SELECT-CT: Symptom Evaluation, Lesion Estimation, and Curated Triage to CT - A Simple Algorithm to Select the Right...

Area of Science:

• Neurology
• Radiology
• Medical Imaging

Background:

• Multimodal "Code Stroke" CT scans are crucial for identifying acute ischemic stroke patients eligible for mechanical thrombectomy.
• Current protocols lead to a high volume of scans, potentially overwhelming radiology services.
• There is a need for efficient algorithms to streamline the identification of thrombectomy candidates.

Purpose of the Study:

• To evaluate the SELECT-CT algorithm for identifying mechanical thrombectomy candidates.
• To assess the algorithm's ability to reduce the number of Code Stroke CT scans.
• To compare SELECT-CT performance against various National Institutes of Stroke Scale score thresholds.

Main Methods:

• Retrospective cohort study of patients undergoing mechanical thrombectomy over two years.
• Application of the SELECT-CT algorithm (NIHSS ≥2 and/or posterior circulation/disabling symptoms).
• Assessment of SELECT-CT and NIHSS thresholds for scan reduction and accurate identification of thrombectomy candidates.

Main Results:

• The SELECT-CT algorithm correctly identified 97% of thrombectomy patients, including all indicated by current evidence.
• Compared to NIHSS ≥6 (83%) and NIHSS ≥2 (95%), SELECT-CT improved identification accuracy.
• SELECT-CT achieved a 30% reduction in Code Stroke CT scan volume, outperforming NIHSS thresholds alone.

Conclusions:

• The SELECT-CT algorithm successfully reduces Code Stroke CT workload.
• It reliably identifies patients who require mechanical thrombectomy.
• This algorithm offers a more efficient approach to managing acute ischemic stroke patients.