Method comparison and validation of a prototype device for measurement of ionized calcium concentrations cow-side against a point-of-care instrument and a benchtop blood-gas analyzer reference method.

The objective of this study was to assess an optimized ion-selective electrode Ca-module prototype as a potential cow-side device for ionized Ca (iCa) measurements in bovine blood. A linearity experiment showed no deviation from linearity over a range of iCa concentrations compared with a commercial point-of-care (POC) device commonly used in the field (POCVS ; VetScan i-STAT, Abaxis North America, Union City, CA) and a laboratory gold standard benchtop blood-gas analyzer [reference analyzer (RA); ABL-800 FLEX, Radiometer Medical, Copenhagen, Denmark]. Coefficient of variation on 3 samples with high, within-range, and low iCa concentrations ranged from 1.0 to 3.9% for the prototype. A follow-up validation experiment was performed, in which our objectives were to (1) assess the performance of the prototype cow-side against the POCVS (farm gold-standard) using fresh non-anticoagulated whole-blood samples; (2) assess the performance of the prototype and the POCVS against the RA in a diagnostic laboratory using blood collected in a heparin-balanced syringe; and (3) assess the agreement of the prototype and POCVS on-farm (fresh non-anticoagulated whole blood) against the RA on heparin-balanced blood. Finally, sensitivity and specificity of the results obtained by the prototype and the POCVS cow-side compared with the results obtained by the laboratory RA using 3 different iCa cut points for classification of subclinical hypocalcemia were calculated. A total of 101 periparturient Holstein cows from 3 dairy farms in New York State were used for the second experiment. Ionized Ca results from the prototype cow-side were, on average, 0.06 mmol/L higher than the POCVS . With heparin-balanced samples under laboratory conditions, the prototype and POCVS measured an average 0.04 mmol/L higher and lower, respectively, compared with the RA. Results from the prototype and POCVS cow-side were 0.01 mmol/L higher and 0.05 mmol/L lower, respectively, compared with results from the laboratory RA on heparinized blood. Sensitivity and specificity for the prototype and the POCVS under farm conditions at 3 potential subclinical hypocalcemia cut points were 100 and ≥93.5%, respectively. This novel ion-selective electrode Ca-module could become a rapid low-cost tool for assessing iCa cow-side, while qualitatively allowing classification of subclinical hypocalcemia on-farm.