Are Doctors More Effective When They Use AI? Results From an Experiment in Indonesia, Kenya and the Netherlands

Description

The integration of artificial intelligence (AI), particularly generative tools like chatbots, into clinical reasoning processes offers new opportunities for enhancing diagnostic decision-making and learning. Yet evidence on the perceived utility of such tools is limited. This session will present the findings of a randomized control study that evaluates whether diagnostic performance was improved by the use of a large language model (LLM) in Indonesia, Kenya and the Netherlands. We will discuss the results of the findings with project team members from across the three countries and at the ILO, as well as discuss perceptions about usage and contextual applicability from the medical practitioners who participated in the study.