Table 1 The baseline recruitment protocol

1) The CPRD advertised the study to practices already participating in CPRD data collection; volunteer practices were identified and made known to the research team. Their practice managers were approached.

2) The research team provided each practice with a model letter from GP to patient introducing the study and the researchers. One generic letter was signed by a doctor in the practice and returned to the research team for copying and inclusion in mailings to patients.

3) The CPRD sent each practice manager an electronic file listing the CPRD-coded identifier and a special study code number for each patient from the practice eligible to take part in the study (those born between 1948 and 1962 inclusive).

4) A member of the practice staff added the name and address of each patient to this file.

5) The GP excluded any patients from the mailing list whom he or she felt should not be approached (e.g., because of terminal illness or recent bereavement).

6) A member of the practice staff printed an address label for each of the remaining patients, including the study code number and the name and address.

7) The research team delivered to each practice a set of sealed envelopes with postage pre-paid, each marked with a study code number.

8) Practice staff attached the appropriate address label to each envelope and mailed them. Envelopes for patients withdrawn from the study were counted and destroyed.

9) Questionnaires were returned directly to the research team. Consent to baseline self-completed information was signified by return of a questionnaire. Additionally, participants were asked to complete and return a signed consent to the further stages of data linkage and postal follow-up, and to provide their name and contact details to the research team to enable follow-up without need for further involvement of the practices.