The differential effects of presenting uncertainty around benefits and harms on treatment decision making.

Abstract

OBJECTIVE: The study aimed to determine the effects of presenting uncertainty of quantitative risk information about medication harms and benefits on intention to take medication and psychosocial outcomes. METHODS: 147 participants were randomized into four hypothetical scenarios: (1) harm/benefit presented as point-estimates, (2) harm as point/benefit as range, (3) harm as range/benefit as point, (4) harm/benefit as ranges. We assessed participants' intention to take medication, understanding, decisional conflict, and perceived benefit and harm. Participants viewed the scenario again with swapped range vs. point-estimate presentations; intention was re-assessed at time 2. RESULTS: There were no significant differences in intention across arms (p=0.93). There was an increase in intention to take medication between time 1 and time 2 in arm 4 (both harm/benefit presented as ranges, then point estimates). Understanding was greater when information was presented as point estimates compared to ranges, for both benefit (all questions correct: 74% vs 42%; χ2=15.28, df=2, p<0.001) and harm (all questions correct: 70% vs 43%; χ2=11.19, df=2, p=0.004) CONCLUSION: Changes in benefit information appear to be more influential in changing intention to take medication than harm information. Presenting uncertainty decreases understanding of information. PRACTICE IMPLICATION: The findings may help develop optimal ways to present uncertainty.