The Influence of Skin Thickness on Flash Glucose Monitoring System Accuracy in Dogs with Diabetes Mellitus

SIMPLE SUMMARY: A flash glucose monitoring system (FGMS) has been validated for use in dogs with diabetes mellitus and diabetic ketoacidosis. It continuously measures the glucose in the interstitial fluid through a small filament (5 mm long) inserted under the skin. Interstitial glucose concentrations are reportedly comparable to whole blood glucose concentrations. However, several factors can influence the performance of interstitial sensors, including the proportion of interstitial fluid in a tissue. The influence of skin thickness on flash glucose monitoring system accuracy has not been investigated in previous studies; therefore, the aim of this study was to evaluate whether FGMS accuracy is affected by skin thickness. On the basis of our results, skin thickness seems to affect FGMS measurements; the mean bias was significantly inversely correlated (p = 0.02; r = −0.6) with the mean skin thickness, and clinical accuracy according to ISO 15197:2013 criteria was observed only in dogs with skin thickness > 5 mm, with 99% of the results falling in zone A + B of the Parkes consensus error grid analysis. In dogs with thin skin (<5 mm), the clinical accuracy was low, and the results should be interpreted with caution. ABSTRACT: A flash glucose monitoring system (FGMS) has been validated for use in diabetic dogs. However, it is unknown whether skin thickness affects FGMS measurements. The aim of this study was to evaluate whether FGMS accuracy is affected by skin thickness. Fourteen client-owned diabetic dogs on insulin treatment were prospectively enrolled in the study. The dogs were divided into two groups according to their ultrasound-measured skin thickness: dogs with skin thickness < 5 mm (Group 1) and dogs with skin thickness > 5 mm (Group 2). On days 1, 7 and 14, glucose curves were obtained simultaneously using the FGMS and a validated portable blood glucose meter. Paired measurements were used to calculate the mean bias and to determine accuracy according to ISO 15197:2013 criteria. The mean bias was significantly inversely correlated (p = 0.02; r = −0.6) with the mean skin thickness. Clinical accuracy was observed only in Group 2, with 99% of the results in zone A + B of the Parkes consensus error grid analysis. In conclusion, skin thickness seems to affect FGMS measurements, and the device is accurate in dogs with thicker skin (>5 mm); in dogs with thin skin (<5 mm), the clinical accuracy is low, and the results should be interpreted with caution.