“…For Zikmund-Fisher et al, the most adequate presentation is to use a large denominator (1,000 or 10,000) so that the data can be presented in whole numbers (Zikmund-Fisher et al, 2008). Nevertheless, Waller et al (Waller, Whitaker, Winstanley, Power, & Wardle, 2014) found small differences in the intention to participate between different versions of numerical information. Information expressed in the proportion 1:3 (one life saved for three overdiagnosed women) was associated with a greater decrease in intention to participate than other information formats, although it did not enable a greater understanding of the term "overdiagnosis".…”