Adoption of AI and its Use Cases in Urban and Rural Hospitals
Nicole M. Summers‐GabrABSTRACT
Purpose
AI can support hospitals’ business workflow as well as their patient care. This study investigates whether rural hospitals' AI adoption is influenced by their proximity to metro areas, which often offer better infrastructure, skilled labor, and other resources, with particular attention to differences across individual clinical and business operation use cases.
Methods
Using the 2024 American Hospital Association's Annual Survey ( n = 6173 hospitals), the likelihood of AI adoption for eight clinical and six business uses by geography was explored. Hospital ownership type and size were controlled. Geography was determined at the county level and used the Rural–Urban Continuum Code (RUCC), with metropolitan hospitals as 1–3, non‐metro or rural as 4–9, and then non‐metro and rural were further divided by metro adjacency.
Findings
Across the country, almost 30% of hospitals adopted AI for clinical tasks and 26% adopted AI for administrative tasks. However, the farther a hospital was from a metro area, the less frequently it adopted AI. Compared to metro hospitals, non‐metro metro‐adjacent hospitals were 29.6% less likely (odds ratio [OR] = 0.70, confidence interval [CI]: 0.53–0.93, p < 0.05) and non‐metro not‐adjacent hospitals were 61.7% (OR = 0.38, CI: 0.26–0.52, p < .001) less likely to use AI for clinical purposes. As for administrative uses, compared to metro hospitals, non‐metro metro‐adjacent hospitals were 48.9% (OR = 0.51, CI: 0.40–0.66, p < 0.001) less likely and non‐metro not‐adjacent hospitals were 70.4% (OR = 0.30, CI: 0.22–0.39, p < 0.001) less likely to use AI. This same pattern persisted for most individual‐use cases.
Conclusions
AI adoption remains limited in hospitals, particularly in non‐metro areas. The impact on adoption rate differences is unclear, but it is an important consideration moving forward to reduce urban–rural health disparities and improve business operations.