Practical problems in recruiting patients with schizophrenia into randomised controlled trials

Editor—Thornley and Adams surveyed the content and quality of trials relating to the care of those with schizophrenia.¹ They found that studies were of short duration (only 19% with six months of follow up), had few patients (mean number 65), and rarely community based (19%). They concluded that “large studies, of long duration ... are needed,” but they did not discuss the possible reasons for these findings.

We are currently undertaking a community based randomised controlled trial to evaluate the effects of a patient held medical record for people with schizophrenia. Power calculations suggest that we need to recruit 100 patients into each arm of the trial (α=0.05, power=90%) to show a clinically meaningful effect on satisfaction with health care² and mental health status.³

As we had anticipated, during the first 12 months of this three year trial we encountered several practical problems affecting recruitment. Potential participants were approached through the community based mental health teams. Substantial variations in working policy and practice between adjacent teams required access to team members and patients to be individually negotiated with the six teams in the mental health trust. This was logistically complicated and extremely time consuming. The positive and negative symptoms of schizophrenia affected patient recruitment. Many of the people under the care of the outreach teams were difficult to track, and at least two home visits were required to make initial contact. Over 30% of potential participants declined to take part in the trial; people with predominantly positive symptoms were suspicious of our motives while those with negative symptoms were unmotivated to take part. Ensuring fully informed consent may also be more time consuming with such potentially vulnerable people than with other groups of patients. Participants could have been recruited through primary care, but evidence suggests that professional stigmatisation and the negative views that some general practitioners hold about people with schizophrenia would simply create a different set of recruitment problems.4

None of the practical problems described above are exclusive to people with schizophrenia, but in combination they may help to explain the lack of large randomised controlled trials called for in Thornley and Adam’s paper. We have recruited 130 people into our trial during the past 12 months, a feat that owes much to the enthusiasm of key workers in each mental health team.