Recruitment strategies for recruitment to RCTs: a systematic review of controlled trials and observational studies

Patrina H. Y. Caldwell, B Med, Dip Paeds ¹²
Jonathan C. Craig, MBChB, MM (ClinEpi), DCH, FRACP, PhD ¹²&sup4;
Sana Hamilton, B Sc in Nursing, MPH ¹
Phyllis N. Butow, BA (Hons), MPH, MClinPsych, PhD ³

1.Centre for Kidney Research, the Children’s Hospital at Westmead
2. Department of Paediatrics and Child Health, University of Sydney
3. Medical Psychological Unit, University of Sydney
4. Department of Public Health and Community Medicine, University of Sydney

Background: Recruitment of participants for randomised controlled trials (RCTs) is often problematic. Many strategies have been tried to improve recruitment. The aim of this study is to assess the effectiveness of different methods of recruiting participants to RCTs.

Method:A comprehensive search of Medline, Embase, the Cochrane Library and reference lists of retrieved papers was done to find all randomised and observational studies that compared two or more methods of recruiting participants to RCTs. Two reviewers independently searched, ascertained eligibility, assessed studies for methodological quality and extracted data. Outcome measures were accrual rates, proportion enrolled by each method and cost per participant recruited by each method. Summary estimators of effects were calculated using a random effects model and expressed as odds ratio with 95% confidence intervals.

Result:From 4340 potentially eligible studies, 41 studies were included -6 RCTs, 2 quasi RCTs, 24 cohort studies, 12 cross sectional analytical studies and 1 before-after study. The highest accrual rates were referrals vs community outreach 5.20 [4.73,5.71]; community outreach vs review of records 5.05 [3.10,8.22]; referrals vs review of records 2.79 [2.06,3.79]; referrals vs mailing 2.77 [2.51,3.05]; and referrals vs media 2.40 [2.22,2.59]. The greatest proportion enrolled were mailing vs community outreach 11.05 [10.63,11.49]; mailing vs media 6.54 [6.31,6.78]; review of record vs referral 6.30 [5.63, 7.05]; community outreach vs personal approach 6.23 [5.56,6.99] and media vs word of mouth 4.97 [3.83,6.45].

Conclusion:Interventions used to enrol patients in trials are very heterogeneous, with referrals have a higher accrual rate compared with other methods. However, the low proportions of participants enrolled by this method means other strategies where large numbers of potential participants are approached may be more effective.

Presented at the International Clinical Trials Symposium, Darling Harbour, Sydney NSW Australia 21-23 Oct 2002

Correspondence
P Caldwell
Centre for Kidney Research
The Children’s Hospital at Westmead
Locked Bag 4001
Westmead NSW 2145 Sydney Australia
Tel: +61 2 9845 3037
Fax: +6 12 9845 3038
Email:Patrina Caldwell