Evaluation of the Accuracy of FreeStyle Libre 2 for Glucose Monitoring in White New Zealand Rabbits

Abstract

Background: Studies are currently being conducted on rabbits requiring serial glucose monitoring. The FreeStyle Libre 2 (FSL2), a serial glucose monitoring device, has been validated in humans, dogs and cats, but not in rabbits.

Objectives: This study aimed to evaluate the accuracy of the FSL2 in rabbits.

Methods: Six healthy rabbits were used in this study. Interstitial glucose (IG) was measured using the FSL2, and blood glucose (BG) was measured using a portable blood glucose meter (PBGM); their results were compared with those from a clinical chemistry analyser. For the first 3 h, IG and BG were measured at 1-h intervals. Subsequently, they were measured every 8 h over a 48-h period. Regular insulin 0.2 U/kg was then administered to the rabbits, and IG and BG were measured every 15 min over a 90-min period.

Results: Before insulin treatment, no measurements fell within the hypoglycaemic range (BG < 100 mg/dL). In the euglycaemic range (BG ≥ 100 mg/dL), the PBGM and FSL2 showed 85.7% and 23.8% accuracies, respectively. After insulin treatment, the PBGM showed 95.5% and 81.3% accuracies in the hypoglycaemic and euglycaemic ranges, respectively. The FSL2 showed 68.1% and 37.5% accuracies in the hypoglycaemic and euglycaemic ranges, respectively. Parkes consensus error grid analysis showed that the PBGM and FSL2 had 100% agreement for Zones A (no effect on clinical action) and B (altered clinical action unlikely to affect outcome) in rabbits with and without insulin treatment.

Conclusions: There was limited agreement between the FSL2 and reference standard BG measurements. However, the FSL2 allows clinically acceptable identification of hypoglycaemic states in rabbits.

Keywords: FSL2; FreeStyle Libre 2; hypoglycaemia; interstitial glucose; rabbit.

FIGURE 1

Correlation between readings from (A) the portable blood glucose meter and the reference method, and (B) the FreeStyle Libre 2 and the reference method. The solid line represents the best fit. FSL2, FreeStyle Libre 2; PBGM, portable blood glucose meter; REF, reference method.

FIGURE 2

Bland–Altman analysis of glucose measurements between (A) the portable blood glucose meter and (B) the FreeStyle Libre 2 and the reference method. (A) A negative proportional bias and heteroscedasticity with more variation for higher blood glucose concentrations is apparent. Constant bias was estimated to be −8.818 mg/dL and the 95% limits of agreement were −24.610 to 6.974 mg/dL. (B) Considerable variation between methods and heteroscedasticity are the most apparent findings. Constant bias was estimated to be 17.553 mg/dL and the 95% limits of agreement were −11.532–46.637 mg/dL. FSL2, FreeStyle Libre 2; PBGM, portable blood glucose meter; REF, reference method.

FIGURE 3

The Parkes consensus error grid (for Type 1 diabetes mellitus) of (A) the PBGM readings for noninsulin‐treated rabbits, (B) the FSL2 readings for noninsulin‐treated rabbits, (C) the PBGM readings for insulin‐treated rabbits and (D) the FSL2 readings for insulin‐treated rabbits. Zones are categorized as follows: (Zone A) no effect on clinical action; (Zone B) altered clinical action unlikely to affect outcome; (Zone C) altered clinical action likely to affect clinical outcome; (Zone D) altered clinical action could have substantial medical risk and (Zone E) altered clinical action could have dangerous consequences. Based on the ISO 15197:2013 criteria, (Zone C) altered clinical action likely to affect clinical outcome; (Zone D)FreeStyle Libre 2; PBGM, portable blood glucose meter.