Objective:

This study aims to show improvement in door-to-needle time using artificial intelligence in case of stroke activation, thereby improving stroke metrics in the community hospital.

Background:

Stroke is one of the leading causes of permanent morbidity in the general population. As Time is of the essence, treating stroke promptly to prevent serious complications and treat the reversible area of the infarct in the brain to prevent further spread of irreversible damage. Advancements in AI and technology provide relevant and apt information for making clinical decisions promptly. AI is used to mine the stroke metrics and provide inputs on where the delay in stroke management happens. This will show where we can approach and improve timely management and door-to-needle time.

Design/Methods:

AI automatically extracts timestamps in the EMR. It sends notifications to the responsible physician to manage stroke promptly so that the patient will get the highest quality of care. If there is no response, AI escalates the delay notification to the supervising/senior physician to ensure treatment. Each order’s timestamp is recorded to monitor stroke metrics and compared against average management times, allowing identification of delays and areas for improvement without the need for manual monitoring.

Results:

 We hypothesize that AI mining of data for the Joint Commission stroke metrics and near-realtime notifications for individual cases will show the areas where improvements must be made for effective stroke management.

Conclusions:

AI-assisted data mining can provide accurate areas for improvement. This will reduce the time and cost of manual analysis to find the issue and help us focus more on solving issues and improving the performance and efficiency of stroke management in real time.