When an emergency medical technician makes a split-second decision about where to transport a stroke patient, the consequences can be profound—especially in rural areas where distance to the right facility may mean the difference between life-altering treatment and missed opportunity.

Grant Brown, associate professor of biostatistics, is part of a team working to transform how those decisions are made, using cutting-edge models, national simulations, and a promising mobile app that could one day guide ambulance routing in real time.

Different Strokes

The project builds on an important insight: strokes vary significantly in their type and required treatment. Large vessel occlusion (LVO) ischemic, non-LVO ischemic, and hemorrhagic strokes—as well as conditions that mimic strokes—all demand different interventions, often with competing time-sensitive windows. Yet in the field, EMS workers can’t distinguish between them definitively without medical imaging.

“Stroke is a somewhat unique condition,” Brown explains. “There are multiple underlying subtypes that we can’t precisely identify in a triage setting. Our goal is to come up with an approach that is more sensitive, is grounded explicitly in a decision model, and takes into account the full context of a particular patient.”

Brown credits his collaborators, Nicholas Mohr and Santiago Ortega-Gutiérrez, both physicians at University of Iowa Health Care, with sparking the idea for the project. “They knew it should be possible to improve on the current state of stroke triage, and we’ve worked together to find a strategy to do that,” he says. “That’s one of the exciting things about biostatistics—we get to work with renowned experts from a variety of fields.”

Developing an App

The research unfolds across three phases. First, the team developed predictive models based on clinical trial data to estimate the likelihood of each stroke type. They then ran nationwide simulations, mapping stroke incidents and modeling driving times to compare triage strategies. The final component is a mobile application designed to suggest an optimal hospital destination tailored to each patient’s situation. Although the app is not yet approved for clinical use, the team is actively testing it in simulated environments with EMS partners.

While the model is national in scope, its relevance is particularly acute in rural states like Iowa. “People are simply farther away from hospitals generally, and from comprehensive stroke centers specifically,” Brown says. “States like Iowa stand to benefit the most from these kinds of decision tools, because the transport decisions are so much more consequential.”

The road from simulation to clinical deployment is long, and Brown acknowledges the hurdles ahead. “We’re currently evaluating what will be required to take the next step,” he says. That includes trial designs, regulatory approvals, funding sources, and identifying sites for future studies.

While there is still much work to be done, Brown and his colleagues remain focused on proving that when it comes to stroke care, smarter decisions at the point of triage could change everything.

This story appeared in the fall 2025 issue of Iowa Public Health Magazine