Development and Validation of Hematocrit Level Measurement in Dried Blood Spots Using Near-Infrared Spectroscopy

Abstract

Background: 

Dried blood spots (DBSs) have gained recent popularity as a sampling method for therapeutic drug monitoring. For patients, DBS sampling has several advantages over venous blood sampling. However, technical issues primarily influenced by hematocrit levels, interfere with the implementation of this method in daily clinical practice. The results of concentration measurements of drugs that are influenced by hematocrit should be corrected for hematocrit levels. In this article, we developed a fast, nondestructive, near-infrared (NIR)-based method for measuring the hematocrit in DBSs.

Method: 

Using a partial least squares algorithm, an NIR-based quantification method was developed for measuring hematocrit levels of 0.19–0.49 L/L. Residual venous blood of 522 patients was used to build this partial least squares model. The validity of the method was evaluated using 40 patient samples. DBSs were created by adding a small amount (50 µL) of blood on a Whatman filter paper and drying for 24 hours in a desiccator cabinet. The robustness was evaluated by measuring 24 additional samples with a high hemolysis, icterus, and lipemia (HIL) index. The hematocrit values obtained using a Sysmex XN hemocytometry analyzer were used as reference.

Results: 

The difference between hematocrit measurements obtained with NIR spectroscopy and a hemocytometry analyzer was <15% for the 40 samples. The accuracy (≤9%) and precision (≤7%) for all the quality control samples were within the acceptance criteria of <15%. The intraassay and interassay coefficient of variability was ≤3% and ≤6%, respectively, for the different quality control levels. There were no deviations in the measurements for the samples with high HIL indices. The stability of hematocrit in DBS was up to 14 days for all levels.

Conclusions: 

We developed and validated a hematocrit model using NIR spectroscopy. This nondestructive, accurate, and reproducible method has a short analysis time (51 seconds), and can be used to analyze DBS samples stored for up to 2 weeks in a desiccator cabinet.