Public attitudes to participating in UK Biobank
UK Biobank is planned as a prospective study, with an expected sample of 500,000 people in the 45 to 69 age range, to examine these interactions in a range of common late onset diseases The costs associated with establishing such a cohort mean that it is essential to make recruitment as successful and cost-effective as possible. It is therefore important to identify those elements of the methodology that will be significant determinants of whether a person approached to be a research participant actually consents to be part of UK Biobank or not.
This public consultation was conducted to ask people about whether they themselves would be interested in participating in UK Biobank, and their preferred options in relation to four issues that had been identified as being important to the general public in previous UK Biobank public consultations.
In the initial interview, participants will be asked about their lifestyles, for example smoking, consumption and exercise habits. At the same time various biochemical and physiological investigations will be performed. These procedures will provide information about the health status of individuals at the start of the study, and their risk of subsequent disease. The first issue is whether participants would want to receive feedback about these tests.
DNA will be extracted from each blood sample for research over the ensuing ten years. Information about any diseases which develop during the follow-up period will be collected via the UK National Health Service medical record. Participants may have concerns about giving consent for their DNA and other information being studied over such a lengthy period.
Participants may withdraw from the study at any time, and may allow the project to continue making use of the DNA and information already provided, or may want all their data to be destroyed.
Participants may have concerns about who will have access to information about their health and DNA, as the database will be an attractive resource to researchers from a range of organisations.
The consultation was conducted as a vehicle for evaluating different methods for consulting with the public. The full report containing the results of the consultation two-fold: firstly, to evaluate the relative effectiveness of conjoint analysis as a tool for consulting the general public; secondly, to determine the preferences of the general public related to issues associated with participating in UK Biobank.
Methods
Traditional quantitative methodologies were used as a baseline against which to assess conjoint analysis. Likert scales were used to measure the strength of agreement with possible options associated with the issues described above in postal and interview elements of the consultation. Respondents were asked to rank their preferences in a separate, similar, questionnaire. Respondents were asked to make pairwise choices between scenarios (which included all four issues and one option in each) to identify how less desirable options might be traded off against the least popular option in the conjoint analysis. Identical questions about preferences relating to the four issues described above were asked in all methodologies. A discursive version of the questionnaires was used to structure the qualitative semi-structured interviews.
Results
Respondents comprised: 1930 to the interview Likert survey; 410 to the postal Likert survey; 367 to the ranking (postal) survey; 1283 to the interview conjoint analysis; 665 to the postal conjoint; 44 to the qualitative interview. This gives a total sample of 4655.
Recruitment to the Likert and conjoint interview elements of the consultation was at 180 locations across Great Britain by market research interviewers. Names for the postal elements (Likert, ranking, and conjoint analysis surveys) were selected from the North East Derbyshire electoral roll. Interviewees for the qualitative element were selected from respondents to the postal surveys.
Feedback
There was overwhelming agreement among consultees in both the postal and interview surveys that there should be some form of feedback. A very high percentage indicated strong disagreement with the option of no feedback being provided. The preference was for any clinical findings to be given directly to the individual, although feedback via their GP was also acceptable. Results from the ranking element of the study were similar, with the lowest rank given for the no feedback option by over 95% of the sample. These results were mirrored by the conjoint analysis, though less emphatically: the interview and the postal samples both wanted feedback. The interview sample had no preference about who received the information, while the postal sample showed a small preference for the individual receiving the feedback. The qualitative enquiry found virtual unanimity in the public´s wish for feedback, with a slight preference for feedback to go to the individual direct.
Consent
A small majority in the Likert interview sample preferred consent to be sought annually, although a substantial minority strongly agreed that consent need only be sought at the start of the project. The postal sample, in contrast, comprised a small majority in favour of a single consent. Similarly, in the conjoint element, a small majority of the interview sample preferred regular consent, while the postal sample did not show a preference. In common with the results of the postal Likert, the postal ranking and the qualitative samples preferred a single consent at the start of the study.
Withdrawal
Retention of both DNA and health information was the preferred option of the postal Likert sample, whose next most popular option was to destroy everything, rather than keeping a person´s health data and destroying the DNA. More people strongly agreed with the `destroy everything´ option in the Likert interview. The results of the postal ranking again agreed with the postal Likert, but a larger majority strongly agreed that both DNA and information should be kept. Almost eighty per cent of the ranking sample were against destroying either the DNA or health information which had already been collected, which was the least favoured option. The postal conjoint analysis sample also preferred not to have everything destroyed, but the interview conjoint sample had no preference on this issue. The qualitative sample tended to see a need to destroy everything if people felt strongly, while regretting the waste of destroying data and DNA already collected.
Access
The favoured options regarding who should have access to the database tended to be NHS and university researchers, with insurance companies as a group to whom access should be denied. However, in the interview Likert sample the police were second favourite, whereas pharmaceutical and biotech companies were ranked second after the NHS and universities in the postal sample. Over 80 per cent of the postal ranking sample gave 1st rank to the NHS and universities, and 4th (bottom) rank to insurance companies.
Participation
The greatest proportion of respondents (24%) in the Likert and conjoint interview samples would definitely not be interested in taking part in UK Biobank compared with 20 per cent who were unsure and 19 per cent who definitely would be interested. These percentages vary from the postal Likert sample of whom only 6 per cent definitely would not be interested, 20 per cent who were unsure, and 25 per cent who definitely would be interested. All except one respondent in the qualitative sample in the 45 to 65 age group expressed eagerness to be a part of the UK Biobank study.
| 
