Improvements in Glycemic Control With a Digital Diabetes Logbook: Secondary Analysis of a Randomized Controlled Trial Enriched by Observational, Real-World Data

Abstract

Background: The treatment of diabetes requires substantial self-management. Digital tools can help reduce the burden of self-management and may improve glycemic control. Objective: This study aims to determine whether the use of a digital diabetes logbook increased the likelihood of achieving optimal glycemic control (glycated hemoglobin [HbA1c] ≤6.5%) after 3 months, based on a secondary analysis of randomized controlled trial (RCT) data. A secondary objective was to evaluate the long-term impact of the logbook on mean blood glucose levels over 3 and 12 months using observational, real-world data (RWD). Methods: Data from 342 participants with type 1 or type 2 diabetes enrolled in the mySugr PRO-RCT were analyzed. A robust logistic regression was performed to examine the likelihood of achieving optimal glycemic control, defined as an HbA1c value ≤6.5% at the 3-month follow-up. The dependent variable was the dichotomous outcome indicating whether this threshold was met. The primary independent variable was group allocation, with baseline HbA1c included as a covariate. For the analysis of RWD, a total of 2861 participants with type 1 or type 2 diabetes were identified using propensity score matching to align their characteristics with those of the RCT participants closely. One-sample t tests were conducted to analyze changes in mean blood glucose separately for each diabetes type, from baseline to 3 months of app use, and from baseline to 12 months of app use (in a subcohort of 1176 participants). Results: The RCT data showed that the likelihood of achieving optimal glycemic control was nearly doubled in the intervention group compared with the control group (odds ratio 2.24, 95% CI 1.12-4.47; P=.02). RWD indicated that mean blood glucose levels significantly improved over 3 months of app use in both groups (type 1: –16.3 mg/dL; 95% CI –20.6 to –12.4; P 8.5% at baseline (before app use) showed the greatest reductions in mean blood glucose (type 1: –82.2 mg/dL; 95% CI –102.0 to –61.8; P