Public preferences for participation in a large DNA cohort study: a discrete choice experiment

Objectives. To determine the general public’s preferences over the design and use of UK Biobank; and the design for optimum recruitment.

Design. Discrete choice face-to-face interviews using a fractional factorial design and multinomial logit regression modelling.

Setting. 180 sampling points across 11 regions of the UK.

Participants. Members of the public.

Main outcome measures. Relative risks of people’s preferences for project design and use.

Results: 34.4% of respondents were willing to take part in UK Biobank (n=1283). The most highly preferred scenario was: individual feedback from the study; consent every time new data is requested; DNA and information destruction on withdrawal; and access to the data by the NHS and Universities but not other third parties. The single most important attribute was access to data. If individuals' insurance companies were to be given access to the data, this would be the largest single impediment to recruitment to the study. Extra resources are likely to be needed to counter the reduced recruitment rate if pharmaceutical companies are allowed access to the data.

Conclusions: The general public do have clear preferences regarding the design of biobanks. Whilst designing the study to meet the most preferred scenario may not be practical within available resources, biobanks can use the type of information provided here to compare the costs and benefits of different study designs. The "price" of discounting public preferences in terms of reduced recruitment should be an important part of the "weighing" process. Pilot studies of recruitment under alternative study designs may be justified.