Randomised non-comparative trial

A randomised non-comparative trial, RNCT (or also non-comparative randomised trial^[1]), is a type of clinical trial where participants are randomised to different conditions (arms), but where the primary analysis involves comparing each arm separately to a historical control, predefined benchmark or something else, with no formal comparison between the two arms.^[2]

The study design appears to have arisen in oncology, where single-arm studies are not unusual. It promises reduced sample size requirements.^[3] An RNCT acts like multiple single-arm designs run concurrently.^[2] A review found RNCTs dating back to 2002, and having been used in high-profile oncology studies and also beyond oncology.^[3]

The design has been criticised by statisticians. It is unclear what benefit randomisation adds compared to running separate single-arm studies.^[3] Pavlos Msaouel described the use of the word "randomisation" as merely being "talismanic" in a 2025 article, it leading readers to think the study has the benefits of a randomised controlled trial (RCT).^[2] While the point of an RNCT is not to have a comparison between arms, about half of reported RNCTs still reported some sort of comparison.^[3]