VARIABLES INFLUENCING THE OPTIMAL SAMPLING DURATION OF INTERMITTENTLY-SCANNED CONTINUOUS GLUCOSE MONITORING TO ASSESS 90 DAYS OF GLYCEMIA RISK INDEX. FINDINGS FROM A MULTICENTRIC REAL-WORLD STUDY.

Background and Aims. The glycemia risk index (GRI) is a novel composite score to evaluate the quality of glucose monitoring profiles. Our study investigated which variables might affect the optimal sampling duration for the assessment of GRI through 90 days of intermittently-scanned continuous glucose monitoring (isCGM). Methods. This real-world, retrospective study included 539 90-day isCGM-data from 367 adults with type 1 diabetes (T1D) on multiple daily injection (MDI) therapy (F/M 164/203, age 42±16 years, diabetes duration 21±14 years). Coefficients of determination (R2) for GRI and its hyper- and hypoglycemia components were computed for sampling periods from 2 to 12 weeks versus the whole 90-day intervals. Correlations were considered strong for R2≥0.88. Results. Strong correlations with 90 days were found for GRI and for its hyperglycemia component with 2-week-long data samplings (R2=0.88 and 0.89, respectively), while 6 weeks were needed to achieve a significant correlation for the hypoglycemia component (R2=0.90). When stratifying by clinical variables, higher R2 values for GRI and for its hyperglycemia component were found with lower coefficient of variation (<36%), lower time below range (TBR) (<4%), lower frequency of hypoglycemic episodes (<30 episodes/90days) and lower mean hypoglycemia duration (<90 minutes); in contrast, the hypoglycemia component showed stronger correlations with higher TBR, lower HbA1c (≤7.5%) and higher frequency of hypoglycemia. Conclusions. Two-week-long isCGM data could provide an effective estimation of GRI through 90 days; however, HbA1c, glycemic variability (expressed as CV), TBR and frequency and mean duration of hypoglycemic episodes might affect the accuracy of the chosen sampling duration.