The effect of positively framing side-effect risk in two different formats on side-effect expectations, informed consent and credibility: A randomised trial of 16-75 year olds in England

The effect of positively framing side-effect risk in two different formats on side-effect expectations, informed consent and credibility: A randomised trial of 16-75 year olds in England
R. Webster, G.J. Rubin
Springer Nature, 2 June 2020
Abstract
Background
Reframing side-effect information in patient information leaflets (PILs) in terms of those who remain side-effect free may reduce negative expectations and side-effects, although there are concerns this may impact informed consent. This study compared two versions of positively framed PILs to current practice to see which reduces side-effect expectations whilst maintaining informed consent and credibility.
Methods
We commissioned Ipsos MORI to conduct an online survey of 16-75s in England. 1067 people completed the study and were randomised to receive a PIL for a hypothetical new antibiotic that either communicated side-effects following current practice (n=356), used positive framing with natural frequencies (n=356), or positive framing with percentages (n=355). After reading the leaflet participants completed measures of their side-effect expectations, absolute risk perceptions, and satisfaction and credibility of the leaflet.
Results
Both positively framed PILs resulted in significantly lower side-effect expectations compared to the current PIL for all side-effects (p’s<.001), apart from seizure. Pairwise comparisons showed no difference in side-effect expectations between the two positively framed PILs (p’s>.626). The positively framed PIL using natural frequencies produced more accurate risk perceptions than the same leaflet using percentages; but performed equally to the current PIL. There was no difference between the leaflets in terms of satisfaction with or credibility of the PILs
Conclusion
Positively framed PILs using natural frequencies significantly reduced side-effect expectations and provided the most accurate risk perceptions without impacting satisfaction or credibility. Replication is needed with patients prescribed new medication and those with lower educational status.