Barriers to recruitment when conducting a commissioned randomised controlled trial of medication versus psychological therapy for generalised anxiety disorder: some lessons learned

Abstract

Background

Poor recruitment is the most common reason for premature discontinuation of randomised controlled trials (RCTs). An RCT of medication versus psychological therapy for generalised anxiety disorder (GAD) was discontinued prematurely by the UK National Institute of Health Research funders because of recruitment failure. In order to inform future research studies, this article explores the reasons for poor recruitment and aspects which could have been improved.

Methods

The trial recruited participants via psychological well-being practitioners (PWPs) employed within local Improving Assess to Psychological Therapies (IAPT) services at four sites in England. For this study, we initially examined the recruitment data to identify reasons why potential participants were reluctant to participate in the trial. We then investigated reasons the PWPs did not identify more potential participants. Finally, we performed retrospective analyses of a computerised clinical records system used by the IAPT services in this study. These analyses aimed to establish the number of potential participants who had not been approached about the trial as well as whether there were additional factors affecting the numbers of people who might be eligible to take part. Data were obtained for all patients assessed during the period from the date on which recruitment commenced until the closure of the trial.

Results

Three quarters of those patients identified as possibly suitable for the trial declined to take part; the great majority did so because they did not want to be randomly assigned to receive medication. Our retrospective database analyses showed that only around 12% of potentially eligible patients for the trial were identified by the PWPs at the pilot sites. The results also indicated that only 5% of those noted at entry to the IAPT services to have a score of at least 10 on the GAD-7 questionnaire (a self-completed questionnaire with high sensitivity and specificity for GAD) would have been eligible for the trial.

Conclusions

Our findings suggest that poor recruitment to RCTs can be significantly affected by participants’ treatment preferences and by factors influencing the recruiting clinicians. It may also be important not to include too many restrictions on inclusion criteria for pragmatic trials aiming for generalisable results.

Trial registration

ISCRTN14845583. Registration date: 5 February 2015.