The objective of this study was to evaluate the suitability of the electronic handheld devices FreeStyle Precision (FSP; Abbott Germany, Wiesbaden, Germany) and GlucoMen LX Plus (GML; A. Menarini GmbH, Vienna, Austria) for the measurement of β-hydroxybutyrate (BHBA) in whole blood in dairy goats. Additionally, glucose concentration was analyzed with the FSP device. For method comparison, the samples were also analyzed in the laboratory by standard methods, which served as the gold standard in our study. A further objective was to evaluate the agreement between samples obtained by minimal invasive venipuncture of an ear vein and measurements of whole blood samples from the jugular vein (vena jugularis). In total, 173 blood sample pairs collected from 28 goats were obtained from an ear vein and from the jugular vein. The Spearman correlation coefficients (rsp) for BHBA concentrations determined with the FSP or GML and the gold standard were 0.95 and 0.85 for the ear vein and 0.98 and 0.88 for the jugular vein, respectively. Bland-Altman plots of differences showed a positive bias of 0.12 (ear vein) and 0.21 (jugular vein) when determination was performed with the FSP and a negative bias of 0.21 (ear vein) and 0.24 (jugular vein) when using the GML. For the FSP, applying the adjusted thresholds determined by ROC analysis of 0.9 (ear vein) and 1.0 mmol/L (jugular vein) allowed to distinguish between healthy goats and animals with hyperketonemia with sensitivities (Se) and specificities (Sp) for samples from the ear vein of 0.98 and 0.85, and from the jugular vein of 0.99 and 0.94, respectively. For the GML, adjusted thresholds were 0.5 mmol/L for the ear vein (Se=0.94, Sp=0.75) and 0.6 mmol/L for the jugular vein (Se=0.88, Sp=0.91). Repeated analyses of defined whole blood samples showed average inter- and intraassay coefficients of variation of 6.6 and 7.3% for FSP, and of 35.6 and 35.4% for GML, respectively. Test characteristics for determining glucose concentration with the FSP compared with the gold standard were poor (ear vein: rsp=0.41; jugular vein: rsp=0.51), with low validity to distinguish between hypo- and normoglycemia (Se=0.71, Sp=0.66). The present study showed good test characteristics for the FSP and moderate for the GML device for determining BHBA concentrations in whole blood. Additionally the results demonstrate the suitability of testing BHBA concentration in a blood drop obtained from an ear vein. Based on the results of this study, the FSP device is not suitable to differentiate normo- from hypoglycemia in goats.
