29 When AI meets QI

Abstract

<h3>Background</h3> The advent of Large Language Models (LLMs) like ChatGPT in late 2022 has revolutionized various fields, including journalism, computer programming, and basic science research. This technological explosion has also significantly impacted the Science of Improvement, presenting both unprecedented opportunities and challenges. <h3>Objectives</h3> This session aims to provide a comprehensive understanding of how LLMs work and their potential applications in Quality Improvement (QI). It also seeks to address the risks and concerns associated with the use of LLMs, such as privacy, accuracy, and efficacy, and how to mitigate them. <h3>Methods</h3> The session involves interactive discussions and presentations on various QI-related use cases of LLMs. It also delves into the technical aspects of LLMs to provide participants with a deeper understanding of their workings. It is based on a 90-day innovation wave, rooted in the Institute for Healthcare Improvement (IHI)’s innovation methodology, completed from July-September 2023. <h3>Results</h3> Artificial Intelligence (AI) has many uses cases for QI, including building data visualizations, surfacing change ideas, and supporting the teaching of QI. In addition, QI methods can be used to introduce AI-powered tools such as predictive analytics clinical approaches. Risks include inaccuracy, privacy violation, exacerbating inequity, and workforce stress due to technological disruption. <h3>Conclusions</h3> The integration of AI with QI presents a promising avenue for enhancing healthcare performance. However, it is crucial to navigate this path with an understanding of both its potential benefits and associated risks. QI practitioners should work to build the knowledge and skills to effectively utilize LLMs in their QI initiatives.

Ähnliche Arbeiten