人工智能干预临床试验研究方案报告规范指南:SPIRIT-AI扩展版(2020)解读

Guidelines for the Reporting of Clinical Trial Research Protocols for Intervention Involving Artificial Intelligence:Interpretation of SPIRIT-AI Extension(2020)

临床试验研究方案的公开与发表有利于提高临床试验研究设计与方法报告的透明度,提高临床试验研究质量,并最终促使试验结果顺利通过同行评议、得到大量推广与应用。为更好地规范临床试验研究方案,临床试验研究方案报告规范指南[Standard Protocol Items:Recommendations for Interventional Trials,SPIRIT](2013)应运而生。2020年9月8日,涉及人工智能(AI)干预的SPIRIT(SPIRIT-AI)扩展版(2020)正式在线发表。SPIRIT-AI扩展版(2020)是在SPIRIT(2013)的基础上进行的扩增,并经历成立工作组、邀请并获取参与者知情同意、生存初始条目清单、德尔菲调查、共识会议和撰写六个步骤;共对12个条目进行了延伸,对另外3个条目进行了更详细阐述,有助于AI干预试验研究方案的撰写和报告,提高AI干预临床试验方案的透明度和完整性,并最终提高临床干预试验质量。现本研究对SPIRIT-AI扩展版(2020)进行解读,以期为相关研究工作者的实践应用提供借鉴和帮助。

The publicity and publication of clinical trial research protocols is conducive to improving the transparency of clinical trial research design and method reports,which can improve the quality of clinical trial research,and ultimately urg the trial results to successfully pass peer review and be widely promoted and applied.In order to better standardize clinical trial research programs,Standard Protocol Items:Recommendations for Interventional Trials(SPIRIT)(2013)came into being.On September 8,2020,the SPIRIT-AI Extension(2020)involving artificial intelligence(AI)intervention was officially published online.The SPIRIT-AI Extension(2020)is an expansion based on SPIRIT(2013),which went through six steps of setting up working group,inviting and obtaining participants'informed consent,a list of initial entries for survival,a Delphi survey,consensus meetings,and writing.And it extends 12 items,and elaborates on the other 3 items.The SPIRIT-AI extension(2020)will contribute to the writing and reporting of AI intervention trial protocols,improve the transparency and integrity of the AI intervention clinical trial program,and improve the quality of intervention clinical trials ultimately.This article is now interpreting the SPIRIT-AI extended(2020),in order to provide reference and help for the practical application of relevant researchers.