Table 2.

Evidence used to inform choice of format of risk presentation

Finding	Inclusion in prototype intervention design/delivery
Pilot work with members of the public and healthcare professionals
When presented in colour, the colour was often more important than the number and dominated their interpretation (Usher-Smith et al., 2017b).	Inclusion of colour in risk presentation while ensuring that the colour scheme reflects current evidence/expert opinion.
Being able to see the impact of changes in lifestyle on their risk was helpful. This included the effect of small changes (increasing fruit and vegetable consumption by one portion per day rather than meeting the requirement of five portions per day). Some also wanted to be able to see the benefits they were already achieving through their current lifestyle (Usher-Smith et al., 2017b).	Incorporation of ways to demonstrate continuous change, both positive and negative, for each modifiable factor.
The first reaction of almost all when presented with their 10-year risk of an individual cancer was that it was low and not concerning, with views on what constituted a high risk ranging widely, from 0.5% to 60 %. As a result, reductions in risk were not always motivating – the risks were considered low and differences small (Usher-Smith et al., 2017b).	Provision of combined risk of multiple cancers.
Review of published literature and best practice guidance
Numerical presentation of risk as opposed to simple risk categories (moderate, high, low) appears to lead to more accurate risk perception (Waldron et al., 2011) and when investigating only the patient’s preferences towards cancer risk communication, the majority of the British women and 50% of the Australian women expressed their preferences for quantitative risk information (Julian-Reynier et al., 2003).	Inclusion of option to see risk as a percentage.
There were strong objections to the word ‘absolute’, which was seen as ambiguous. For many participants it conveyed that the risk score was ‘conclusive’, or in some way ‘definite’ that a person would suffer a cardiovascular event rather than a probability (Kirby & Machen, 2009; Hill et al., 2010).	Avoidance of the term ‘absolute risk’ and clarity throughout that risks are estimates and apply to people with the same characteristics as the individual rather than the individual person.
People need comparisons between the probabilities of different risks in order to be able to interpret absolute risk information (Julian-Reynier et al., 2003; Hill et al., 2010).	Provision of relative risk in addition to absolute risk information and comparison to individuals with a recommended lifestyle.
Presenting relative risk as number alone has been criticised as many participants did not know how to translate 2.3 times in absolute terms (Dorval et al., 2013) or because it was ‘too alarming because the risks appeared bigger’ (Fortin et al., 2001).	Inclusion of option to see risk as an absolute percentage and comparison with individual with recommended lifestyle
Treatment decisions are sensitive to the way a treatment’s effectiveness is presented. The relative risk reduction format appears to encourage the treatment the most and number needed to treat format leads to the least acceptance (Waldron et al., 2011).	Presentation of relative risk to encourage behaviour change.
Shorter timeframes (less than 10 years) may lead to more accurate risk perceptions and increased intention to change behaviour, than 10-year risk or longer, especially for older patients (Waldron et al., 2011). Some participants thought 10 years was too remote (Hill et al., 2010).	Decision made to present 10-year risk to be consistent with cardiovascular disease within primary care.
Display of risk information visually can enhance understanding compared with written information alone, particularly amongst those with low numeracy (Lipkus, 2007)	Display risk information with a simple visual for ease of understanding.
Graphical formats are perceived as helpful (Hill et al., 2010) but one format does not fit all (Dorval et al., 2013). Several formats were reported as confusing, such as line graphs, and icons, particularly those with larger numbers (Hill et al., 2010).	Inclusion of graphical presentation but avoid line graphs and icons.
People found formats which combined information helpful, such as colour, effect of changing behaviour on risk or comparison with a healthy older person (Hill et al., 2010).	Inclusion of colour, effect of changing behaviour and comparison to individual with a recommended lifestyle.
Provision of feedback from the consultation to the counselee appears to be welcomed and the interest in other tools that complement the consultation has been pointed out (eg, leaflets, CDs and other media to promote self-help) including the tailored print communication through a personal letter summarising the consultation for the counselee (Sheridan et al., 2009).	Inclusion of option to print a tailored information sheet summarising the risk assessment.
Several explained they might take their risk more seriously if they knew exactly what the calculation is based on and how the numbers affect the final percentage (Sheridan et al., 2009).	Provision for individuals to change all the modifiable factors to see how that changes the final risk estimate and provided information on the development of the risk score as additional information.
Consultation with experts and PPI members
To enable understanding of risk, incorporation of colour into the risk presentation. For this to be of use, it must have meaning.	Inclusion of a colour scale from green to red to demonstrate level of risk where green corresponds to a relative risk of 1 and then the colour changes gradually to be orange at a relative risk of 2 and then to red at a relative risk of 4
Use of relative risk is acceptable in the context of this study; however, this must be made clear to the recipient.	Clarity throughout that risks are estimates and apply to people with the same characteristics as the individual rather than the individual person.

CD = compact disc; PPI = patient and public involvement.