Effects of ChatGPT Use on Self-care Cardiac Rehabilitation Adherence in Patients with Coronary Artery Disease.

Abstract

Objective: To clarify the possible effects of ChatGPT use on self-care cardiac rehabilitation (CR) adherence in patients with coronary artery disease (CAD).

Method: Patients in the ChatGPT group participated in a standard cardiac rehabilitation program supplemented with personalized digital support via ChatGPT (n=48). The control group participated in conventional cardiac rehabilitation only, without AI-based support (n=48). Both groups completed a 6-month structured rehabilitation program supervised by a multidisciplinary team. Baseline demographic and clinical data, patient satisfaction assessed using a Likert scale, physical activity levels assessed using the International Physical Activity Questionnaire (IPAQ), and metabolic equivalent of task (MET) values were recorded.

Results: The postrehabilitation MET score was 1537 in the ChatGPT group and 1413 in the control group (P = 0.001). Similarly, the change in MET score after rehabilitation was greater in the ChatGPT group than in the control group (1029 vs. 875, P = 0.001). SF-36 parameters, including physical functioning, role-physical, mental health, bodily pain, and general health, were significantly higher in the ChatGPT group than in the control group (P < 0.05 for all). Program compliance was higher in the ChatGPT group than in the control group (91.7% vs. 75.0%, P = 0.028). Patient satisfaction was also higher in the ChatGPT group (4.5 vs. 3.8, P = 0.001).

Conclusion: ChatGPT use during CR was associated with higher patient compliance rates. Moreover, ChatGPT-supported CR was associated with higher physical activity levels, greater patient satisfaction, and better selected quality-of-life scores.