Comparison of a portable blood glucose meter analyzer with a benchtop point-of-care chemistry analyzer for measurement of blood glucose concentration in client-owned ferrets (Mustela putorius furo)

Abstract

Hypoglycemia is common in pet ferrets (Mustela putorius furo) because of the high prevalence of insulinoma in this species. The objectives of this study were to evaluate agreement of a portable blood glucose meter (PBGM) with a benchtop point-of-care (POC) chemistry analyzer for measurement of blood glucose concentration in ferrets, and to assess the clinical impact of using a PBGM for blood glucose measurement. The benchtop POC analyzer was used as the reference analyzer (modified hexokinase method). Glucose concentration was measured from 82 blood samples from client-owned ferrets with the benchtop POC chemistry analyzer and the PBGM. Agreement and bias between measurements were assessed using the Bland-Altman method and a Passing-Bablok regression analysis. A modified Clarke error grid analysis was modelized to evaluate the clinical effect if the PBGM was used rather than the benchtop POC chemistry analyzer. Glucose values obtained with the PBGM were not in agreement with the benchtop POC chemistry analyzer, and it underestimated blood glucose concentration in many cases. However, variation was unpredictable (mean bias, -3.97 mg/dL; range, -52.2 to 64.8 mg/dL), with wide 95% LOA (-47.3 to 38.5 mg/dL). A Passing-Bablok linear regression analysis had a slope of 1.13 (95% confidence interval: 1.00–1.36), and an intercept of -20.78 (95% confidence interval: -38.92 to -9.90), highlighting presence of a proportional and a constant bias. Important clinical error would have occurred in 1% of cases with the PBGM. Unpredictable variation of glucose results obtained with the PBGM could have an important impact on clinical decision making. Thus, the use of a benchtop POC analyzer using hexokinase method for measurement of blood glucose concentration should be favored in ferrets for rapid onsite result obtention.