Statistical validation of H NMR protocol vs standard biochemical assay in quality control of RBC packed units.

Background: Time dependent quantification of endogenous metabolites in biological samples (blood, urine, biological tissues extracts) in normal and pathological conditions as well as following therapeutic protocols is well established. In the clinical practice, such a dynamic flux of information allows the physician to identify and appreciate alterations associated to biochemical pathways of specific organs. In the years, many biochemical assays have been developed to detect, selectively, this vast array of molecules.

Methods: The Proton Nuclear Magnetic Resonance (H NMR) spectrum allows the identification and quantification of more than 30 RBC-associated metabolites with minimum manipulation of the sample. To validate the use of H NMR spectroscopy for quality control purposes in transfusion medicine, a series of statistical tools have been employed to analyse and compare accuracy and precision of the H NMR results with respect to the ones obtained by standard biochemical assays.

Results: Among the many metabolites that can be detected and quantified by H NMR spectroscopy we selected creatinine and lactate, since they are routinely quantified by standard biochemical assays and because they are characterized by a wide concentration dynamic range. We show that 1D H NMR spectroscopy is an accurate a precise method for metabolite quantification.

Conclusion: These results validate the use of H NMR spectroscopy in transfusion medicine as a method to evaluate the quality of RBC packed units and to develop novel and more efficient RBCs storage protocols.