearson and intraclass correlations were .96 to .97. However, there were statistically significant biases (P < .001); the mean ratios of nephelometric to functional assay values were 1.05 to 1.07, and the mean optical to von Clauss assay ratio was 1.05 (data analyzed as logs and expressed as antilogs). The 95% limits of agreement demonstrated that 5% of the ratios for the von Clauss and optical methods were outside the 0.83 to 1.32 range. Analyses restricted to cases with a fibrinogen value of less than 2.0 g/L resulted in somewhat greater biases and higher upper limits of agreement. Laboratorians need more than correlations to make informed judgments regarding the interchangeability of these methods.