Clinical Chemistry and near Infrared Spectroscopy: Multicomponent Assay for Human Plasma and its Evaluation for the Determination of Blood Substrates

Abstract

A multicomponent assay for the blood substrates of total protein, glucose, total cholesterol, triglycerides and urea in human EDTA-plasma by FT-IR spectroscopy is described. Transmission near infrared spectra of plasma were recorded using a 1 mm cell. Partial least-squares was applied for multivariate calibration taking into account optimising the spectral ranges of absorbance or logarithmised single beam spectra. The standard errors of prediction for the population of 124 plasma samples of different patients calculated by cross-validation are in the range of clinical acceptance for protein, cholesterol and triglycerides, whereas glucose assay performance is slightly above the recommended level. The relative standard error of prediction for urea is 12.7% based on the mean concentration of the population studied. Within an error variance analysis, the contributions of the reference method and the spectrometric measurement to the average (RMS) prediction error have been estimated for each substrate giving evidence of the spectrometric method limitations. The potential for further improvement is discussed.