Scholarlink

<p>Artificial intelligence (AI) has the potential to improve healthcare delivery, but uneven adoption and implementation can reinforce existing care gaps and inefficiencies. We analysed data from 3,560 US hospitals using the 2023 American Hospital Association (AHA) Annual Survey, the 2023–2024 AHA Information Technology Supplement, community-level socioeconomic indicators, and the 2023–2025 Centers for Medicare &amp; Medicaid Services hospital quality metrics to examine where AI is implemented, what factors are associated with implementation and the patterns of early AI adoption across geographical regions. Here we found that hospital AI implementation is considerably clustered, with hotspots and coldspots of adoption. Regions with greater healthcare access needs were less likely to have hospitals with AI-based predictive models. Geographically weighted regression showed that factors associated with predictive AI implementation vary by region, suggesting that adoption patterns reflect diverse local contexts and institutional characteristics. These findings provide a baseline snapshot of early AI deployment patterns in US hospitals in 2023 and 2024, highlighting the uneven and context-dependent nature of implementation. Future efforts should develop standardized, detailed, model-specific AI implementation metrics and account for local contexts rather than pursuing uniform deployment strategies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The landscape of AI implementation in US hospitals

  • 期刊论文

  • 发表日期：

    2026年1月15日

查看全文
  • Yeon-Mi Hwang,
  • Madelena Y. Ng,
  • Malvika Pillai,
  • Michelle P. Sahai,
  • Tina Hernandez-Boussard

摘要

Artificial intelligence (AI) has the potential to improve healthcare delivery, but uneven adoption and implementation can reinforce existing care gaps and inefficiencies. We analysed data from 3,560 US hospitals using the 2023 American Hospital Association (AHA) Annual Survey, the 2023–2024 AHA Information Technology Supplement, community-level socioeconomic indicators, and the 2023–2025 Centers for Medicare & Medicaid Services hospital quality metrics to examine where AI is implemented, what factors are associated with implementation and the patterns of early AI adoption across geographical regions. Here we found that hospital AI implementation is considerably clustered, with hotspots and coldspots of adoption. Regions with greater healthcare access needs were less likely to have hospitals with AI-based predictive models. Geographically weighted regression showed that factors associated with predictive AI implementation vary by region, suggesting that adoption patterns reflect diverse local contexts and institutional characteristics. These findings provide a baseline snapshot of early AI deployment patterns in US hospitals in 2023 and 2024, highlighting the uneven and context-dependent nature of implementation. Future efforts should develop standardized, detailed, model-specific AI implementation metrics and account for local contexts rather than pursuing uniform deployment strategies.