Can the trustworthiness of a randomised controlled trial really be judged by whether it shares its raw data?

Researchers from Melbourne’s Monash University are advocating for the increased publication of individual participant data, based on their meta-epidemiological study published in eClinicalMedicine in July.

Examining 265 randomised controlled trials on labour induction, the study found that trials that shared these data were more likely to report consistent results and meet trustworthiness criteria, such as being prospectively registered and producing realistic-looking results.

Trials that did not share these data were more likely to show exaggerated treatment effects and had a higher likelihood of statistical inconsistencies, the team said.