Development and Validation of a Risk Score for Predicting ICU Admission in Adults with New-Onset Encephalitis

Area of Science:

• Neurology
• Critical Care Medicine
• Epidemiology

Background:

• Encephalitis diagnosis requires timely intensive care unit (ICU) admission for improved patient prognosis.
• Optimal management and resource allocation are crucial for encephalitis patients.

Purpose of the Study:

• To develop and validate a clinical risk score for predicting ICU admission in adult patients diagnosed with encephalitis.
• To assist healthcare providers in making informed decisions regarding the intensity of care upon hospital admission.

Main Methods:

• A binary logistic regression model was used to identify predictive variables for ICU admission in 372 patients with encephalitis (cohort 1).
• The developed risk score was externally validated using a separate cohort of 262 patients with encephalitis (cohort 2).
• International Encephalitis Consortium (IEC) criteria were used for patient selection in both cohorts.

Main Results:

• Of 634 patients with encephalitis, 40% required ICU admission.
• Key predictors for ICU admission included focal neurological signs, new-onset seizures, low Full Outline of Unresponsiveness (FOUR) score (≤14), leukocytosis, and chronic kidney disease.
• The risk score demonstrated good predictive performance with an area under the receiver operating characteristic curve (AUROC) of 0.77 in the primary cohort and 0.76 in the validation cohort.

Conclusions:

• Intensive care unit admission is frequently necessary for patients with encephalitis, irrespective of the underlying cause.
• The developed risk score, incorporating neurological and systemic factors, can support clinical decision-making for encephalitis patient care.
• This tool may enhance the appropriate allocation of critical care resources for encephalitis management.