Interstitial glucose monitoring has acceptable clinical accuracy in juvenile dogs

Abstract

Abstract OBJECTIVE To compare the performance of an interstitial glucose monitor (IGM) versus a portable blood glucose monitor (PBGM) in sick juvenile dogs in a veterinary ICU. ANIMALS 16 client-owned dogs admitted to the university teaching hospital under 1 year of age with systemic illness. PROCEDURES Paired interstitial and blood glucose samples were collected. A third glucose measurement with a reference method was obtained when IGM and PBGM values were clinically disparate. Analytical accuracy was measured by Pearson correlation and agreement statistics, including mean absolute relative difference (MARD), bias, and 95% limits of agreement. The Parkes consensus error grid analysis was performed to assess clinical accuracy. RESULTS 159 paired glucose measurements were available for analysis. Comparison of IGM readings to PBGM measurements resulted in an MARD of 15.4% and bias of –2.6%, with the 95% limits of agreement ranging from –42.5% to 37.4%. Positive correlation between IGM and PBGM (Pearson r = 0.65) was found. On consensus error grid analysis, 100% of the pairs fell into clinically acceptable zones (74.2% in zone A, and 25.8% in zone B). When disparate IGM and PBGM readings were compared to a laboratory reference standard (n = 13), both methods resulted in high MARD and wide limits of agreement. CLINICAL RELEVANCE The IGM provides clinically acceptable glucose measurements compared to PBGM to monitor glucose levels in juvenile dogs in a clinical setting. Further clinical studies with a larger sample size, particularly in the hypoglycemic range, are needed to assess IGM performance in the lower glucose range.