Clinical scales currently in use do not accurately identify large-artery occlusion (LAO) in patients with acute ischemic stroke, researchers from France report.

Clinical scales are commonly used to identify stroke patients with LAO so that they can be triaged to the nearest comprehensive stroke center (CSC) for thrombectomy, yet data on the predictive value of these scales are scarce.

Dr. Guillaume Turc and colleagues from Hospital Sainte-Anne and Paris Descartes University investigated the accuracy of 13 published clinical scores in predicting LAO in 1,004 patients with acute ischemic stroke admitted within six hours of symptom onset. Among these patients, 328 (32.7%) had an LAO, according to the April 28 online report in Stroke.

A National Institutes of Health Stroke Score (NIHSS) score of 11 or greater had the highest accuracy (79%), but false-negative rates (FNR) were 16% or higher for all but four of the published cutoffs: out of hospital NIHSS/Cincinnati Prehospital Stroke Severity Scale (CPSS) 1 or greater (4%); abbreviated NIHSS 1 or greater (5%); NIHSS score 4 or greater (7%); NIHSS score 5 or greater (10%).

“We found that although several clinical scores showed a good accuracy to predict LAO, at least 20% of patients with LAO would be missed when applying published cutoffs,” the researchers noted. “Moreover, cutoffs reducing the FNR to 10% (or greater) were associated with an FPR of at least 45%. We were unable to identify a reliable cutoff to rule out LAO.”

“Altogether, our findings call into question the usefulness of a clinical score to identify the best candidates for thrombectomy,” they concluded. “Indeed, even in the best case scenario (all patients in our cohort had a confirmed ischemic stroke and were examined by a stroke neurologist), clinical scales could not reliably identify patients with LAO. Therefore, although clinical scores may provide a rough estimate of the probability of LAO, MRA or CTA should be performed in all patients with symptomatic ischemic stroke presenting within 6 hours of symptom onset.”

Dr. Ashutosh Jadhav from the University of Pittsburgh Medical Center in Pennsylvania told Reuters Health by email, “A major issue in acute stroke care is triaging the right patient to the right hospital. In our recent analysis of the SWIFT PRIME trial, we found that patients with large vessel occlusions who initially presented to a non-endovascular-capable center had greater than 90 minutes’ delay in symptom onset to groin puncture times compared to those who presented directly to the endovascular-capable center. This highlights the importance of being able to identify patients directly in the field that may be better candidates for direct presentation to an endovascular center.”