Predicting Discharge Outcomes from In-Hospital Characteristics After Cerebral Arterial Aneurysm Rupture: A Single...

Area of Science:

• Neurology
• Clinical Prediction Models
• Public Health

Background:

• Existing models for aneurysmal subarachnoid hemorrhage (aSAH) outcomes use a dichotomized approach, failing to capture the full clinical spectrum.
• This limitation leads to underestimation of rehabilitation needs and economic costs.
• The Hospital Assessment Scale (HAS) provides a basis for a more nuanced outcome assessment.

Purpose of the Study:

• To develop and validate an ordinal predictive model for in-hospital outcomes following aSAH.
• To improve the accuracy of outcome prediction beyond dichotomized methods.
• To better inform rehabilitation planning and resource allocation.

Main Methods:

• A retrospective cohort study of 489 aSAH patients in Ukraine.
• Outcome categorization using four HAS scores.
• Ordinal logistic regression with penalized maximum likelihood estimation and multiple imputation for missing data.
• Internal validation via bootstrap resampling (B=200).

Main Results:

• Key predictors of worse HAS outcomes identified: aneurysm re-rupture, larger aneurysm size, limb paresis, conservative treatment, cerebral vasospasm, and hospital-acquired pneumonia.
• A nine-variable ordinal predictive model was constructed, including demographic, clinical, and complication data.
• The model demonstrated strong discrimination and calibration (C-index ≈ 0.82) upon internal validation.

Conclusions:

• An internally validated ordinal predictive model for HAS outcomes after aSAH has been developed.
• The model integrates baseline patient characteristics with in-hospital clinical changes.
• This approach offers a more comprehensive assessment of aSAH patient outcomes.