Evaluation framework for assessing the quality and safety of artificial intelligence solutions for the acute care surgeon: a narrative review from the Eastern Association for the Surgery of Trauma quality, safety, and outcomes committee.

Abstract

Artificial intelligence (AI) has been developing as a field for decades. As acute care surgeons, we have been using AI in the clinical realm and in our everyday lives for years. However, with the public release of conversational large language models, there has been a rapid increase in the uptake of AI, with numerous new products on the market and evolving conversations around what this means for the future of surgery. Given acute care surgeons are leaders in healthcare systems, interfacing with patients and impacting quality and safety through critical timepoints in the emergency department, operating rooms, wards, and intensive care units, it is imperative that we have an understanding of this technology, its potential for bias, its assessment and implementation, and its governance. In this review, we cover these critical concepts and aim to provide a framework for acute care surgeons to understand and evaluate the quality and safety of AI solutions.