Debi Taylor, MSN, RN, SCRN

Widening the Scope of Stroke Care With AI

I was recently reading an article about a stroke treatment delay that resulted in a significant malpractice verdict against a large certified comprehensive stroke center in the northeast. As a former stroke program leader, this gave me pause as I started to dissect the published facts and how this could potentially happen within many stroke programs globally.

The stroke victim in this case was taken to a nationally certified stroke center by first responders who suspected they were having a stroke. They likely initiated their respective prehospital stroke protocol that included prenotification of the stroke center emergency department team.  Upon arrival at the stroke center, it was reported that the emergency department team initiated their acute stroke protocol and immediately performed head imaging.

All accounts indicate that the stroke process started out using the best practices in favor of expediting care for this patient who had arrived within 40 minutes of their last known well time.

So what happened? Why such a hefty verdict in favor of this stroke patient and their family?

  1. There was no board-certified radiologist onsite to read the stroke imaging study
  2. The initial stroke imaging study read was done by residents who determined there was no clot
  3. Once the attending radiologist read the stroke imaging –  three hours later – a basilar artery occlusion was observed

According to the Radiological Society of North America (RSNA), the radiology field is facing a global shortage. This has been attributed to factors such as an aging population that is increasing the need for imaging, physician burn-out after the COVID pandemic that led to early retirement or seeking new careers and not enough new radiologists entering the field due to a shortage of new resident positions. In this type of environment, the demand is high and the supply is low, which can inevitably lead to patient safety risks.

While it may be shocking to hear that there was no board-certified radiologist onsite to read this patient’s stroke imaging, it’s important to understand that radiologists have been practicing in remote environments for several years – effectively. What needs to be understood is the sheer volume of imaging studies that a radiologist is required to interpret within one shift. One study reported that “…the number of images requiring interpretation each minute of every workday for staff radiologists [is] 16.1.” This radiologist was likely attending to other acute patient images, and this stroke patient’s images were not prioritized appropriately due to an error on the residents’ part.

AI can support providers and clinical care teams as they work to improve stroke treatment timelines and patient outcomes, thus mitigating patient safety risks and avoiding costly verdicts. Let’s break down how AI technology can help improve stroke workflows:

1. Analyze and identify suspected vessel occlusions

There is one healthcare AI company that provides a comprehensive set of AI algorithms along with care coordination solutions to flag suspected positive findings and help cross-specialty physicians and care teams develop timely and relevant stroke patient treatment plans. Aidoc’s full brain solution includes algorithms that identify and flag suspected large and medium vessel anterior, posterior and basilar artery occlusions.

2. Prioritize patient findings and notify care teams

Once the AI flags a suspected large, medium or basilar vessel occlusion, an AI-triggered notification is sent to relevant care team members who can ensure the appropriate and timely patient triage and further evaluate the patients’ imaging.

3. Coordinate care teams and speed up treatment times

AI-triggered notifications can be received by any care team member who plays a relevant role in stroke patient treatment, such as emergency providers, radiologists, neurologists, neurointerventionalists and other clinical care team members. Care team members can immediately engage in data sharing and clinical decision-making that is aimed at improving stroke patient treatment times and clinical outcomes.

As organizations look to mitigate risks due to staff shortages and lack of training, failed prioritization processes or poor team coordination, it is important to consider what role AI can and should play in a stroke workflow. One missed stroke is too many and has proven to be quite costly – for patients, families, providers and healthcare organizations.

Explore the Latest AI Insights, Trends and Research

Debi Taylor, MSN, RN, SCRN