Effectiveness of an AI-Based Secondary Prevention Management Model for Coronary Heart Disease on the Prognosis of Patients After Coronary Intervention For Acute Myocardial Infarction: A Randomized Controlled Trial (Preprint)

BACKGROUND The suboptimal implementation and limited accessibility of secondary prevention measures have led to adverse outcomes for myocardial infarction, creating a significant burden on healthcare systems. OBJECTIVE Our study evaluates the efficacy of an artificial intelligence (AI)-based secondary prevention management model for coronary heart disease in improving the long-term prognosis of myocardial infarction patients following percutaneous coronary intervention. METHODS In this single-blind, parallel, randomized controlled trial, we enrolled myocardial infarction patients aged 18 and older who underwent percutaneous coronary artery intervention at a major tertiary hospital in Anhui, China. Participants were randomly assigned (1:1) to either an AI-managed group or a conventional management group. The AI-managed group received secondary prevention management through an AI-based platform, while the conventional group received standard cardiology outpatient follow-ups without formal secondary prevention. Evaluations were conducted at baseline and at 12 months. The primary endpoint was the one-year incidence of major adverse cardio-cerebral events (MACCE), encompassing all-cause mortality, recurrent myocardial infarction, target vessel revascularization, and stroke. Secondary endpoints included the incidence of bleeding events, readmissions due to heart failure/angina, lifestyle modifications, and medication adherence. This study is registered with the China Clinical Trial Registry, registration number Chi-CTR-2200065344. RESULTS A total of 1,328 patients who met the entry criteria were enrolled in the trial between October 2021 and October 2022, of whom 1,104 (95.3%) patients completed the 1-year follow-up, and 44 (7.6%) and 8 (1.4%) patients were lost to follow-up in the intervention and usual Management groups, respectively. There were no significant differences in gender, age, medical history, laboratory tests, ejection fraction, and discharge medications among those who completed the follow-up. A total of 111 MACCEs events occurred in both groups at 12 months, with 43 MACCEs (9.1%) occurring in the intervention group compared with 68 MACCEs (11.9%) in the conventionally managed group (P = 0.031). Statistically significant differences were observed between the intervention and standard care groups in cardiac mortality (1.9% vs. 4.0%), postoperative myocardial infarction (1.3% vs. 4.2%, P = 0.03), readmissions for recurrent heart failure/angina (11.6% vs. 17.7%, P = 0.004), and total bleeding events (13.7% vs. 18.4%, P = 0.032). However, there were no significant differences in stroke (2.6% vs. 3.0%, P = 0.550), target vessel revascularization (2.2% vs. 3.6%, P = 0.204), and in-stent thrombosis (0.6% vs. 1.2%, P = 0.233). CONCLUSIONS This study demonstrates that the use of an AI-based secondary prevention management model for Myocardial Infarction significantly improves the prognosis of myocardial infarction patients and optimizes the allocation of health resources, which is beneficial for the modern healthcare system. CLINICALTRIAL This study is registered with the China Clinical Trial Registry, registration number Chi-CTR-2200065344.https://www.chictr.org.cn/