3.4. Recruitment, randomisation and blindingThe nurse consultant of the participating community mental healthteam was asked to nominate patients who met the study inclusion criteria. Each eligible participant was given a unique identificationnumber in sequence according to their alphabetical order of surname. Alist of random identification numbers was generated by an online external randomisation service (researchrandomizer.com) and this listwas used to determine the order in which patients were approached bytheir keyworker. Participants were recruited in three consecutive cohorts over a three-month period.