Background: External validity is central to pragmatic trials. Recruitment of a 'representative' sample is fundamental to this. However, recruitment is a complex process, requiring, first researcher access to eligible participants. Questions have been asked about the match between study samples and real-world clinical patients and the impact this has on generalisability of findings.
Method: Using interview and observational data from a grounded theory process evaluation of a multicentre trial, we investigated factors influencing referral.
Results: We found evidence of a fundamental disjunction between research and clinical practice manifested in varying 'filtering' processes. Organisational culture and the knowledge and attitudes of service providers regarding research generally and specifically in this RCT resulted in exclusion of very many potentially eligible participants.
Conclusions: These findings have policy and practical implications and highlight the need for development of genuinely collaborative partnerships between the research and clinical communities. Importantly, we suggest that the CONSORT recommendations be further refined to identify the potential study population.