Comparison of human point-of-care glucometer (Guide Me) and lactometer (Lactate Plus) to an automated chemistry analyzer for measurement of blood concentrations of lactate and glucose in juvenile commercial pigs

Abstract

Porcine glucose and lactate concentrations measured with a point-of-care glucometer (POC-Glu; Guide Me) and a POC lactometer (POC-Lac; Lactate Plus) were compared to those from a comparative method (CM; COBAS-501c). Fresh whole blood samples (n = 175 glucose; n = 272 lactate) from 10 healthy juvenile commercial-cross pigs (5 barrows, 5 gilts; 79–91-d-old) were collected over 12 d under various conditions. Comparisons were made with the Passing–Bablok regression and Bland–Altman method with multiple measurements per subject. The allowable total error (TE_a) was set at 20% for glucose and 40% for lactate. Correlation with CM was high for POC-Glu (r = 0.886; 95% CI [0.849, 0.914]) and very high for POC-Lac (r = 0.935; 95% CI [0.918, 0.949]) with constant and proportional biases (intercept = –0.52, 95% CI [–0.98, −0.13] mmol/L, slope = 0.06, 95% CI [0.06, 0.07] mmol/L for glucose; intercept = 0.075, 95% CI [0.050, 0.104] mmol/L, slope = 0.797, 95% CI [0.775, 0.819] mmol/L for lactate). Acceptance limits based on combined inherent imprecision (CIP) were ±5.2% for glucose and ±8.1% for lactate. The observed total errors (TE_obs) were 5.42% for POC-Glu and 29.9% for POC-Lac at the decision threshold of 3.0 mmol/L, both within the set TE_a. Both POC-Glu and POC-Lac are practical and reliable tools for use in juvenile pigs, with satisfactory agreement to the reference method, although interpretation of POC-Lac results requires greater caution due to a significant matrix effect, underscoring the need for method- and matrix-specific RIs, particularly for lactate.