EN06 - Comparison of Two Continuous Glucose-Monitoring Systems in Diabetic Dogs

Abstract:

Background: Continuous glucose monitoring systems (CGMs) are increasingly used in monitoring of canine diabetes. Older models of FreeStyle Libre (FSL) have been validated in dogs but the FSL3+ and the Dexcom G7 (DG7) have not been previously evaluated.

Objectives: Assess the accuracy of the FSL3+ and DG7 compared with the AlphaTrak3 in diabetic dogs.
Animals: Five clinically stable, client-owned diabetic dogs.

Methods: Prospective study. On each dog, DG7 and FSL3+ sensors were placed on opposite sides of the craniodorsal thorax. At least 24hr after placement, blood glucose concentrations (BG) were measured (AlphaTrak3) q30 minutes over 5 hours in each dog and compared with interstitial glucose concentrations (IG). CGMs were left in place for approximately 12-14 days. Method comparisons were assessed using Bland-Altman analysis, Parke’s Error Grid analysis (PEGA), and mean absolute relative differences (MARD).

Results: Compared to BG, mean (± SD) bias was 13.26 mg/dL ± 18.96 and -23.56 mg/dL ± 36.61 and mean percent difference was 7.84% ± 10.02 and -7.49% ± 14.16 for the FSL3+ and DG7, respectively. Compared to BG, MARD was 8.13% and 13.54% for the FSL3+ and DG7, respectively. Both CGMs had 100% of their readings within zones A and B of the PEGA, both met clinical accuracy standards, but neither met analytical accuracy standards. Mild skin reactions were observed with both the DG7 and FSL3+.

Conclusions: The FSL3+ tends to overestimate BG concentrations and the DG7 tends to underestimate BG concentrations compared to the AlphaTrak3. Both CGMs met clinical accuracy standards.