Potential Risk Factors for Hospital-Onset Bloodstream Infections - Initial Findings Within the RISK PRINCIPE Project...

Area of Science:

• Infectious Diseases
• Public Health
• Health Informatics

Background:

• Healthcare-associated infections (HAI) represent a major challenge in patient safety and healthcare costs.
• Current infection prevention and control (IPC) measures lack personalized risk assessment for individual patients.
• Hospital-onset bloodstream infections (HOBSI) are a critical subset of HAI requiring targeted prevention.

Purpose of the Study:

• To identify candidate risk parameters for hospital-onset bloodstream infections (HOBSI) from routine hospital data.
• To lay the groundwork for an application supporting decision-making in personalized IPC.
• To re-analyze existing patient data for novel risk factor identification.

Main Methods:

• Re-analysis of a dataset comprising 4290 patients from a large-scale study.
• Application of logistic regression modeling to identify significant risk factors.
• Utilization of a random forest approach for enhanced predictive parameter identification.

Main Results:

• Initial findings from the re-analysis are presented.
• Candidate risk parameters for HOBSI have been identified using statistical and machine learning models.
• The identified parameters are derived from routinely available hospital data.

Conclusions:

• Routine hospital data contains valuable information for predicting HOBSI risk.
• Personalized risk assessment for IPC is feasible and can be supported by data-driven applications.
• Further development of decision support tools for personalized HOBSI prevention is warranted.