# LiVE-STROKE: live video from bystander to the medical dispatcher in potential stroke cases

## TL;DR

This study explores using live video from bystanders during emergency calls to help medical dispatchers assess potential stroke cases.

## Contribution

The study evaluates the use of live video in prehospital stroke assessment during emergency calls in Copenhagen.

## Key findings

- Live video was successfully used in 550 stroke cases, but only 52% were correctly examined for facial paresis.
- Only 16% of cases were examined for both facial and extremities paresis, and 46% of stroke cases were missed by dispatchers.
- Missed stroke cases had less accurate examinations and were more likely to involve hemorrhagic stroke.

## Abstract

Identification of stroke during an emergency call remains challenging. We aimed to explore performed stroke examination with prehospital live video transmission during an emergency call from a bystander’s smartphone to the medical dispatcher.

Since 2019, the Emergency Medical Services (EMS) in Copenhagen, consisting of the Emergency Medical Coordination Centre (EMCC) and the Out-of-Hours Health Service (OOHS), have enabled medical dispatchers to add live video to emergency calls. Dispatchers are registered nurses or paramedics. Live video is initiated via text message link sent to the caller’s smartphone. We included all cases where the dispatcher added live video to the emergency call and stroke was suspected or later identified according to the Danish Stroke Registry. The video-recorded on-site stroke examinations were retrospectively analyzed. No specific guidance for video-based stroke examination was in place.

A successful live video transmission was achieved in 550 potential stroke cases, where 52% were correctly examined for facial paresis and 16% for both facial and extremities paresis. In six patients, hemiparesis was observed without an examination. The group with ‘considered stroke’ had the highest amount of examined cases. Dispatchers suspected stroke in 450 cases, and stroke was confirmed in 16% (n = 73). Stroke was diagnosed in 126 cases and was not suspected in 46% of these cases (missed stroke n = 63, TIA n = 10). Patients with missed stroke had significantly less accurate examinations and more often hemorrhagic stroke, compared to patients with confirmed stroke.

Video is not routinely used for stroke assessment during emergency calls, and only a small number of cases are sufficiently examined for stroke with video. Implementation of a structured protocol for live video guided stroke examinations during emergency calls could be beneficial but best practice is still unknown.