Patient and Facility Factors Associated with Uptake of NTAP-Billed AI to Identify Suspected LVO in Ischemic Stroke

PURPOSE: Medicare's New Technology Add-On Payment (NTAP) incentivizes the adoption of innovative technologies. We examined factors associated with the use of NTAP-billed artificial intelligence (AI) software for detecting large vessel occlusion (LVO) in acute ischemic stroke (AIS) patients requiring thrombolytic treatment. METHODS: Using nationally representative Medicare 5% Research Identifiable Files, inpatient AIS episodes billed under diagnosis-related groups (DRGs) 61-63 from October 2020 to December 2023 (NTAP period) were analyzed. AI use was identified via NTAP ICD-10 code 4A03X5D. Stroke severity and access to care were assessed using validated scales and the Area Deprivation Index. Associations between NTAP-billed AI use and patient, clinical, and facility characteristics were evaluated using chi-square tests. Odds of NTAP-billed AI use at the AIS inpatient visit level were evaluated with logistic regression models employing random effects to account for facility-level clustering. RESULTS: Among 2,116 AIS episodes across 1,076 facilities, NTAP-billed AI use among inpatient AIS episodes requiring thrombolytic treatment increased annually, peaking at 21% in 2022. NTAP-billed AI use was most common in episodes with CT imaging, at comprehensive stroke centers, and in facilities with ≥1000 beds. No disparities in NTAP-billed AI use were observed across patient demographics or stroke severity measures. In adjusted logistic regression models, odds of NTAP-billed AI use were highest in 2022 (OR 6.0, 95% CI 2.7-13.3), among beneficiaries in the Stroke Belt (OR 2.0, 95% CI 1.3-3.0), and at comprehensive stroke centers (OR 1.5, 95% CI 1.1-2.0). CONCLUSION: NTAP-billed AI adoption for LVO detection in AIS episodes requiring thrombolytic treatment increased from 2020 to 2022 and appeared to be driven by institutional factors, suggesting facility characteristics, rather than patient factors and stroke severity, may influence NTAP-billed AI use, potentially contributing to geographic variability in stroke care.