Development of a prediction model for poor outcomes after thrombolysis in mild non-disabling ischemic stroke

Mild non-disabling ischemic stroke (MNDIS) is increasingly treated with intravenous thrombolysis, yet a substantial proportion of patients still experience poor functional outcomes, and robust tools for individualized risk prediction are lacking.

In this retrospective cohort study, we analyzed 713 consecutive MNDIS patients who received intravenous thrombolysis within 4.5 h of symptom onset at an advanced stroke center between January 1, 2022 and December 31, 2024. Poor outcome was defined as a 90-day modified Rankin Scale (mRS) score ≥2. Candidate predictors, including demographic, clinical, laboratory, hemodynamic and imaging variables, were first screened by univariable analysis and then entered into a stepwise multivariable logistic regression model (entry p < 0.05, removal p > 0.10). A nomogram incorporating independent predictors was constructed in R, and its performance was evaluated using receiver operating characteristic (ROC) analysis, bootstrap calibration, and decision curve analysis.

Of the 713 patients, 91 (12.8%) had poor 90-day outcomes (mRS 2-6) and 622 (87.2%) had good outcomes (mRS 0-1). Admission NIHSS score (OR 1.37; 95% CI 1.10-1.72), 24-h NIHSS score (OR 1.78; 95% CI 1.52-2.10), diastolic blood pressure (OR 1.02 per mmHg; 95% CI 1.00-1.05), and coronary heart disease (OR 1.88; 95% CI 1.05-3.35) were independently associated with poor outcome. The resulting nomogram showed good discrimination (AUC 0.835; 95% CI 0.805-0.861; sensitivity 71.4%; specificity 84.1%), excellent calibration (bootstrap mean absolute error 0.014), and provided positive net clinical benefit across a wide range of risk thresholds (0.03-0.89).

Admission and 24-h NIHSS scores, diastolic blood pressure, and coronary heart disease are key predictors of poor 90-day outcomes after thrombolysis in patients with MNDIS. The derived nomogram offers accurate, well-calibrated, and clinically useful individualized risk estimation, and may assist clinicians in early post-thrombolysis risk stratification and tailoring the intensity of monitoring and follow-up.