Technological functionality and system architecture of mobile health interventions for diabetes management: a...

Despite advancements in digital health, systematic evaluations of mobile applications (Apps) for diabetes management are limited.

Researchers conducted searches on PUBMED, EMBASE, COCHRANE, SCOPUS, and WEB OF SCIENCE from inception to August 2024. The researchers included randomized controlled trials (RCTs) that investigated the effectiveness of app-based interventions in health management among diabetic patients. Reviewers were paired and independently conducted the screening of studies, data extraction, and evaluation of study quality. The primary outcome of interest was the modification of hemoglobin A1c (HbA1c). The researchers utilized a random effects model to calculate the weighted mean differences (WMDs) and 95% confidence intervals (CIs) and used the I² statistic to assess study heterogeneity. Publication bias for the primary outcomes underwent assessment. Studies were Appraised for quality using the Cochrane Risk of Bias assessment.

41 studies of 3911 initially identified articles that met the selection criteria. The results showed that Apps' intervention significantly improved glycemic control in diabetic patients, with a mean reduction in HbA1c levels of 0.49% (95%CI: -0.65 to -0.32%) compared to standard care. The analysis also revealed that Apps enhanced patient self-management behaviors. Subgroup analyses failed to resolve heterogeneity, but studies consistently observed improved HbA1c levels. The quality assessment results indicated that most studies performed well in the completeness of outcome data and selective reporting.

This meta-analysis confirms that mobile health applications with practical technological functionalities and system architectures are beneficial in managing diabetes. These applications significantly reduced HbA1c levels and improved self-management behaviors. Although some studies exhibited a moderate risk of bias, the overall evidence supports using these applications as valuable tools in diabetes care. Future research should standardize application features, refine system architectures, and address bias issues to enhance.

PROSPERO (CRD42023441365).