Key points are not available for this paper at this time.
In 2021, US healthcare spending surpassed 4. 3 trillion, about 18. 3% of GDP, and is projected to grow over 5% annually until 2028 1, 2. The US healthcare system, which often fails to provide the best patient outcomes at the lowest cost, is fragmented and inefcient. It is organized around silos of providers, payers, and pharmacies, not centered on patients' needs, leading to duplication, waste, and poor care coordination. Built on a fee-for-service payment model, the system prioritizes service volume over value and quality 3. Despite efforts to link care to value, this model incentivizes overuse, underuse, and misuse of resources. The fragmented care and lack of accessible patient outcome data hinder informed decision-making. The focus remains on billable services rather than disease prevention or health promotion, resulting in higher rates of chronic conditions, complications, mortality, and acute care spending. In searching for solutions to these pervasive healthcare challenges, articial intelligence (AI) offers promising possibilities. AI has tremendous potential to transform clinical care by optimizing resource utilization, enhancing care delivery, and improving patient outcomes. There are three major areas in which we envision AI to be valuable for improving care: (1) streamlining administrative tasks to increase physician productivity, (2) rapidly assisting physicians in diagnosis and treatment, and (3) vastly tracking quality metrics to improve the healthcare system.
Belkin et al. (Mon,) studied this question.