Evaluation of a point-of-care calcium device in bovine plasma and serum

Abstract

Hypocalcemia is a common metabolic disease in dairy cows, and it is defined as total calcium (tCa) blood concentration <2.0 mmol/L. The alternatives for the gold standard test to measure tCa in bovine blood are limited. Therefore, our objective was to compare the performance of the calcium (Ca) point-of-care compact analyzer (POC; ARKRAY Inc.) device with the gold standard method to measure bovine blood tCa concentration. Blood samples (n = 151) from dairy cows were collected within 24 h postpartum from multiparous and primiparous dairy cows for serum and plasma. Then, serum and plasma were stored at −80°C until further analyses with the gold standard method on an automatic analyzer (Cobas C501 analyzer; Roche Diagnostics) and the POC device. The tCa blood concentration was measured in the laboratory in plasma and serum samples using both methods within 10 mo of sample collection. Correlation coefficients (Spearman), coefficients of variation (CV, %), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, Passing and Bablok regression, and Bland-Altman agreement test were performed between the gold standard and the POC device. The range and median tCa plasma concentrations measured with the POC device were 1.1 to 2.8 mmol/L and 2.4 mmol/L, respectively. The range and median tCa serum concentrations measured with the POC device were 1.1 to 2.7 mmol/L and 2.3 mmol/L, respectively. The tCa blood concentrations range and median with the gold standard were 1.1 to 2.6 mmol/L and 2.3 mmol/L. The hypocalcemia prevalence of our study population was 11.2%. The CV were 1.89% and 0.55% for low and high tCa in plasma samples measured with the POC, respectively. The CV were 2.57% and 1.58% for low and high tCa in serum, respectively. The Spearman correlation coefficient showed a strong correlation between the gold standard and the POC device for both serum and plasma tCa concentration. The sensitivity of the POC device for both plasma (41.1%) and serum (64.7%) Ca was poor. However, the specificity of the POC device was perfect in plasma (99.2%) and serum (99.2%). The PPV in plasma and serum were 87.5% and 91.6%, respectively. Negative predicted values were 93.0% and 95.6% in plasma and serum. The mean (95% CI) difference between the gold standard and the POC device in plasma and serum were 0.35 (−0.52, 1.23) mmol/L and 0.19 (−0.53, 0.92) mmol/L, respectively. Finally, we observed a strong correlation between the POC device and the gold standard method for tCa plasma and serum. However, the clinical application of the POC device should be carefully considered because its ability to detect cows with hypocalcemia in serum or plasma samples was poor. However, the device performed better than previously analyzed POC devices and needs further improvement to be a valuable tool for the dairy industry.