AIMS: Continuous glucose monitoring (CGM) may complement or potentially replace fasting blood glucose (FBG) for basal insulin dose titration in type 2 diabetes (T2D). This retrospective analysis compared CGM-based titration with FBG-based titration using 7354 pairs of FBG and blinded CGM data from a clinical study in 68 people with T2D.
METHODS: Based on pharmacokinetic/pharmacodynamic simulations, the median of 3 lowest CGM values in the hour preceding the FBG timepoint ("1-h am nadir") was selected as basis for titration. Basal insulin doses were determined using 3 algorithms (Canadian INSIGHT, Treat2Target, AT.LANTUS). Absolute/relative dose adjustment differences, mean absolute relative differences, and relative dose errors between CGM- and FBG-based doses were calculated.
RESULTS: The 1-h am nadir was essentially equivalent to FBG for basal titration across all 3 algorithms. There was >90 % probability of the absolute dose adjustment difference being within tolerance. Mean absolute relative difference values were generally low, although higher for Treat2Target. Relative dose errors were mostly between -10 % and 10 %, indicating high agreement between CGM- and FBG-based titration and low clinical risk.
CONCLUSIONS: This study established that the 1-h am nadir can potentially be used as an FBG surrogate for basal insulin titration in T2D.