Implementing a Real-time Complex Event Stream Processing System to Help Identify Potential Participants in Clinical...

Area of Science:

• Computer Science
• Health Informatics
• Clinical Research

Background:

• Real-time data analysis is crucial for timely clinical decisions.
• Identifying eligible patients for clinical studies in the Emergency Department (ED) presents a significant challenge.
• Existing alerting systems may lack the efficiency for dynamic patient data streams.

Purpose of the Study:

• To implement Event Stream Processing (ESP) for real-time patient notification in the ED.
• To identify Emergency Department patients potentially eligible for clinical studies.
• To evaluate the performance of the ESP system against a legacy alerting system.

Main Methods:

• Developed and implemented an Event Stream Processing system on the STRIDE platform.
• Focused on inferring event occurrences from data streams without database reference.
• Evaluated the system's performance in real-time notification of potential clinical study participants.

Main Results:

• The implemented ESP system successfully provided real-time notifications for potential clinical study participants.
• Performance was evaluated and found to be adequate when compared to a standalone legacy alerting system.
• The system demonstrated extensibility for more complex research alerting scenarios.

Conclusions:

• Event Stream Processing is a viable computational approach for real-time event inference from data streams.
• The STRIDE platform effectively supports ESP for clinical research alerting in the ED.
• The system has the potential to evolve into an enterprise-level service for complex research needs.