The AI-enabled regional tele-ECG cloud platform increased diagnostic accuracy at primary healthcare institutions from 82.30% to 98.11% and reduced median clinical decision-making time from 10 to 3 minutes.
Observational (n=1,166)
Yes
Does an AI-enabled regional tele-ECG cloud platform improve diagnostic accessibility and reduce delays in emergency cardiovascular care in primary healthcare settings?
An AI-enabled tele-ECG platform significantly reduced clinical decision-making time and improved diagnostic accuracy for chest pain patients in resource-constrained primary healthcare settings.
Absolute Event Rate: 98.11% vs 82.3%
Background Significant disparities in cardiovascular disease outcomes persist between urban centers and resource-constrained primary healthcare (PHC) settings due to geographic barriers and uneven access to specialist expertise. Digital health networks offer a potential strategy to improve diagnostic accessibility. This study evaluated the real-world implementation of an AI-enabled regional tele-ECG cloud platform within an integrated urban medical group. Methods A longitudinal real-world evaluation was conducted using 1,998 tele-ECG transmissions, including 53 confirmed myocardial infarction (MI) cases at the PHC level within a network of 1,166 MI patients. The platform integrated AI-assisted ECG interpretation with centralized specialist review through a cloud-based B/S architecture. Analyses included inter-tier comparisons, subgroup analyses by geographic location and age (≥65 years), post-hoc power assessment, and a Budget Impact Analysis (BIA) with sensitivity analysis. Results Platform implementation was associated with reduced delays in emergency cardiovascular care. Among patients presenting with chest pain at PHC institutions, median clinical decision-making time decreased from 10 to 3 min (70.0% reduction), while report turnaround time (TAT) was 3.79 ± 1.81 min. No significant differences were observed between PHC institutions and the tertiary hospital in TAT ( P = 0.4384) or report review times ( P = 0.8102). Diagnostic accuracy at the PHC level increased from 82.30% to 98.11%. PHC institutions achieved a survival-to-discharge rate of 75.00% among acute MI cases. The BIA showed an average patient saving of 26 CNY per encounter and an annual net social benefit of 154,182.50 CNY for the regional network. Conclusions The AI-enabled regional collaborative model was associated with improved access to cardiovascular diagnosis across geographically diverse settings. Despite limited statistical power in the PHC subgroup (mean power: 32.4%) and potential confounding from seasonal population migration, the platform may provide a scalable approach for strengthening cardiovascular diagnostic capacity in resource-constrained regions.
Xu et al. (星期三) 进行了一项关于心肌梗死的观察性研究 (n=1,166)。评估了人工智能驱动的区域远程心电图云平台与实施前基线（传统孤立诊断模型）的诊断准确性。在初级卫生保健机构中，人工智能驱动的区域远程心电图云平台将诊断准确性从82.30%提高到98.11%，并将中位临床决策时间从10分钟减少到3分钟。