A Novel, Nondestructive, Dried Blood Spot-Based Hematocrit Prediction Method Using Noncontact Diffuse Reflectance Spectroscopy.

Abstract

Dried blood spot (DBS) sampling is recognized as a valuable alternative sampling strategy both in research and in clinical routine. Although many advantages are associated with DBS sampling, its more widespread use is hampered by several issues, of which the hematocrit effect on DBS-based quantitation remains undoubtedly the most widely discussed one. Previously, we developed a method to derive the approximate hematocrit from a nonvolumetrically applied DBS based on its potassium content. Although this method yielded good results and was straightforward to perform, it was also destructive and required sample preparation. Therefore, we now developed a nondestructive method which allows to predict the hematocrit of a DBS based on its hemoglobin content, measured via noncontact diffuse reflectance spectroscopy. The developed method was thoroughly validated. A linear calibration curve was established after log/log transformation. The bias, intraday and interday imprecision of quality controls at three hematocrit levels and at the lower and upper limit of quantitation (0.20 and 0.67, respectively) were less than 11%. In addition, the influence of storage and the volume spotted was evaluated, as well as DBS homogeneity. Application of the method to venous DBSs prepared from whole blood patient samples (n = 233) revealed a good correlation between the actual and the predicted hematocrit. Limits of agreement obtained after Bland and Altman analysis were -0.076 and +0.018. Incurred sample reanalysis demonstrated good method reproducibility. In conclusion, mere scanning of a DBS suffices to derive its approximate hematocrit, one of the most important variables in DBS analysis.