LAUSR

Background It is widely acknowledged that many clinical trials do not always provide results that are meaningful to patients. We sought to utilise market research techniques, widely used in non-health consumer product development, to understand patient preferences that could improve the design of clinical trials. We use an example from preventative treatments in patients with preclinical asymptomatic RA where many trials are being designed or ongoing. Objectives To consider how to inform trial design for:i) which outcome/s should be primary? ii) what difference in the primary outcome between arms is important? iii) will patients want to use the intervention if it meets its primary endpoint/s?, and iv) does an alternative strategy exist that patients would prefer? Methods We developed a discrete choice experiment and surveyed first-degree relatives of patients. Focus groups of RA patients, first-degree relatives of RA patients and rheumatologists identified 5 key attributes of treatment (reduction in risk of RA, how treatment is taken, chance of side effects, certainty in estimates, health care providers opinion). DCE data was analysed using a conditional logit regression models to estimate the significance and relative importance of attributes in influencing preferences. We predicted uptake using estimates from the opt-out data analysed using a logit model. Results 288 first-degree relatives of people with RA started and completed all tasks in the survey. The majority of the sample were aged between 18 and 39 years (60%), and 60% female. All attributes levels significantly influenced preferences for treatments, but how treatment is taken (oral vs. infusion β0.983, p<0.001) was the most influential, followed in similar magnitude by increasing risk reduction (60 to 24 in 100) (β0.922, p<0.001), matching of patient and health care professional preferences (β0.900, p<0.001), and reducing risk of side effects (β0.839, p<0.001). If a risk reduction of RA from 60 in 100 over 5 years to 44 in 100 is realised with only minor, reversible side-effects likely, then the uptake of hydroxychloroquine was predicted to be 86%. If all treatments currently under study in the pre-clinical phase of RA were assumed to be options for the asymptomatic phase and met hypothesised outcomes, the uptake of oral methotrexate was predicted to be 46% and hydroxychloroquine 20%. Predicted uptake of biologic drugs was 6% for abatacept and 4% for rituximab. Conclusions The study illustrates how market research can be used to design clinical trials that address patient centred priorities and outcomes. The results illustrate that a trial of preventative treatments for RA should: i) be powered to detect both a difference in preventing the development of RA, and the increase in minor side-effects, ii) require a significant reduction in risk of developing RA if any side-effects are possible. We calculate iii) that hydroxychlorquine would be likely to be used by pre-clinical asymptomatic patients, but biologics would likely not, iv) and that methotrexate should also be explored as an earlier option. Acknowledgements CIORA, BC SUPPORT Unit Disclosure of Interest None declared