A Prospective Multicenter Analysis of Mobile Stroke Unit Cost‐Effectiveness

Objective Given the high disease and cost burden of ischemic stroke, evaluating the clinical efficacy and cost‐effectiveness of new approaches to prevent and treat ischemic stroke is critical. Effective ischemic stroke management depends on timely administration of thrombolytics after stroke onset. This study evaluates the cost‐effectiveness associated with the use of mobile stroke units (MSUs) to expedite tissue plasminogen activator (tPA) administration, as compared with standard management through emergency medical services (EMS). Methods This study is a prospective, multicenter, alternating‐week, cluster‐controlled trial of MSU versus EMS. One‐year and life‐time cost‐effectiveness analyses, using the incremental cost‐effectiveness ratio (ICER) method, were performed from the perspective of CMS's Medicare. Quality‐adjusted life years (QALYs) estimated using patient‐reported EQ‐5D‐5L data were used as the effectiveness measure. Health care utilizations were converted to costs using average national Medicare reimbursements. ICERs excluding patients with pre‐existing disability, and limited to stroke‐related costs were also calculated. Results The first‐year ICER for all tPA‐eligible patients using total cost differences between MSU and EMS groups was $238,873/QALY; for patients without pre‐existing disability was $61,199/QALY. The lifetime ICERs for all tPA‐eligible patients and for those without pre‐existing disability were $94,710 and $31,259/QALY, respectively. All ICERs were lower when restricted to stroke‐related costs and were highly dependent on the number of patients treated per year in an MSU. Interpretation MSUs' cost‐effectiveness is borderline if we consider total first‐year costs and outcomes in all tPA‐eligible patients. MSUs are cost‐effective to highly cost‐effective when calculations are based on patients without pre‐existing disability, patients' lifetime horizon, stroke‐related costs, and more patients treated per year in an MSU. ANN NEUROL 2025;97:209–221