A Telemedicine System Intervention for Patients With Type 1 Diabetes: Pilot Feasibility Crossover Intervention Study

**Background:** Diabetes mellitus is a growing global health problem, with an estimated 8.4 million people worldwide living with type 1 diabetes in 2021. Telemedicine systems have the potential to improve diabetes self-management by enabling real-time data collection, analysis, and feedback, but their adoption remains limited. The Diani telemedicine system integrates a web app, mobile diabetes diary (Diabetesdagboka), Bluetooth glucometer, Fitbit smart bracelet, and optional continuous glucose monitor and smartwatch to support patients in managing their diabetes.

**Methods:** This was a 24-week randomized crossover feasibility study conducted at a single internal medicine clinic in the Czech Republic. Ten patients with type 1 diabetes (5 men, 5 women; mean age 47.7 years, SD 19.3; mean diabetes duration 10.5 years, SD 8.6; mean HbA1c 59.5 mmol/mol, SD 6.7) were enrolled. Patients were randomized to group A (Diani system for first 12 weeks, then usual care) or group B (usual care first, then Diani system for 12 weeks). Data were collected at three visits (baseline, week 12, week 24). Primary outcomes were HbA1c and body weight. Secondary outcomes included quality of life (WHOQOL-BREF questionnaire), patient-reported system evaluation, and system usage analytics via Matomo. Statistical analysis used linear mixed effects (LME) models, Wilcoxon tests, and Friedman tests.

**Key Results:** After the intervention, mean HbA1c decreased by 4.35 mmol/mol (P=.01 by LME analysis). The effect comparing V2 with V1 was −6.54 mmol/mol (P=.04); the effect at V3 versus V1 was −2.16 mmol/mol (P=.45). Mean body weight did not change significantly (81.4 kg at V1, 81.1 kg after intervention, 81.5 kg at V3; P>.05). The proportion of patients measuring blood glucose at least 3 times daily increased from 70% at baseline to 100% after intervention; those measuring ≥4 times daily increased from 40% to 60%. Using LME, patients measured blood glucose on average 1.8 times more often during system use (P=.06). Quality of life showed a trend toward improvement: 60% rated quality of life as good/very good at V1, rising to 78% at V2 and 89% at V3, but Friedman and Wilcoxon tests did not show statistically significant differences. Patients spent a mean of 18.6 (SD 6.8) minutes daily using the app. The median number of days of system use was 84. Seven (70%) patients reported feeling safer using the system, and 90% reported that using the system helped them think more about their diabetes and dose insulin better. Positive feedback included graphical visualization, automatic glucometer data transmission, and device interoperability. Negative feedback included difficulty using the smartwatch, desire for trend graphs, and occasional server connectivity issues.

**Clinical Implications:** This small feasibility study suggests that the Diani telemedicine system may improve glycemic control in type 1 diabetes, as evidenced by a statistically significant reduction in HbA1c of 4.35 mmol/mol. The system also appeared to increase blood glucose monitoring frequency and patient engagement with diabetes self-management. Patients reported improved understanding of the relationships between carbohydrate intake, physical activity, and insulin dosing. However, the study had significant limitations: a very small sample (10 of 30 planned patients were recruited), 80% questionnaire return rate, technical issues during the first phase, and lack of statistically significant effects on quality of life or body weight. The findings support the potential value of interactive telemedicine systems for type 1 diabetes but underscore the need for larger, adequately powered studies to confirm these effects and identify factors that promote long-term patient engagement.