105-LB: Does HbA1c Accurately Predict Time-in-Range?

Abstract

Background: Onduo’s Virtual Diabetes Clinic (VDC) incorporates a continuous glucose monitor (CGM), telemedicine, and a coaching app into the management of patients with type 2 diabetes (T2D). A recent meta-analysis of studies involving CGMs (Vigersky & McMahon 2019) showed HbA1c and time in range (%TIR) are strongly correlated. Methods: All patients in the program with CGM usage in the 45 days prior to a follow-up HbA1c were analyzed. Mean HbA1c and %TIR were added to the previously published meta-analysis plot where each study or cohort is represented by a single dot. Additionally, a regression model was fit on the patient-level data. Results: Patient-level results from the VDC confirm the two metrics are strongly correlated (r=.-78). At the population-level, a mean %TIR of 84% in 194 VDC patients with mean HbA1c of 7.0% is 18% higher than predicted from the meta-analysis regression line. Cohorts associated with the 3 T2D studies in the meta-analysis were the 3 with the highest %TIR versus predicted, and 1 study with mixed type 1/2 was also higher than predicted. Conclusions: Our findings add support to the strong correlation observed between TIR and A1c overall, and suggest potentially distinct calibrations in T1D vs. T2D populations. Additional CGM data in T2D will clarify if separate prediction formulae are necessary. Disclosure R.F. Dixon: Employee; Self; Onduo LLC. D.P. Miller: Employee; Self; Verily Life Sciences LLC. Stock/Shareholder; Self; Verily Life Sciences LLC. A. Majithia: Consultant; Self; Onduo LLC. A. Armento Lee: Employee; Self; Verily Life Sciences LLC. S. Robertson: Employee; Self; Verily Life Sciences LLC. M. Germanos: None. H. Zisser: Employee; Self; Verily Life Sciences LLC. Stock/Shareholder; Self; Verily Life Sciences LLC.