Background: In a randomized trial, individual diabetes patient education resulted in better short-term glucose, psychosocial, and behavioral outcomes compared to group education (using conversation maps) and usual care (no assigned education). Long-term analysis is of interest.
Methods: 623 adults with diabetes and glycosylated hemoglobin (A1c) >7% were randomized to individual education (IE), group education (GE), or usual care (UC). Trajectories of A1c values 360 days post-randomization were analyzed using a random effects generalized additive model with smooth splines. Problem Areas in Diabetes (PAID), Diabetes Self-Efficacy (DES-SF), Recommended Food Score (RFS), and Physical Activity (PA) trends were evaluated using survey at baseline and 4 follow-up time points using linear mixed models.
Results: A1c trajectory for IE showed a stepper decline compared to GE and UC in the first 150 days post randomization. However, the A1c reduction was not sustained. By 220 days, there was no treatment group A1c difference. The model fit test (LR-T) for intervention trends was significant for three distinct non-linear trajectories (p=.02). Compared to GE and UC, IE resulted in improved DES and PAID scores in the short-term (DES, p=.008 and PAID, p=.03) and long-term (DES, p=.004 and PAID, p=.008), with no significant long-term differences for RFS or PA.
Conclusions: Higher self-efficacy and lower diabetes distress were observed after an individual educational intervention and sustained over the long-term. However, short-term improvements in glucose control, nutrition, and physical activity were not sustained. Continued educational support may be needed to achieve lasting patient changes in behavior and glucose control.