Comparison of Two Methods for the Measurement of Blood Plasma and Capillary Blood Glucose in Tropical Highland Grassing Dairy Cows

SIMPLE SUMMARY: Blood glucose is crucial for milk production in dairy cows. The use of some human-customized blood glucometers has been evaluated for the measurement of cow blood with acceptable outcomes. However, there is scarce information validating the performance of cow-customized blood glucometers, like Centrivet GK (CVGK). We evaluated and compared the performance of the glucometer CVGK in tropical highland grazing cows against the measurement of glucose concentrations in plasma and serum via an enzymatic/photometric assay, considering the glucose measurements in serum as the reference method. Our study indicates that the measurement of glucose concentrations in plasma or by using CVGK is not reliable compared to the reference method used in our research. Thus, in the environmental and technical conditions of the study, the use of this cow-side glucometer cannot be recommended for glucose measurement in cows. ABSTRACT: (1) Background: There is lack of published studies validating specific cow-side glucometers such as Centrivet GK (CVGK). (2) Methods: The aims were (1) to measure and compare the blood glucose concentrations in 52 tropic highland grassing cows by using CVGK and the traditional enzymatic/photometric assay (EPA) in plasma and serum (reference method) and (2) to establish if glucose concentrations obtained via these methods could be affected by several demographic and zootechnical parameters of the dairy herd evaluated. (3) Results: Glucose concentrations were significantly (p = 0.00) affected by the method used for their measurement. The intra-assay coefficient of variation (CV) for glucose concentrations in plasma EPA and for CVGK was 14% for both methods with serum EPA, whereas the inter-assay CV for plasma EPA and CVGK was 8% and 13.7%, respectively, with serum EPA. Pearson correlation coefficient calculations between the reference method in serum and plasma presented a slightly positive significant (p = <0.000) correlation (r = 0.56), whereas there was not a significant (p = 0.413) correlation between serum EPA and CVGK (r = 0.135). The Passing and Bablok regressions were out of the ideal expected values for the slope (β = 1) and the intercept (α = 0) (11), whereas the Bland–Altman plots showed a bias of 5.29 ± 11.73 (mg/dL) for serum and plasma and 11.01 ± 15.74 (mg/dL) for serum and CVGK. The ROC curve showed no sensitivity in detecting normoglycemic cows (area = 53.7 %, e.d = 12.5 %, p = 0.759) for CVGK when compared to plasma EPA (area = 36.1 %, e.d = 14.2 %, p = 0.256). Plasma EPA exhibited a better but not significant effect in detecting hyperglycemic cows (area = 63.9%, e.d = 14.2%, p = 0.256) when compared to HHD (area = 46.3 %, e.d = 12.5 %, p = 0.759). General glucose concentrations, independently of the method used, were significantly (p = <0.001) greater in young cows when compared to adult and old cows. (4) Conclusions: Glucose concentration measurement in plasma by using EPA or in capillary blood via CVGK were not reliable methods when compared with the reference method.