Real-time continuous glucose monitoring did not significantly reduce time below range but improved secondary measures of time in range, time above range, and mean glucose among patients with insulin-treated type 2 diabetes receiving maintenance hemodialysis.
Researchers conducted a prospective, randomized crossover trial at Emory Healthcare dialysis centers in Atlanta, Georgia. Eligible patients were aged 18 to 80 years, had type 2 diabetes and end-stage kidney failure, had received hemodialysis 3 times weekly for at least 3 months, and were receiving basal insulin alone or in combination with prandial insulin or incretin therapy.
Participants were randomly assigned to use the Dexcom G6 real-time continuous glucose monitoring system or standard capillary blood glucose testing for 30 days, followed by a 14-day washout period and crossover to the alternate strategy.
During the continuous glucose monitoring period, patients received instruction on sensor trend arrows, glucose patterns, and alarms, as well as written guidance for self-adjusting insulin. During the standard-care period, patients were instructed to perform capillary testing 2 to 4 times daily and maintain a glucose diary. Blinded continuous glucose monitoring was used during that period to collect comparative glucose data.
Of 52 enrolled patients, 39 were included in the modified intention-to-treat analysis, which required at least 10 days of continuous glucose monitoring data during each treatment period. Thirteen patients were excluded because of hospitalization, acute illness, lost sensors or devices, incomplete glucose data, or inability to complete study procedures.
The primary outcome was the percentage of monitored time with glucose below 70 mg/dL. Secondary outcomes included time below 54 mg/dL, time in the target range of 70 to 180 mg/dL, time above 180 and 250 mg/dL, mean glucose, glycemic variability, and hypo- and hyperglycemia measures.
Primary Outcome Did Not Differ
Time below 70 mg/dL was low during both monitoring periods and did not differ significantly, accounting for approximately 1.2% of monitored time with real-time continuous glucose monitoring and 1.3% with capillary testing. Time below 54 mg/dL also did not differ significantly.
The proportion of patients who experienced glucose below 70 mg/dL was 79% during real-time continuous glucose monitoring and 72% during capillary testing. Approximately half of patients in each period experienced glucose below 54 mg/dL.
The mean number of level 1 hypoglycemic episodes was higher during real-time continuous glucose monitoring, at 7.6 vs 6.0 during capillary testing, whereas level 2 episode counts did not differ. Overall time below range and the proportion of patients experiencing hypoglycemia did not differ significantly between monitoring strategies.
Secondary Glycemic Measures Improved
Mean time in the target range was 63% during real-time continuous glucose monitoring vs 55% during capillary testing.
Time above 180 mg/dL averaged 35% during real-time continuous glucose monitoring vs 44% during capillary testing. Time above 250 mg/dL averaged 12% vs 19%, respectively. Mean glucose was 174 mg/dL during continuous monitoring and 188 mg/dL during capillary testing.
The researchers reported that the improvements were driven primarily by less time spent above the hyperglycemia thresholds rather than by changes in time below range.
Hyperglycemic episodes nevertheless remained common. Every patient experienced glucose above 180 mg/dL during both periods. Glucose above 250 mg/dL occurred in 95% of patients during real-time continuous glucose monitoring and all patients during capillary testing.
The researchers reported no evidence of carryover effects for the time-in-range and time-above-range outcomes.
Findings Require Confirmation
The researchers acknowledged several limitations, including the small sample size, inclusion of only insulin-treated patients, and selection of hypoglycemia as the primary outcome. Because time below range was low in both groups, there was little opportunity for the intervention to improve the primary outcome.
They also noted the challenges of conducting research in this medically complex population, including illness, hospitalization, and the high burden of care that prevented some patients from starting or completing the study.
The researchers concluded that real-time continuous glucose monitoring improved time in range and reduced time above range during the 30-day intervention. They called for larger studies to determine whether improvements in glycemia correlate with better clinical outcomes.
Disclosures can be found in the published research.
Source: Diabetes Care
