Information about a treatment’s benefits and harms available to a patient often relies on text. However, for many medical conditions, patients must trade off benefits and harms across multiple competing treatments. It remains unknown how to appropriately communicate information on benefits and harms to patients.

We compared three communication tools using textual information (Cochrane summary of findings table) or increasing combinations of textual and graphical information (Kilim and Vitruvian plots, respectively) to convey the available evidence.

Communication of Benefit–Risk Information, an online randomised controlled trial, is a three-group, parallel, open-label, automated, randomised controlled trial (no. NCT05917639). We recruited participants aged between 18 and 65 years from the general population. Participants were randomly allocated (1:1:1) to one of the three communication tools providing information on competing fictional treatments for social anxiety, and were asked to choose one based on externally provided preferences. The primary outcome was the perceived level of decisional conflict when selecting a treatment (decisional conflict scale (DCS): 0 = best, 100 = worst). Because this was an all-or-nothing, single-visit trial, only those participants providing data contributed to the primary analyses (modified intention to treat).

We recruited 2178 adults between 1 June and 27 November 2023. Vitruvian and Kilim plots outperformed the Cochrane summary of findings table on the primary outcome (adjusted mean difference −10.9, 95% CI −13.5 to −8.2, P < 0.0001 and −9.7, 95% CI −12.4 to −7.1, P < 0.0001), respectively). Results varied by participants’ literacy and numeracy skills, lived experience of the condition of interest, ethnic group, gender assigned at birth and age.

Combining graphical and textual information, as opposed to text only, improved communication and reduced decisional conflict when choosing across multiple competing medical interventions. Organisations involved in disseminating scientific evidence should consider endorsing a combined graphical and textual approach and adopting more intuitive and accessible communication methods. We identified several prognostic factors that should inform the development of future patient decision aids and communication of scientific findings.

NCT05917639.