Abstract IA20: Maximizing the impact of risk prediction models: Leveraging lessons learned from risk communication research

The most promising new medications are doomed to be ineffective if not adequately prescribed and taken as directed by patients, highlighting the grave importance of understanding the vicissitudes of human behavior. The same might be said of risk prediction tools. Irrespective of their quality and validity, the successful use and impact of such tools hinges firmly on a thorough understanding of human motivation, emotion, and cognition – the building blocks of human behavior and decision-making. In general, people desire to minimize loss, uncertainty, and ambiguity, and they hold defensive self-serving beliefs about their risk and risk factors, particularly when they compare themselves to other people. People also construe risk in terms of dimensions such as dread, absolute frequency, controllability, and intuition rather than objective likelihood, and fail to consider base rates when assessing risk (rendering their risk judgments non-Bayesian). They also endeavor to appear to themselves and others as rational actors, often leading to the paradoxical choice of non-dominant options, and are influenced – often beyond awareness – by incidental emotions and secondary motives such as managing existential anxiety when evaluating personal risk and making consequential decisions. People also vary greatly in how they use and comprehend numerical information and in the comfort with which they do so. Risk communication strategies have been developed to reduce the undesired consequences of these phenomena on risk perception and decision making, and in some cases can be implemented quite easily into the outward design of a risk prediction tool as well as the manner in which it is used in clinical practice. Citation Format: William MP Klein. Maximizing the impact of risk prediction models: Leveraging lessons learned from risk communication research. [abstract]. In: Proceedings of the AACR Special Conference: Improving Cancer Risk Prediction for Prevention and Early Detection; Nov 16-19, 2016; Orlando, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2017;26(5 Suppl):Abstract nr IA20.