Infection Control Through Clinical Pipelines Built with Arden Syntax MLM Building Blocks

Area of Science:

• Medical Informatics
• Infectious Disease Epidemiology
• Clinical Decision Support

Background:

• Healthcare-associated infections (HAIs) pose significant risks, including a 25% 30-day mortality rate for sepsis.
• HAIs incur substantial economic costs, estimated at €7 billion annually for EU member states.
• Existing tools like MONI require clinical and technical updates for effective infection monitoring.

Purpose of the Study:

• To develop and validate a novel, configurable pipeline-based system for defining and processing HAI definitions.
• To enhance clinical decision support for infection monitoring, early warning, and alerts.
• To provide a flexible and open approach for HAI definition management.

Main Methods:

• Developed a configurable pipeline-based system using a network of data access points and clinical rules.
• Implemented rules using Arden Syntax-based medical logic modules within a clinical pipeline framework.
• Validated the system by creating a pipeline for the ECDC PN5 pneumonia rule and testing on intensive care patient data.

Main Results:

• The new system achieved high performance in detecting pneumonia, with a sensitivity of 93.8% and a specificity of 99.8%.
• Results were validated against the established MONI system, demonstrating comparable or improved accuracy.
• The clinical pipeline approach proved effective for processing HAI definitions.

Conclusions:

• Clinical pipelines offer a promising, open, and configurable method for HAI definitions.
• This approach facilitates human-readable and machine-executable HAI definitions.
• The validated system demonstrates potential for improving HAI surveillance and patient safety.