# Treatment‐Specific Risk Scales for Identifying High‐Risk Patients With Poor Prognosis in Acute Ischemic Stroke: A Cohort Study From the National Neurological Medical Center of China

## TL;DR

This study developed a new risk scale called PAIST to predict poor outcomes in acute ischemic stroke patients, helping doctors make better treatment decisions.

## Contribution

The novel contribution is the development of treatment-specific risk models (PAIST) for acute ischemic stroke prognosis.

## Key findings

- PAIST outperformed benchmark models in predicting adverse outcomes in both thrombolysis and non-thrombolysis groups.
- High-risk patients identified by PAIST had significantly higher poor prognosis rates.
- The model integrates multidimensional clinical variables and is externally validated.

## Abstract

To develop and validate a user‐friendly scale for predicting acute‐phase adverse outcomes in acute ischemic stroke (AIS), thereby optimizing clinical management.

This retrospective study enrolled AIS patients within 72 h of onset (excluding thrombectomy), stratified according to thrombolysis status to develop treatment‐specific prognostic models. The prognostic scale of AIS acute stage based on treatment stratification (PAIST) was developed using clinical variables, with discharge mRS as the primary endpoint, followed by external validation.

A total of 1971 AIS patients (437 thrombolyzed) were included. Both thrombolysis‐specific and non‐thrombolysis‐specific models incorporated core predictors (baseline NIHSS, deep vein thrombosis, neuron specific enolase, neutrophil percentage) but differed in cut‐off values and weightings. Additionally, the non‐thrombolysis‐specific model integrated three extra variables: age, fasting blood glucose, and serum potassium. External validation demonstrated PAIST outperformed the benchmark model (AUCs: thrombolysis group 0.759 vs. 0.698; non‐thrombolysis group 0.850 vs. 0.801; all p ≤ 0.05). PAIST‐based risk stratification effectively identified high‐risk patients, with poor prognosis rates of 76.92% (thrombolysis group) and 61.11% (non‐thrombolysis group).

The PAIST scale is an effective and practical tool for acute‐phase prognostic risk stratification in AIS. Its treatment‐stratified design enables accurate risk assessment, thereby supporting individualized clinical decision‐making.

The PAIST prognostic tool, developed through thrombolysis‐based stratification and externally validated, effectively identifies AIS patients at high risk of acute‐phase adverse outcomes by integrating multidimensional variables, demonstrating superior predictive performance to support clinical management.