实时持续葡萄糖监测 vs 自我血糖监测在2型糖尿病中的随机对照试验

原文

BACKGROUND: Type 2 diabetes is the most common metabolic disorder worldwide, accounting for about 90% of people living with diabetes. Glycated haemoglobin (HbA1c), a measure of chronic glycaemic exposure, correlates with the risk of long-term complications, which can result in substantial morbidity for people with diabetes and major costs to health-care systems. The value of continuous glucose monitoring (CGM) in people with type 2 diabetes managed with basal insulin and modern therapies remains unclear. FreeDM2 aimed to evaluate the effectiveness of real-time CGM in adults with type 2 diabetes. METHODS: This open-label, parallel-design, randomised controlled trial conducted across 24 primary and secondary care centres in the UK enrolled adults with type 2 diabetes managed with basal insulin and SGLT2 inhibitors or GLP-1 receptor agonists or dual GIP/GLP-1 receptor agonists with HbA1c 7·5-11·0%. Participants were assigned (2:1; using permuted block randomisation by study site, generated by Sealed Envelope) to CGM (intervention) or continuation of self-monitoring of blood glucose (SMBG; control), across two phases: weeks 1-16, self-management with basal insulin self-titration; and weeks 17-32, clinician-supported where additional therapies could be initiated in line with national guidance. Participants and study site staff were not masked to group allocation. The primary outcome was difference between groups in HbA1c concentrations at 16 weeks, and the key secondary outcome was the difference between groups at 32 weeks, both in the treatment policy estimand. Safety analysis included all randomly assigned participants. The FreeDM2 randomised controlled trial is registered at ClinicalTrials.gov (NCT05944432) and is complete. FINDINGS: Between July 26, 2023, and Jan 31, 2025, 469 individuals underwent screening for potential study inclusion, 140 were excluded due to not meeting inclusion criteria, and 329 were included in the baseline phase of the study. 26 individuals were then excluded due to insufficient data capture or withdrawal, and 303 participants were randomly assigned; 198 to the CGM intervention group and 105 to the SMBG control group. 204 (67%) participants were male and 99 (33%) were female, the mean age of the cohort was 60·7 years (SD 9·8), and mean diabetes duration was 16·7 years (6·9). Baseline HbA1c concentration was 8·8% (SD 1·0) in the CGM group and 8·8% (1·1) in the control group, decreasing to 8·0% (0·9) in the CGM group and to 8·7% (1·1) in the control group at week 16 (adjusted difference -0·6 [95% CI -0·8 to -0·3]; p<0·0001) and decreasing further to 7·8% (0·9) in the CGM group and to 8·3% (1·2) in the control group at week 32 (adjusted difference -0·5 [95% CI -0·7 to -0·2]; p<0·0001). There was a similar incidence of non-device-related adverse events in both groups, and two instances of severe hypoglycaemia in the control group. INTERPRETATION: In adults with type 2 diabetes on basal insulin plus modern therapies, real-time CGM improved glycaemic control versus SMBG during self-management and under clinician-supported management. FUNDING: Abbott Diabetes Care.