For example, trial 36 had named a ‘patient representative’ as a member of the TSC at the application stage then subsequently, in direct response to peer reviewer comments, the team had indicated that they would consider increasing the number of ‘patient representatives’ on kinase inhibitor Nilotinib the TSC from one to two, in order to provide ‘mutual support’. The team proceeded to include two PPI contributors on the TSC, thereby achieving their documented plans. Despite having prior experience of PPI, however, the researcher divulged no personal expectations for PPI within this particular trial and referred to PPI as a ‘tick box’ exercise: It

was a requirement of…that we had representation on our steering committee and therefore I went through that […] We can say [the PPI contributors] are there and therefore it’s, if you like, ticking a political box. (CI 36) The documentation for trial 2 included no plans for PPI during the trial but did state that there had been ‘several stages of user involvement’ prior to the

grant application, “to confirm that the research question is pertinent to both the needs of the NHS and the NIHR programme of research development.” Two grant reviewers commented on the lack of ‘service user representation’ on the team and suggested membership ‘on the research team or steering group’. The TSC did include PPI membership but during the interview the researcher spoke of his initial ‘tokenism’ and ‘ignorance’ about how PPI ‘should and could work’. When asked about the expectations of their role, the PPI contributors in two other oversight trials (115 and 96) implied similar uncertainties when they spoke of not knowing what was expected of them and of feeling ‘bewildered’ in meetings: I can’t understand why they use me… they seem to find me useful but I just sit there bewildered. I’m there as a sort of grey background while the others do all the sparky stuff. (PPI 115) In the next section we describe planned and implemented PPI in 14 trials which incorporated a managerial role of PPI. Unlike the six trials with a mainly oversight mode, many of the managerial mode trials had utilised

more than one form of PPI. Beyond oversight, into managerial mode (n=14) Most of these 14 trials had indicated some type of managerial involvement in the documented plans, usually to include PPI contributors as co-investigators (table 1). Two trials (4 and 27) did not have PPI contributors AV-951 as co-investigators but planned to include PPI contributors on the trial management group, and interviews with informants indicated that this had been implemented. It was unclear in one ongoing trial whether there was a PPI co-investigator, but documented plans stated that a named PPI collaborator would be “directly involved in decision making of trial processes and then relay back information to user groups”; according to the PPI contributor interview these plans were being implemented (trial 18).