Abstract WP172: A Novel Score to Predict Early Time From Symptom Onset to Hospital Arrival in Patients With Intracerebral Hemorrhage

Abstract

Introduction: Up to 40% of ICH patients have unknown onset time, leading to uncertainty regarding the use of hyperacute ICH treatments and their exclusion from acute ICH treatment trials. We aimed to develop a prediction score that can reliably identify ICH patients presenting within 3 hours of symptom onset. Methods: Spontaneous ICH patients participating in the iDEF trial with available prerequisite data were analyzed. A stepwise multivariable logistic regression model was used to determine predictors of early presentation (symptom onset-to-baseline CT scan time ≤3 hours) with a p-value < 0.15. To develop the prediction score, we allocated scores proportional to the standardized odds ratio (OR) of each binary variable. Continuous variables were dichotomized using the lowest threshold above which there was no additional gain in the positive predictive value (PPV) for early presentation. Results: Of 291 iDEF participants, 132 (45%) presented early. Baseline NIHSS (OR 1.05 per 1-score increase; 95% CI, 0.99-1.11), Glasgow Coma Scale (1.18 per 1-score increase; 1.01-1.39), ICH volume (1.17 per 10-ml increase; 0.98-1.39), absolute perihematomal edema volume (0.81 per 10-ml increase; 0.66-1.00), irregular hematoma shape (2.24; 1.12-4.48), blend sign (0.45; 0.21-0.97), white blood cell count (0.89 per 10 3 cell/ml increase; 0.82-0.96), and blood glucose levels (0.91 per mmol/l increase; 0.82-0.99) were associated with early presentation (c-statistic=0.71). The PPV for an early ICH presentation score (Figure) of >7 was 69%. Conclusion: This novel early ICH presentation score using readily available clinical, neuroimaging, and laboratory data has good performance for identification of ICH patients presenting within 3 hours of symptom onset. If validated in external datasets, the early ICH presentation score can be used to improve recruitment feasibility in acute ICH trials and support clinical decision-making in ICH patients with an unknown symptom onset time.