Effects of a Digital Decision Support System on Diabetes Management in a Real-World Home Care Setting: Prospective Quasi-Experimental Feasibility and Implementation Study.

Background: Older adults with diabetes are vulnerable, facing multimorbidity and challenges in reaching glycemic targets. When insulin therapy is required, limited knowledge of geriatric therapeutic goals often leads to regimens that increase the risk of hypoglycemia. Simplified approaches such as once-daily basal insulin are recommended, yet titration protocols and unclear regimens remain barriers in home care nursing. Digital decision support systems (DDSSs) such as GlucoTab address this gap by offering evidence-based titration guidance integrated into nursing workflows. DDSSs also strengthen nurse autonomy, which leads to improved care processes and better outcomes in home care settings. Objective: This study assessed the effects of a nurse-guided intervention including a DDSS for diabetes management in older adults with type 2 diabetes receiving home care nursing. Methods: The prospective quasi-experimental feasibility and implementation study was conducted using GlucoTab for diabetes management as part of a multicomponent intervention at 2 home care nursing sites. Nine adults with type 2 diabetes receiving insulin therapy were enrolled (n=5 women; mean age 77, SD 10 years; mean BMI 27.7, SD 4.7 kg/m Results: Before implementation, 5 diabetes-related emergency admissions led to 4 hospitalizations, compared with none during implementation and 4 after implementation, mostly due to hyperglycemia. The median hospital stay was 9 (IQR 8-15) days (before and after implementation). Diabetes management shifted from predominantly premixed insulin therapy at baseline to basal insulin with or without bolus insulin during implementation. Diabetes-related home care visits decreased from 1.81 to 1.01 per day (before vs during implementation), and general practitioner contacts for insulin adjustments declined (19 before implementation vs 16 after implementation). Mean morning blood glucose decreased from 212 (SD 72) mg/dL to 162 (SD 42) mg/dL, and the proportion of measurements within the target range (70-180 mg/dL) increased from 39.0% to 70.8% before vs during implementation. No incidence of hypoglycemia requiring third-party help was observed during implementation. Adherence of nurses to DDSS recommendations exceeded 94%. Conclusions: This study suggests that integrating a DDSS in a multicomponent intervention into home care nursing yielded longitudinal changes in diabetes outcomes. Emergency visits and hospitalizations were lower during implementation, no severe hypoglycemia was observed, glycemic control improved throughout the observation period, and therapy was simplified in routine care. Nursing visits decreased, suggesting potential improvements in care processes and resource use. High adherence and acceptance by nurses indicate feasibility and usability in home care practice. These findings align with evidence from hospitalized patients and telehealth studies, suggesting the potential for clinical, safety, and efficiency benefits.