This three-arm pragmatic randomized controlled trial enrolled 666 African-American adults with uncontrolled diabetes (HbA1c ≥ 8%) and multiple chronic conditions from 19 Mid-South US primary care practices. Participants were randomized to educational materials (EM) alone, health coaching plus EM (HC+EM), or text messaging plus EM (TM+EM). The primary outcome was patient-reported healthy eating, exercise, and medication adherence self-care behaviors; secondary outcomes included HbA1c, body weight, quality of life, and primary care engagement.
At 12 months, healthy eating days increased in all arms: EM-alone by 0.67 ± 0.19 day/week (95% CI, 0.30-1.04), HC+EM by 0.99 ± 0.15 day/week (95% CI, 0.70-1.28), and TM+EM by 1.36 ± 0.15 day/week (95% CI, 1.07-1.64). The TM+EM arm showed significantly greater improvement than EM-alone (0.69 ± 0.24 day/week). Exercise behaviors were similar across arms. All three arms significantly improved HbA1c from baseline to 12 months, with an average reduction of 0.76% ± 0.12% (p ≤ 0.0001); among 349 participants with complete data, 90 (25.8%) achieved HbA1c below 8%.
Safety and tolerability were not reported. Limitations include little known about comparative effectiveness beyond these outcomes. The randomized design supports causal inference for intervention effects on outcomes, but generalizability is limited to African-American adults with uncontrolled diabetes and multiple chronic conditions.
These low-cost interventions should be made more readily available in primary care settings nationwide, though exercise behaviors did not differ between arms.
View Original Abstract ↓
BACKGROUND: Although studies have demonstrated that many patient-centered behavioral interventions are effective in improving diabetes self-care, little is known regarding their comparative effectiveness.
OBJECTIVE: The Management of Diabetes in Everyday Life (MODEL) Study sought to compare the effectiveness of promising alternative patient-centered behavioral interventions, including educational materials (EM) alone, health coaching (HC)+EM, and text messaging (TM)+EM, for improving diabetes self-care.
DESIGN: Three-arm pragmatic randomized controlled trial (ClinicalTrials.gov NCT02957513).
PARTICIPANTS: A total of 666 African-American adults with uncontrolled diabetes (hemoglobin A1c [HbA1c] ≥ 8%) and multiple chronic conditions from 19 Mid-South US primary care practices randomly assigned to EM, HC+EM, or TM+EM in a 1:2:2 ratio.
INTERVENTIONS: EM-alone participants received culturally tailored and patient-vetted diabetes EM. HC+EM participants received motivational interviewing-based coaching from certified lay health coaches. TM+EM participants received personalized and tailored messages.
MAIN MEASURES: Primary outcomes included patient-reported healthy eating, exercise, and medication adherence self-care behaviors. Secondary outcomes included HbA1c, body weight, quality of life, and primary care engagement.
KEY RESULTS: All three arms significantly improved healthy eating, exercise, and HbA1c from baseline to 12 months. EM participants increased healthy eating days by 0.67 ± 0.19 day/week (95% CI, 0.30-1.04), HC+EM by 0.99 ± 0.15 day/week (95% CI, 0.70-1.28), and TM+EM by 1.36 ± 0.15 day/week (95% CI, 1.07-1.64). Improvements in exercise behaviors were similar in all arms. TM+EM participants experienced significantly greater improvement in healthy eating than EM-alone participants (0.69 ± 0.24 day/week). Average reduction in HbA1c was 0.76% ± 0.12% (p ≤ 0.0001) and 90 of 349 participants (25.8%) with complete data for HbA1c achieved reductions below 8%.
CONCLUSION: The MODEL Study demonstrates that EM-alone, HC+EM, and TM+EM are effective for improving self-care behaviors and HbA1c. TM+EM is more effective than EM-alone in improving healthy eating. Low-cost EM, HC+EM, and TM+EM should be made more readily available in primary care settings nationwide.