What is the accuracy of the iScore (ischemic stroke score) stroke prediction tool?

Accuracy of the iScore Stroke Prediction Tool

The iScore demonstrates excellent accuracy for predicting stroke outcomes, with observed versus predicted outcomes showing correlations of 0.94-0.99 across multiple validation studies, and consistently outperforms expert clinician judgment. 1, 2

Performance Characteristics

Superior to Clinical Judgment

• In the JURaSSiC study comparing 111 stroke experts to the iScore, clinicians accurately predicted death or disability at discharge in only 16.9% of cases, while 90% of iScore-based estimates fell within the 95% confidence interval of observed outcomes 1
• Nearly half (48%) of expert clinicians could not accurately predict outcomes in any of 5 test cases, and none accurately predicted all 5 cases 1
• The iScore's superiority over clinical judgment suggests it should be the preferred tool for outcome prediction and decision-making following acute ischemic stroke 1

Validation Across Multiple Cohorts

• The iScore showed high correlation between observed and predicted outcomes in two large Canadian cohorts: 0.988 and 0.940 for mortality or disability, and 0.985 and 0.993 for mortality or institutionalization 2
• External validation in the VISTA collaboration (7,140 patients) confirmed the iScore's ability to stratify risk and predict outcomes after thrombolysis 3
• A revised version substituting OCSP for TOAST classification maintained comparable discrimination (area under ROC curve 0.767-0.801) with correlation coefficients of 0.985-0.995 4

Clinical Applications and Thresholds

Risk Stratification

• The iScore effectively stratifies patients into low-risk (iScore <200) and high-risk (iScore ≥200) groups, with approximately 10-19% of patients falling into the high-risk category 5, 3
• Patients with iScore ≥200 have dramatically worse outcomes: only 5.5-15.4% achieve favorable outcomes compared to 38.9-58.7% in those with iScore <200 5, 3

Predicting Treatment Response to tPA

• In patients with iScore <200, tPA administration increases odds of favorable outcome by 47% (OR 1.47,95% CI 1.30-1.67) 3
• However, tPA shows no significant benefit in patients with iScore ≥200 (favorable outcome 15.4% vs 13.4% placebo, p=0.77), suggesting limited efficacy in this high-risk group 5, 3
• The iScore predicts hemorrhagic complications: patients with iScore ≥200 receiving tPA had 15.4% symptomatic ICH versus 3.9% with placebo (p=0.04) 5

Multiple Outcome Predictions

• The iScore accurately predicts death or disability at discharge (modified Rankin Scale 3-5), with prevalence ranging from 44-56% depending on the cohort 2
• It also predicts death or institutionalization at discharge (prevalence 16-17%) and 30-day mortality 2
• The tool maintains accuracy for 3-month outcomes, with discrimination improving slightly over time (AUC 0.801-0.810 at 3 months) 4

Important Limitations and Context

Not Designed for LVO Detection

• The evidence provided focuses on the iScore's ability to predict outcomes after stroke, not its ability to diagnose large vessel occlusion (LVO) 6
• For LVO prediction, other tools like NIHSS, RACE, LAMS, and CPSSS are more appropriate, though none achieve both high sensitivity and specificity 6, 7

Risk Stratification After Stroke Remains Primitive

• According to AHA/ASA guidelines, no risk stratification systems have been generally recommended for use after stroke in existing guidelines for secondary prevention 6
• Even "low-risk" groups defined by various stroke prediction instruments have annual stroke recurrence rates of 3.2-4%, well above the 2% threshold for high-risk equivalents 6
• Current risk stratification schemes ignore clinically important outcomes including functional decline, disability, and dementia 6

Practical Implementation

• The iScore is freely available online at www.sorcan.ca/iscore and can be calculated at the bedside 1, 2, 5
• The tool should be used to guide discussions about prognosis, treatment decisions (particularly regarding thrombolysis), and goals of care 1, 3
• An iScore ≥200 should prompt careful consideration of the risk-benefit ratio for aggressive interventions, given the high hemorrhage risk and limited treatment benefit in this population 5, 3