Self-management education and support for type 2 diabetes in Thailand: a cluster randomized trial (2019–2021)

Background Diabetes self-management education and support (DSMES) improves glycemia via improved care coordination and self-management behaviors driven by educational programs tailored to individual needs, but scalable models for its delivery in low- and middle-income countries are lacking. We evaluated the effectiveness of two scalable models of DSMES delivery: two sessions with a nurse (nurse-led) or one each with a nurse and a lay volunteer (peer-assisted). Methods A three-arm cluster randomized trial was conducted comparing two DSMES interventions (nurse-led or peer-assisted) to routine care. The interventions consist of two DSMES sessions (once at baseline and a refresher at 6 months). The nurse-led group was delivered only by nurses, while the peer-assisted group was delivered by nurses with peer assistance. Seven primary care units were allocated to each arm. Newly diagnosed (at enrollment) or high risk glycemia (baseline HbA1c >10%) people with type 2 diabetes were eligible. Primary outcomes were differences in HbA1c and Thai-CV risk score (estimates ten-year risk of cardiovascular events in Thai population based on various clinical parameters) at 12 months relative to routine care. Secondary outcomes were individual cardiovascular risk factors. Findings 693 individuals were enrolled (mean age 59 years, 269 (39%) male, and 424 (61%) female) and 664 completed the study (4% total drop-outs). For the control, nurse-led, and peer-assisted arms respectively, the means (SD) for primary outcomes are as follows. For HbA1c: at baseline; 8.5% (1.4), 8.5% (1.9), and 8.8% (2.0), at 12-month; 7.8% (1.6), 7.7% (1.3), and 8.1% (1.6). For Thai-CV risk scores: at baseline; 22 (15), 23 (15), and 23 (14) at 12-month; 21 (14), 23 (15), and 22 (13). Mean differences at 12-month from control (95% confidence intervals) for nurse-led and peer-assisted arms respectively were: HbA1c; −0.1% (−0.4 to 0.1) and 0.1% (−0.2 to 0.5), Thai-CV risk scores; 0% (−2 to 2) and −0% (−2 to 2). Interpretation While there were no improvements in HbA1c or CV risk scores, low drop-out rates observed demonstrate the feasibility of the models. Future explorations of the appropriate balance between the intensity and feasibility of program delivery are needed. Funding UK Medical Research Council (MR/R020876/1). Thailand Research Fund (DBG6180007).