Evaluation of the Analytical and Diagnostic Performances of a Point-of-Care Device for Hematocrit Measurement in Calves

Abstract

Objective: To evaluate the performance of a point-of-care hemoglobin/hematocrit measurement device (POCHD), validated for human medicine, for measuring hematocrit in calves. Design: Prospective study conducted between July 2022 and April 2023. Setting: University teaching hospital. Animals: One hundred thirty-three calves <1 month old, suffering from various diseases. Interventions: The microhematocrit method (MHM) and a POCHD were used to measure hematocrit in venous blood samples. The MHM served as the reference method. The analytical performance of the POCHD was evaluated using Bland-Altman plots and calculating the intraassay coefficient of variation and observed total error (TEobs). Diagnostic performance was assessed by implementing receiver operating characteristic curve analysis and Cohen's kappa statistic. Measurements and main results: Bland-Altman plots showed a mean bias of -0.03 L/L (-3.26%) (95% confidence interval: -0.04 to -0.03 [-4.02% to -2.51%]). The intraassay coefficient of variation for the POCHD was 2.4% for low hematocrit values and 0.7% for high hematocrit values. TEobs ranged from 3.2% to 32.9% (median: 18.1%). Receiver operating characteristic curve analysis at hematocrit thresholds of <0.22 L/L (22%) for anemia and >0.33 L/L (33%) for hemoconcentration showed optimized thresholds of <0.19 L/L (18.5%) and >0.33 L/L (33.2%), respectively, for the POC device to differentiate anemic from nonanemic calves, as well as calves with dehydration-related hemoconcentration from those without. At the thresholds of 0.22 L/L (22%) and 0.33 L/L (33%), Cohen's kappa statistic found agreements of 90.2% (kappa = 0.536) and 91% (kappa = 0.820), respectively, between the POCHD and the MHM. Conclusions: Clinicians should be aware that the POCHD underestimates hematocrit values and exhibits high TEobs; however, when adjusted thresholds are applied, the POCHD shows good sensitivity and specificity for distinguishing anemia and dehydration-related hemoconcentration.