Time to start using checklists for reporting artificial intelligence in health care and biomedical research: a rapid review of available tools

Abstract

While the volume of using artificial intelligence (AI) and machine learning (ML) in medical research has grown considerable over the past years, the reporting quality for the majority of such studies has been poor, raising concerns about the replicability, biasedness, validity and overall value for a vast amount of research. This rapid review aims to summarize reporting guidelines for medical AI studies. Following a systematic search in the PubMed database up to May 2022 and the reference lists of previously published reviews in the field, we identified 22 reporting checklists published or under development for a variety of study designs and clinical fields or general use. The main aims, the target audience and specific focus of the identified checklists has been summarized. Given the documented positive impact of checklists on the reporting quality of medical research, we encourage researchers using AI or ML in medicine to start using them.