Healthcare professionals across the United States are increasingly overworked, and physician shortages are projected to worsen in the coming decade. In regions such as the State of New York, emergency department (ED) utilization for non-emergent conditions remains disproportionately high. Our previous study identified a strong association between public insurance type and avoidable ED use, demonstrating persistent gaps in healthcare access and navigation. Patients with Medicare or Medicaid often experience limited primary care availability, appointment delays, and complex referral systems, leading to greater reliance on ED services for non-emergent needs. This paper explores strategies to reduce avoidable ED visits and improve healthcare efficiency using artificial intelligence (AI). Integrated AI-based tools are promising for identifying high-risk patients for non-emergent ED use and directing them to preventive and primary care resources. By linking these models with others that support healthcare systems in allocating resources, streamlining patient flow, and improving clinical decision-making, we can help reduce ED congestion. Future initiatives should prioritize integrating AI models into care settings and hospitals, and expanding this nationwide to improve patient well-being.
Georgy et al. (Sat,) studied this question.