ESPE Abstracts

Introduction: Continuous Glucose Monitoring (CGM) devices have introduced new metrics to assess metabolic situation of patients with type 1 diabetes mellitus (T1DM). Among them, the glucose management indicator (GMI) provides a proxy to HbA1c, however, these measures do not always match.

Purposes: To evaluate the concordance between GMI and HbA1c, and determinate in which scenarios GMI could replace HbA1c.

Subjects and Methods: Retrospective study of children and adolescents aged ˂18 years with T1DM using FreeStyleLibre® Flash Glucose Monitor. Data of the CGM are collected from the Libreview® cloud system and correspond to 2 weeks period and sensor active ≥70% of the time. The HbA1c is measured capillary by the DCA® Vantage analyzer. Variables analyzed: average glucose, CV, GMI, HbA1c, time above range (TAR), time very high (TVH), time in range (TIR), time bellow range (TBR), time very low (TVL), time sensor is active, and glucose lectures/day. Inclusion criteria: DM last ≥6 months, time between HbA1c determination and CGM evaluation ≤15 days. Variables are analyzed with the Stata14 software. Qualitative and quantitative variables are described as frequencies and mean (SD) respectively. GMI and HbA1c are compared by the t Student-Fisher test. Linear regression analysis examines the influence of the reported variables on the HbA1c-GMI difference, in case of association the cutoff point of each variable at which GMI and HbA1c can be considered to be equivalent is determined by the t Student-Fisher test: first not significant value. A p≤0.05 is considered significant.

Results: 201 subjects are enrolled (56.72% males), with 4.59 years (3.12) of T1DM evolution. The time elapsed between the determination of HbA1c and the CGM evaluation is 0.0003 days (0.01). The mean HbA1c is 7.2% (0.8) and the mean GMI is 7.49% (0.72), and they are significantly different (p˂0.001). In 32.85% of the patients the HbA1c-GMI discordance is ≤0.2%.

Conclusions: In case of average glucose ≤161 mg/dl, TAR ≤30%, TVH ≤6%, TIR ≥70%, TBR ≤0% or lectures/day ≥30, GMI agrees with HbA1c, and HbA1c does not add any information compared to the sensor, avoiding its determination is both time and cost saving.