TOPLINE

A clinical trial found that a 1-week, on-site role-play-based training program for physicians managing type 2 diabetes (T2D) improved patient outcomes. Physicians simulated diabetic exams and complications, attended lifestyle workshops, and held face-to-face conversations with patients. This program led to better A1c control at 6 and 12 months in patients with T2D than did standard training.

METHODOLOGY

• Researchers conducted a clinical trial to evaluate the effectiveness of a role-play-based intensive training program for physicians in improving glycemic control and metabolic outcomes among patients with T2D.
• They included 205 physicians (one per center; mean age, 36.2 years; 75.6% women) who were randomly assigned to either a role-play-based intensive training group (n = 103) or a regular training group (n = 102); each physician consecutively enrolled and followed up on approximately 10 patients with T2D.
• Both groups received online training on the diabetes management workflow, the study protocol, and the use of the Metabolic Management Center platform. Physicians in the intensive training group then participated in an additional 1-week, on-site training program in which they role-played as patients, undergoing diabetes-related health examinations, complication scenarios, lifestyle interventions, and face-to-face conversations.
• The trial enrolled 2017 patients with T2D (mean age, 53.0 years; 64.7% men), with a median diabetes duration of 4.0 years, mean BMI of 26.8, and mean A1c of 8.7% at baseline; all had inadequate glycemic control despite using one to two noninsulin hypoglycemic medications for at least 2 months without regimen changes.
• The primary outcome was the proportion of patients achieving A1c levels < 7.0% at 6 months; the secondary outcomes included A1c control at 12 months; changes in clinical, metabolic, and lifestyle parameters; reported adverse events; and improvements in physicians’ professional knowledge at 12 months.

TAKEAWAY

• At 6 months, significantly more patients in the intensive training group achieved an A1c level < 7.0% than those in the regular training group (58.0% vs 42.9%; adjusted between-group difference, 16.6%; P < .001); this difference persisted at 12 months (60.9% vs 44.6%; adjusted between-group difference, 17.0%; P < .001).
• At 12 months, patients in the intensive training group also showed significantly greater reductions in BMI (P < .001), waist circumference (P < .001), fasting blood glucose (P = .006), and systolic blood pressure (P = .04) than those in the regular training group. They also showed healthier lifestyle behaviors, including lower rates of current drinking, higher rates of vegetable and fruit intake (≥ 500 g/d), and a significantly greater increase in the proportion of those with ideal nighttime sleep duration (P = .04).
• Physicians in the intensive training group exhibited greater improvements in diabetes-related knowledge (P < .001) and showed a significantly greater reduction in BMI (P = .01) than those in the regular training group, along with improved dietary behaviors, including a 225 g/wk increase in vegetable and fruit intake (P = .03).
• Serious adverse events were comparable between the two groups.

IN PRACTICE

“[The study] findings suggest that the role-playing-based on-site experiential training effectively bridges the guideline-implementation gap in diabetes care and reaffirms the importance of human-centered offline training in digital healthcare systems for medical education,” the authors of the study wrote.

SOURCE

The study was led by Yifei Zhang, MD, Shanghai Institute of Endocrine and Metabolic Diseases, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China. It was published online on August 6, 2026, in JAMA Network Open.

LIMITATIONS

Each physician enrolled only about 10 patients, limiting the ability to assess variations in intervention effects at the physician level. Patients were enrolled after center-level randomization, raising the potential risk for selection bias. Additionally, the open-label design could have introduced performance bias because participants were aware of their group assignments.

DISCLOSURES

The study was supported by grants from the Noncommunicable Chronic Diseases-National Science and Technology Major Project, the National Key Research and Development Program of China, the Capacity Building for Multidisciplinary Cooperation in Diagnosis and Treatment of Major Metabolic Diseases, and the Leader Project of the Oriental Talent Program. The authors declared having no relevant conflicts of interest.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.