EVD-Associated Infections in Subarachnoid Hemorrhage: Risk Factors and Clinical Predictions-A Retrospective...

Area of Science:

• Neurosurgery
• Infectious Disease Epidemiology
• Clinical Risk Prediction

Background:

• External ventricular drain (EVD)-associated infections are a significant complication in subarachnoid hemorrhage (SAH) patients, with incidence rates ranging from 1-45%.
• Current prediction models for EVD infections have limited performance and often overlook dynamic risk factors, focusing instead on static variables at the time of drain insertion.

Purpose of the Study:

• To identify independent predictors of EVD infections in SAH patients.
• To develop a practical clinical prediction model for EVD infections in SAH patients that accounts for changing risk over time.

Main Methods:

• Retrospective analysis of 198 SAH patients with EVDs, collecting 4757 laboratory observations.
• Univariate and multivariate logistic regression analyses were employed to identify independent risk factors and construct a prediction model.

Main Results:

• EVD duration was the sole independent predictor of infection, with a 3.5-fold increased risk for each additional day the drain was in place.
• Infection rates showed a clear dose-response relationship with EVD duration: 11.1% for ≤7 days, 19.3% for 8-14 days, and 31.6% for >14 days.
• The final prediction model demonstrated good discrimination with an AUC of 0.737.

Conclusions:

• EVD duration is the dominant, traditional predictor of infection risk in SAH patients.
• Clinical protocols should prioritize minimizing EVD duration to mitigate infection risk.
• Management strategies should shift focus from complex risk scoring to time-based management of EVDs.