Knowledge-based surveillance for preventing postoperative surgical site infection

Area of Science:

• Medical Informatics
• Public Health
• Computer Science

Background:

• Hospital-acquired infections (HAIs) affect at least 5% of Canadian hospital admissions, leading to significant mortality and healthcare costs.
• Current HAI surveillance relies on manual data abstraction from multiple sources, limiting efficiency and effectiveness.
• Surgical Site Infections (SSIs) represent a substantial portion of all HAIs.

Purpose of the Study:

• To apply knowledge modeling and semantic technologies to automate HAI surveillance.
• To facilitate the integration of disparate data sources for enhanced HAI detection.
• To enable automatic reasoning for identifying clinical events of interest, specifically SSIs.

Main Methods:

• Utilized knowledge modeling to structure clinical data.
• Employed semantic technologies for data integration from diverse healthcare information systems.
• Developed automated reasoning capabilities to identify potential SSIs.

Main Results:

• Demonstrated the feasibility of integrating heterogeneous data using semantic technologies.
• Showcased the potential for automated reasoning to detect clinical events like SSIs.
• Highlighted the limitations of current manual surveillance methods.

Conclusions:

• Knowledge modeling and semantic technologies offer a promising approach to automate and improve HAI surveillance.
• Automated surveillance can enhance the identification of SSIs, leading to better prevention and control.
• This research paves the way for more efficient and effective infection control practices in hospitals.