Compliance With FDA iCGM Special Controls is Dependent on Study Design and Procedures

In a recent article, we reported the results of a performance evaluation of three current-generation continuous glucose (CGM) monitoring systems. ¹ Two of these CGM systems were the FreeStyle Libre 3 (FL3) and Dexcom G7 (DG7), both of which received “integrated” CGM (iCGM) clearance by the U.S. Food and Drug Administration (FDA).^2-4 In our study, 24 adults with type 1 diabetes mellitus wore one sensor of each CGM system in parallel for up to 15 days. The sensors of DG7 were replaced on day 5 to account for the different lifetimes of the sensors. Three seven-hour in-clinic sessions with 15-minute comparator blood glucose level measurements using the YSI 2300 STAT PLUS (YSI, venous) and Cobas Integra 400 Plus (INT, venous) laboratory analyzers and the Contour Next (CNX, capillary) home-use blood glucose monitoring system were conducted on days 2, 5, and 15. Simultaneously, glucose level excursions with transient hyperglycemia and hypoglycemia were induced.

Aside from CNX, which is not an FDA-accepted comparator, the study design and procedures can be considered to comply with iCGM special controls. Therefore, we analyzed the data from our study accordingly, using a clustered bootstrap approach (treating individual in-clinic sessions as clusters) to calculate the lower confidence bounds (LCB) for accuracy metrics of criteria (A)-(G). ⁵ The results are shown in Table 1.

Table 1.

iCGM Special Controls Results for Two iCGM Systems Using Three Different Comparators.

For requirements (A)-(G), results are given as point estimates with 95% lower confidence bounds in parentheses. Green cells indicate that both point estimate and confidence bound comply with iCGM special controls, orange cells indicate that point estimate but not confidence bound comply with iCGM special controls and red cells indicate that neither point estimate nor confidence bound comply with iCGM special controls.

YSI: YSI 2300 STAT PLUS laboratory analyzer used with venous samples, INT: Cobas Integra 400 Plus laboratory analyzer used with venous samples, CNX: Contour Next blood glucose monitoring system used with capillary samples. CGM: continuous glucose monitoring, Comp: comparator measurement, RoC: rate of change. [See online article for color table.]

Evaluating all combinations of CGM systems and comparators, neither FL3 nor DG7 met all iCGM special controls (A) to (G). This was particularly evident in the results against YSI, which served as a comparator in the pivotal studies.^2,3 However, it is important to note that for several accuracy metrics of criteria (A) to (G), the point estimates exceed the specified limits, while the LCBs did not (orange cells), indicating that a larger sample size may have led to compliance. This was especially true for FL3 against CNX. All combinations of CGM systems and comparators fulfilled requirements (H) to (K).

Apart from the smaller sample size, the discrepancies in iCGM compliance compared to the pivotal studies likely stem from: (1) the more demanding in-clinic session procedures with rapid changes toward hypoglycemia and hyperglycemia resulting in a mean absolute rate of change of ~1.3 mg/dL/min in our study compared to ~0.8 mg/dL/min in the pivotal studies, and (2) a different bias of FL3 and DG7 against the YSI comparator with +10.1 mg/dL (FL3) and +9.0 mg/dL (DG7) in our study compared to approximately −5.5 mg/dL (FL3) and −1.3 mg/dL (DG7) in the pivotal studies.

Despite these findings, our results show that FL3 and DG7 are safe and accurate CGM systems. However, our study also demonstrates that any minimum performance criteria for CGM systems, such as the iCGM special controls or the proposed eCGM designation for CGMs in Europe, ⁶ must incorporate a rigorous definition of requirements for study design and procedures. Specifically, requirements for the distribution of comparator data as well as the accuracy and traceability of comparator measurements need to be established. Currently, the working group on CGM of the International Federation of Clinical Chemistry and Laboratory Medicine is preparing a proposal for requirements to be introduced this year.