Continuous glucose monitoring in diabetes

Clinical question

Compared with self-monitoring blood glucose (SMBG), does continuous glucose monitoring improve clinical outcomes or hemoglobin A_1c (HbA_1c) levels for adults with diabetes taking insulin?

Bottom line

Continuous glucose monitors use subcutaneous sensors and include real-time monitors (RTMs) and flash monitors (FMs). In patients with type 1 diabetes, RTM use is associated with lower rates of severe hypoglycemic events (6%) compared with SMBG use (8%), with no reported events in those with type 2 diabetes. Effects on HbA_1c may not be clinically meaningful. Compared with SMBG, FMs do not differ regarding rates of severe hypoglycemic events, and effects on HbA_1c are inconsistent (type 1) or similar (type 2). Cost may limit use.

Evidence

Results were statistically different unless indicated. A minimal clinically important HbA_1c change was defined as 0.5%.¹

• Type 1 diabetes, RTMs vs SMBG (8 systematic reviews; 11 to 22 RCTs; 1399 to 2461 patients): At 4 to 12 months, the percentage of severe hypoglycemic events requiring third-party assistance (3 systematic reviews without substantial methodologic flaws^1-3) ranged from 3.5% to 8% RTMs vs 6.5% to 10% SMBG, number needed to treat=30 to 55.

  • - There was no difference in patient satisfaction.¹

  • - The mean difference in HbA_1c was about 0.2% to 0.4% lower with RTMs at 4 to 6 months.^1-3

• Type 1 diabetes, FMs vs SMBG at 6 months: No differences found in rates of severe hypoglycemic events,¹ discontinuation rates, or HbA_1c in 2 systematic reviews.^1,4

  • - A recent RCT (N=156) showed no difference in rates of severe hypoglycemic events, and HbA_1c was 0.5% lower with FMs.⁵

• Type 2 diabetes, RTMs vs SMBG (most taking insulin): In 2 systematic reviews there were no reports of severe hypoglycemia,^6,7 and HbA_1c was about 0.25% to 0.5% lower with RTMs for 3 to 6 months.^6,7

• Type 2 diabetes, FMs vs SMBG: At 2.5 to 6 months, no differences were found in rates of severe hypoglycemic events or HbA_1c.⁸

  • - Discontinuation: 6% (FMs) vs 15% (SMBG) (calculated by authors), number needed to treat=12.

• Limitations: Most RCTs were unblinded and industry funded. Quality-of-life scores were inconsistently reported.

Context

• Real-time monitors automatically display readings (eg, Dexcom 6) whereas FMs require manual upload (eg, FreeStyle Libre). Replace sensors every 7 to 14 days.⁹

• Cost per year: approximately $2500 to $6000 vs $1200 for 4-times daily SMBG.⁹

Implementation

New diabetes guidelines emphasize the use of medications with proven cardiorenal benefit in patients at high risk, with glycemic control a secondary focus.¹⁰ Exercise, healthy dietary choices, and medication adherence should continue to be encouraged. Advising non–insulin dependent patients to check glucose readings frequently is likely unnecessary. Continuous glucose monitor readings may lag behind SMBG by up to 15 minutes, particularly after exercising or eating.¹¹