Predictors of false-positive stroke thrombectomy transfers

Area of Science:

• Neurology
• Interventional Radiology
• Emergency Medicine

Background:

• Large vessel occlusion (LVO) stroke patients frequently require transfer for thrombectomy.
• Current transfer decisions often rely on clinical scales due to time constraints, despite high diagnostic error rates.
• This study evaluates the overdiagnosis susceptibility in current LVO stroke transfer protocols.

Purpose of the Study:

• To quantify the rate of non-treatment transfers for suspected LVO stroke.
• To identify the primary reasons for patients not undergoing thrombectomy after transfer.
• To determine predictors associated with unnecessary transfers for thrombectomy.

Main Methods:

• Retrospective review of clinical and transfer data for stroke code activations.
• Analysis of consecutive patients transferred for potential thrombectomy.
• Statistical analysis including chi-squared tests and multivariate logistic regression.

Main Results:

• 54% of transferred patients (105/192) did not undergo thrombectomy between 2015-2016.
• The main reason for non-treatment was the absence of LVO on computed tomography angiography (CTA) post-transfer (68%).
• Lower NIHSS scores (<10) were associated with non-treatment, while helicopter use and delayed arrival (>5 hours) predicted no treatment; early arrival (<5 hours) predicted treatment.

Conclusions:

• Clinical scales appear to overdiagnose LVO, contributing to a majority of non-treatment transfers.
• Primary stroke centers should develop rapid CTA capabilities for LVO assessment before transfer.
• Implementing pre-transfer CTA can significantly reduce costs associated with unnecessary thrombectomy transfers.