CONSORT-AI

An extension of the CONSORT checklist that tells you what to report when a randomised trial tests an intervention with an AI component.

What it is¶

CONSORT-AI adds 14 AI-specific items to the standard CONSORT checklist for reporting randomised controlled trials. The new items ask you to describe the AI intervention and the skills needed to use it, the setting it was built into, how its inputs and outputs were handled, how people interacted with it, and what you learned from analysing its errors. It was developed through an international consensus process together with SPIRIT-AI and published in 2020 in several journals at once, including Nature Medicine and The BMJ.

In healthcare¶

Written for clinical trials of AI interventions; use it directly when you report a trial of a diagnostic, triage or treatment support system.

Where to use it in The ELSA Way¶

• Reporting and transparency › International reporting guidelines: This is the standard to follow when your evaluation is a randomised trial of an AI intervention.

• Clinical utility and safety › The clinical evaluation plan: If your clinical evaluation plan points to a randomised trial, the checklist shows what you will need to record about the AI, its inputs and its use so that the trial can be reported in full.

References¶

• Liu X, Cruz Rivera S, Moher D, Calvert MJ, Denniston AK; SPIRIT-AI and CONSORT-AI Working Group. Reporting guidelines for clinical trial reports for interventions involving artificial intelligence: the CONSORT-AI extension. Nature Medicine. 2020;26:1364-1374. Liu et al. (2020)

• Liu X, Cruz Rivera S, Moher D, Calvert MJ, Denniston AK; SPIRIT-AI and CONSORT-AI Working Group. Reporting guidelines for clinical trial reports for interventions involving artificial intelligence: the CONSORT-AI extension. BMJ. 2020;370:m3164. Liu et al. (2020)

Related tools¶

• SPIRIT-AI: An extension of the SPIRIT checklist that tells you what to put in a clinical trial protocol when the intervention includes AI.

• DECIDE-AI: A reporting checklist for small, early clinical studies of AI decision support, where you first test a tool with real users before a large trial.

• TRIPOD+AI: A 27-item checklist for writing up studies that develop or validate a clinical prediction model, whether it uses regression or machine learning.