Abstract
Introduction
Brain MRI with diffusion-weighted imaging provides sensitive detection of acute cerebral ischemia in patients with suspected transient ischemic attack (TIA), while MRA permits evaluation of the intracranial and cervical vasculature. Despite these advantages, CTA head and neck is frequently obtained in patients with resolved neurologic symptoms who subsequently undergo MRI, potentially resulting in duplicative imaging and unnecessary exposure to ionizing radiation and iodinated contrast. We evaluated the effect of implementing a structured MRI/MRA-first pathway for suspected TIA on institutional neurovascular imaging utilization.
Methods
In August 2025, an MRI/MRA-first imaging pathway was implemented at a single hospital for non-stroke-alert patients undergoing evaluation for suspected TIA in the Observation Unit. The pathway utilized brain MRI with diffusion-weighted imaging, MRA of the head, and MRA of the neck, collectively termed the “triple rule-out” protocol. CTA head/neck remained available when MRI could not be completed within the targeted 12-hour timeframe. Monthly MRI/MRA triple-rule-out and CTA head/neck examination volumes were retrospectively reviewed for the 12 months before implementation (August 2024–July 2025) and 12 months after implementation (August 2025–July 2026). Because available CTA data included examinations performed for all clinical indications, CTA volume was evaluated as an institutional utilization measure rather than as a TIA-specific cohort. Differences in the relative utilization of MRI/MRA versus CTA head/neck between periods were assessed using the chi-square test.
Results
MRI/MRA triple-rule-out volume increased from 760 examinations during the preimplementation period to 1,483 examinations after implementation, representing a 95.1% increase. Mean monthly MRI/MRA volume increased from 63.3 to 123.6 examinations. Monthly volume ranged from 39–82 before implementation and 93–171 afterward, with every postimplementation month exceeding the highest monthly volume observed before implementation. MRI/MRA accounted for 50.3% of combined MRI/MRA and CTA head/neck examinations before implementation compared with 67.3% afterward, an increase of 17.0 percentage points (p 0.001). Overall CTA head/neck volume remained relatively stable, decreasing from 752 to 720 examinations (4.3%).
Conclusion
Implementation of a structured MRI/MRA-first pathway for suspected TIA was associated with a nearly twofold and sustained increase in utilization of the dedicated MRI/MRA triple-rule-out protocol and a significant shift in institutional neurovascular imaging utilization toward MRI/MRA. The persistence of increased MRI/MRA utilization throughout the 12-month postimplementation period suggests sustained adoption of the pathway. Because CTA head/neck data included examinations performed for all clinical indications, the effect of the intervention on CTA utilization specifically among patients with suspected TIA cannot be determined from the available data. Patient-level analysis would be required to assess changes in TIA-specific CTA utilization and duplicative imaging.