The Evidence Base for Continuous Glucose Monitoring

Abstract

Twenty-seven published randomized controlled trials (RCTs) assessing outcomes of continuous glucose monitoring (CGM), involving a total of 3,826 patients, have been published to date. Although the number of patients in each study has been small compared to drug trials, cumulative evidence indicates a benefit of CGM for patients treated with either continuous subcutaneous insulin infusion (CSII) or a multiple daily injection (MDI) insulin regimen. Additionally, some data suggest that CGM may benefit people with type 2 diabetes who do not use insulin therapy. Overall, RCTs have shown improved glucose control in patients with higher initial A1Cs (often in the range of 7.8– 8.8%) using CGM compared to self-monitoring of blood glucose (SMBG). People who wear their CGM device most consistently derive the most benefit. Time spent in the designated hypoglycemia range (usually <70 mg/dL) was reduced in some studies, particularly in those with patients selected for having a higher risk of hypoglycemia. These patients tended to have lower baseline A1Cs (in the range of 6.5–7.5%). Rates of severe hypoglycemia generally have not differed between CGM and non-CGM groups, and these rates have been low across all studies. Studies fall into a few basic categories: adults with type 1 diabetes (8 trials, 698 patients), adults with type 2 diabetes (4 trials, 547 patients), children with type 1 diabetes (2 trials, 227 patients), adults plus children with type 1 diabetes (7 trials, 1,084 patients), adults with type 1 or type 2 diabetes (3 trials, 655 patients), and women during pregnancy with either type 1 diabetes or gestational diabetes mellitus (GDM) (3 trials, 585 patients). Table 1 lists general findings from all of these trials. It is important to note that some trials used A1C or time in range as the primary endpoint, whereas others used time in a hypoglycemic range as the primary outcome. Readers should also be aware that Table 1 is not a meta-analysis per se, but rather includes studies identified through a literature search of PubMed and Ovid MEDLINE, as well as all prior reviews and studies in their reference lists. Only RCT data are included; observational studies and extension phases of RCTs also have been performed but are not represented here. The first trials, from the early 2000s, used intermittent CGM. Some used “professional” CGM, in which patients were blinded to the CGM data (see the article on p. 8 of this compendium), and others followed an intermittent use schedule. As time progressed, the trials reflected evolving use of CGM to the current day. That is, earlier studies began to suggest that CGM could improve outcomes, but lack of access to real-time data limited benefit. More recent studies of real-time CGM, in which around-the-clock data are available, have shown more benefit in terms of reduction in both A1C and time spent in a hypoglycemic range. A major impediment to interpreting CGM studies is that no uniform standard has been employed for teaching people with diabetes how to use continuous data, and no standard follow-up is provided to ensure that dose adjustments are made. In some trials, written instructions were provided to patients regarding insulin dose adjustments, but in many others, targeted education was not provided beyond how to use the device. Additionally, rapid advances in technology are not well represented in the literature, although data from newer systems, such as the Dexcom G5 Mobile (Dexcom, San Diego, CA) and the FreeStyle Libre (Abbott, Alameda, CA), are becoming available. The Evidence Base for Continuous Glucose Monitoring