Showing personalized health risks is not enough: understanding requires more than an attractive visualization
With the rapid rise of artificial intelligence and predictive models, citizens and patients are increasingly receiving personalized estimates of their health risks. This information can support decisions about treatment or prevention—but only if people truly understand it. Research by Inge van Strien-Knippenberg shows that this is far from self-evident, and that the way risks are presented can have major societal consequences.
The study focused on breast cancer treatment and breast cancer screening. Through interviews and co-creation, the researcher examined the ideas, expectations, and information needs of patients and citizens when they receive personalized risk information. Together with them, informational materials and visualizations were developed and subsequently tested for comprehensibility and usability. The research also explored differences between people with varying levels of health literacy and numeracy.
The results show that “laypeople” approach health risks differently from experts. What seems logical or important to professionals often does not align with what citizens need or expect. Existing beliefs play a major role in how people interpret risk information. Effective communication therefore needs to connect with what people already know as well as with what they want to know.
Notably, the use of visualizations - such as graphs or pictograms - does not automatically lead to better understanding. Nor do these tools reduce the gap between people with higher and lower levels of health literacy or numeracy. Visualizations are therefore no magic solution, the study shows.
If personalized risk information is not presented carefully, there is a risk that especially people with lower skills will disengage or draw incorrect conclusions. This can further widen existing health disparities. The research therefore underscores the importance of clear guidelines for risk communication and a structured, well-organized presentation of information.
For healthcare professionals and organizations that develop health information, for example for leaflets, websites, dashboards, or use in clinical consultations, this means looking beyond simply adding attractive charts. Involving the target audience from the very beginning of the design process is crucial. Only by developing and testing materials together with users can it be determined whether people truly understand the information and can use it when making health decisions.
Further research is needed to better support people with lower levels of health literacy and numeracy. In this way, personalized risk information can contribute to better and fairer healthcare, rather than to growing health inequalities.
More information on the thesis
