To estimate the proportion of attrition at which results of drug trials for people with schizophrenia lose enough credibility to become mistrusted by relevant groups of stakeholders. A piloted questionnaire was sent to 128 local clinicians, 100 relevant researchers and 104 service users and carers.

We received the biggest number of responses from the service user and carer group (n=81, 76%); 43% of clinicians and 32% of researchers responded. All three groups suggested that the follow-up rate for a 12-week schizophrenia drug trial should be around 70–75% for the trial to be credible.

This survey suggests that relevant stakeholders, including researchers, fundamentally mistrust results of the majority of drug trials in schizophrenia. Adopting a more pragmatic trial design can help address this.