Risk-based innovations in cancer screening and diagnosis: A discrete choice experiment to explore priorities of the UK public

ObjectiveTo understand the importance and potential impact on uptake of different attributes of risk-based innovations in the context of risk-stratified healthcare for cancer screening and symptomatic diagnosis.DesignThe online survey comprised a discrete choice experiment (DCE) in which participants chose between two risk assessment options or to opt out of risk stratification. There were six attributes: test method, type (genetic or non-genetic), location, frequency, sensitivity and specificity. Participants were randomly allocated to consider the choice in an asymptomatic or symptomatic context.SettingMembers of the public in the UK.Participants1202 participants completed the DCE.Outcome measuresConditional logistic regression and latent class analysis informed modelling of predicted preferences for a range of innovations with different features.ResultsOverall, participants preferred risk assessments over opting out and prioritised sensitivity, with test method and specificity also important. Genetic and non-invasive tests were favoured. With sensitivity and specificity of 80% or better, participants would be more likely to take up a risk assessment than not. Comparing the asymptomatic and symptomatic contexts, 65% and 73% of participants would be very likely to participate regardless of the innovation used, and 29% and 13% of participants might participate depending on the method, sensitivity and specificity. A minority showed strong dislike of risk-based innovations, particularly within screening.ConclusionsThere are high levels of public support for risk-based innovations within risk-stratified cancer healthcare, especially for referral decision-making and using genetic and non-invasive tests. Optimising risk-based innovations is needed to engage those whose participation is contingent on test methods and performance metrics.