AI Utilization in Stroke Care: Impacts and Implementation Challenges
A recent study published in the American Journal of Neuroradiology explores the adoption patterns of artificial intelligence (AI) technology for detecting large vessel occlusion in acute ischemic stroke, specifically under Medicare's New Technology Add-On Payment (NTAP) program. Conducted by the Harvey L. Neiman Health Policy Institute in partnership with Northwell Health, the research reveals that AI integration is heavily influenced by the characteristics of healthcare facilities rather than patient-related factors.
Utilizing comprehensive Medicare data, the study assessed 2,116 inpatient acute ischemic stroke episodes in 1,076 facilities from October 2020 to December 2023. During this period, the NTAP code facilitated temporary additional reimbursement for cutting-edge technologies not yet fully covered by established payment systems. The use of AI technologies billed through NTAP peaked at 21% of eligible stroke cases in 2022 but decreased to 14.8% as the availability of the temporary payment code waned in 2023.
AI Utilization Driven by Facility Characteristics
Lead author Casey Pelzl, MPH, emphasized that while AI tools hold potential to enhance the speed and accuracy of stroke identification, their utilization is more contingent on the hospital's capabilities than the clinical needs of patients. This has significant implications for inconsistencies in stroke care.
The study identified no disparities in AI application based on patient demographics, such as age, sex, or race/ethnicity, nor on the severity of strokes. Adoption was predominantly higher in comprehensive stroke centers and facilities within the Stroke Belt, whereas hospitals serving socioeconomically disadvantaged areas showed lower use of NTAP-billed AI, raising access concerns.
Maria X. Sanmartin, PhD, co-author and Assistant Professor at Zucker School of Medicine at Hofstra/Northwell, noted the critical role of operational readiness and existing clinical infrastructures at major stroke centers, stating, "AI tools may enhance faster stroke evaluations, but their implementation hinges on existing systems and workflows."
Elizabeth Rula, PhD, Executive Director of the Neiman Institute, highlighted the importance of reimbursement strategies in propelling AI adoption. However, Rula stressed the need for tailored economic models to accommodate smaller, less-resourced facilities, suggesting options like centralized service models to ensure AI accessibility without necessitating full independent investment.
The study received support from the American Heart Association Grant #24SCEFIA1249718. For further inquiries or interviews with a representative, contact Nichole Gonzalez at [email protected]
The Harvey L. Neiman Health Policy Institute continues to spearhead research on the socioeconomic aspects of medical imaging, examining the evolving role of radiology in healthcare delivery and its broader implications on cost, quality, and efficiency. Additional information can be found at www.neimanhpi.org.