Leeson and Tyrer aim to investigate ‘the reasons behind difficulties in recruiting patients to randomised controlled trials in psychiatry’ (Leeson & Tyrer, Reference Leeson and Tyrer2013). Their paper fails to do this and is methodologically not suited for this purpose. What it does is illustrate that there are difficulties in both instigation and recruitment; that, at least for England, these issues do not seem to be getting better; and lists some problems for consideration. Other important difficulties that apply everywhere are omitted and these have been thoroughly documented elsewhere (Duley et al. Reference Duley, Antman, Arena, Avezum, Blumenthal, Bosch, Chrolavicius, Li, Ounpuu, Perez, Sleight, Svard, Temple, Tsouderous, Yunis and Yusuf2008) (see Table 1).

Table 1. Some [but probably not all] difficulties instigating and recruiting people to randomized trials

*Mentioned in Leeson & Tyrer (Reference Leeson and Tyrer2013).

Setting research priorities is made problematic by being far from the clinical coal-face and those undertaking this task have to balance political expediency and time scales with research and clinical need. This can delay instigation of studies. Initiatives such as the James Lind Alliance, helping set answerable priorities for research, should begin to help erode this difficulty across specialities (Lloyd & White, Reference Lloyd and White2011; Lloyd et al. Reference Lloyd, White and Chalmers2012).

Little needs to be said about acquisition of funding. Delay of instigation of research can occur in spite of every effort to structure and govern funding panels broadly and fairly. It is difficult for them, in turn, not to prioritize one subspecialty against another, one district over another, and easy to remain blinkered and culture-bound. The collectivism of the funding board guards against ill-advised radical impulsive decision making and can also mitigate against innovative, imaginative and, perhaps, risky [but important] work.

Difficulties with research regulation and governance are discussed by Leeson & Tyrer (Reference Leeson and Tyrer2013), as are issues around health care partnerships and data acquisition. There is, however, no mention of how Research Ethics Committees may be acting unethically and in this way delaying either instigation (Savulescu et al. Reference Savulescu, Chalmers and Blunt1996) or undertaking of the work (Roberts et al. Reference Roberts, Prieto-Merino, Shakur, Chalmers and Nicholl2011).

A randomized trial is a collaborative effort. Leeson & Tyrer (Reference Leeson and Tyrer2013) list issues with the professional carers that may hinder progression but notably omits other issues specific to lay carers, trial participants and, most importantly, researchers themselves. These protagonists are interconnected. Why should clinical staff not be anything but apathetic about study recruitment if they feel no study ownership? This too applies to the lay carers or trial participants. When entry criteria are seen as irrelevant to everyday care, interventions are unlikely to be accessible once recruitment closes, and outcomes are time-consuming to collect and not the routine data that could change care, recruitment is likely to drag. Many specialities have not learnt the value of truly pragmatic, collaborative, design (Thorpe et al. Reference Thorpe, Zwarenstein, Oxman, Treweek, Furberg, Altman, Tunis, Bergel, Harvey, Magid and Chalkidou2009), but perhaps ours more than most. Tools, however, are now available to help considering these at design and funding stages (Tosh et al. Reference Tosh, Soares-Weiser and Adams2011).

Beyond the day to day running are issues with the design of the analysis. Why would a person with schizophrenia sign informed consent to enter a trial that collects data that may be redundant [especially is not statistically significant] and that will be analysed in a less than transparent manner (Goldacre, Reference Goldacre2012)? If more informed consents involved stipulation that the full analysis plan along with ‘dummy tables’ for the results was in the public domain perhaps informed consent would be less of a broken contract (Hayden, Reference Hayden2012). Hopefully the All Trials initiative will help (Creative Commons, 2013).

Staff intimately involved in running the trial may inhibit process for a large variety of reasons. The wrong person may end up in the role. Running a trial is an enormous demand on a person's academic, logistic and social reserve. Even for the right person, the training or support may be inadequate. Many studies are now designed and run in conjunction with a clinical trial unit (CTU). These can be enormously helpful but their ethos or understanding can also hinder progress of smooth instigation or recruitment. If a unit works primarily with cardiologists it may be problematic for them to enter the world of mental health trials. If the CTU works primarily for industry that dynamic may effect design and subsequent recruitment in other types of trials when past priorities have been focused on shareholder and regulatory authority need (Goldacre, Reference Goldacre2012) rather than those of patients or carers.

Part of the difficulty for anyone in this area is lack of knowledge of what is effective in helping instigate trials and make recruitment successful. Just as for the interventions being tested in trials, a trial-conduct evidence-base is needed. It is increasing (Treweek et al. Reference Treweek, Lockhart, Pitkethly, Cook, Kjeldstrøm, Johansen, Taskila, Sullivan, Wilson, Jackson, Jones and Mitchell2013) but remains limited. We have little knowledge about best techniques for trial-specific training of researchers, managers, clinicians, economists, statisticians, CTUs, funders, ethicists, consumers, policy-makers and commissioners. Here is where psychiatry and psychology along with colleagues in sociology and anthropology, with our expertise in understanding behaviour – and its change – may be able to lend an especially constructive hand.