Telemedicine for Type 2 Diabetes: Meta-analysis Shows HbA1c and Cardiometabolic Benefits

A systematic review and meta-analysis of randomized trials in adults with type 2 diabetes compared telemedicine interventions with usual care and reported a pooled reduction in HbA1c favoring telemedicine (mean difference [MD] –0.38% HbA1c, 95% CI –0.49 to –0.27), alongside substantial heterogeneity (I²=96%). The authors framed this glycemic outcome as the primary endpoint and also examined cardiometabolic secondary outcomes, including whether study or intervention features appeared to modify effects. Overall, the report emphasizes a pooled improvement in glycemic control versus usual care while describing variability across studies and related physiologic measures.

The review included 58 randomized controlled trials encompassing 13,942 participants, defined as adult type 2 diabetes populations in which telemedicine was compared against face-to-face, standard, or traditional care. For the quantitative synthesis of HbA1c, the authors note that 47 trials provided sufficient data for meta-analysis, and that some trials contributed more than one intervention group, yielding multiple comparisons. Outcomes were pooled as mean differences, using random-effects models when heterogeneity was high; prespecified outcomes included glycemic measures and cardiovascular risk–related indicators reported across trials. In the authors’ presentation, the HbA1c analysis is the main reference point for the overall comparison between telemedicine and usual care.

Secondary pooled analyses favored telemedicine over usual care for several cardiometabolic measures, with reported mean differences of –11.29 mg/dL for fasting blood glucose, –1.33 kg for weight, –0.43 kg/m² for BMI, –2.14 mmHg for systolic blood pressure, and –1.24 mmHg for diastolic blood pressure (95% CI –2.02 to –0.46). The authors reported no statistically significant between-group differences for HDL cholesterol or LDL cholesterol. They describe these results as between-group, trial-level averages that extend the synthesis beyond glycemic control into weight and blood pressure domains. Overall, the secondary outcomes are presented as modest pooled shifts across multiple measures within a heterogeneous evidence base.

In HbA1c subgroup analyses, the authors report larger reductions in categories where telemedicine was delivered by physicians, dietitians, or researchers. They also describe significant improvements for short-term and long-term interventions, while medium-term interventions did not reach statistical significance. In meta-regression, however, they report that no statistically significant moderators were identified, including location, setting, telemedicine type, provider, and intervention duration. Interventions spanned synchronous, asynchronous, remote monitoring, and hybrid modes, with commonly used components including monitoring, counseling, and education, and tools such as telephone calls, remote monitoring devices, apps, and text messages. The paper presents subgroup patterns alongside an overall meta-regression finding that statistically supported effect modifiers were not detected.

The authors caution that substantial between-study heterogeneity was a persistent feature of the pooled analyses. They describe publication bias assessments suggesting possible asymmetry, including a borderline Egger test result (P=.054) and a trim-and-fill analysis that imputed potentially missing studies and produced an adjusted effect estimate. They also note interpretive limits tied to variability in intervention content, modality, providers, and duration, along with study-conduct constraints such as lack of blinding in telemedicine trials and the exclusion of non-English studies as a potential source of bias. In describing the evidence base, they report that a large share of included studies originated from high-income countries, which they discuss as part of the broader context for how the findings may be viewed. In the authors’ overall interpretation, pooled benefits were observed versus usual care, while study-to-study variability shaped how confidently summary estimates could be generalized.

Key Takeaways:

• The authors reported that telemedicine (vs usual care) was associated with a pooled reduction in HbA1c in adults with type 2 diabetes.
• The pooled analyses also reported improvements in fasting glucose, weight, BMI, and blood pressure outcomes with telemedicine, while HDL-C and LDL-C outcomes were not statistically significant.
• The authors reported high heterogeneity and described subgroup patterns by provider and intervention duration, but meta-regression did not identify statistically significant moderators.