Family Practice Advance Access originally published online on September 13, 2007
Family Practice 2007 24(6):601-603; doi:10.1093/fampra/cmm051
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Short report: How often do UK primary care trials face recruitment delays?
a National Primary Care Research and Development Centre, University of Manchester, Manchester
b Department of Primary Care & General Practice, University of Birmingham, Birmingham
c Institute of General Practice and Primary Care, University of Sheffield, Sheffield, UK
Correspondence to Peter Bower, National Primary Care Research and Development Centre, University of Manchester, Manchester M13 9PL, UK; Email: p.bower{at}manchester.ac.uk
Received 19 February 2007; Revised 21 June 2007; Accepted 9 July 2007.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Recruitment to trials is often viewed as problematic but data are scarce. This study surveyed authors of published primary care trials to assess the scale of recruitment problems. Seventy trial authors were surveyed with a response rate of 56%. Less than one-third of trials recruited to their original timescale. Recruitment requiring GPs to gain patient consent was significantly associated with recruitment problems. The data may be useful in the wider drive to improve recruitment in primary care.
Keywords. Randomized controlled trails, recruitment.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Trials need sufficient patients to ensure statistical power and validity, but recruitment remains problematic.1 A previous study of UK trials found that less than one-third recruited to target, but reported few significant factors associated with success.2 Although recruitment in primary care is always viewed as a particular challenge, data concerning the exact magnitude of difficulties are scarce.
We surveyed authors of published trials to examine:
- (a) the extent of recruitment difficulties;
- (b) responses to recruitment problems; and
- (c) the relationship between trial characteristics and recruitment.
- (b) responses to recruitment problems; and
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
We identified randomized trials through an online search of three medical journals which publish primary care trials routinely (British Medical Journal, Family Practice and British Journal of General Practice) and included trials from UK primary care requiring individual patient consent, published during 2000–2005.
We emailed a questionnaire to the corresponding authors, with two reminders. Where authors had published more than one eligible trial, we chose one randomly.
Analysis was largely descriptive. We identified potential predictors of recruitment success from a previous study (see Table 1).3 Factors associated with recruitment success were analysed using cross-tabulations, comparing study characteristics with a dichotomous measure of recruitment ‘success’ (defined as recruiting to time or overrunning by less than 50% of the planned time).
|
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
We identified 213 trials and excluded 125 trials on one or more exclusion criteria: outside the UK (n = 71); no patient consent (n = 45) and not in general practice (n = 36). We removed 18 trials by the same author, leaving 70 eligible, and the response rate was 56% overall (n = 39), although data on the main dependent variable was reported for 34. Responders were involved in larger trials, although the difference was not significant (mean difference 399, 95% confidence interval –214 to 1011). Table 1 shows key trial characteristics.
- (a) Extent of recruitment difficulties—the mean planned sample size was 1086 patients (SD 1562), and planned recruitment duration was 12 months (SD 8.3, range 1–36). The mean achieved sample size was 1002 (SD 1585). Ten trials (29%) recruited to timetable, 12 (35%) required up to 50% greater time than planned and 12 (35%) required over 50% additional time.
- (b) Responses to recruitment problems—these included extending the recruitment period (56%); seeking additional funds (31%); introducing other recruitment methods (18%); increasing the number of sites (44%); recalculating power (21%) and finishing with insufficient patients (18%).
- (c) The relationship between trial characteristics and recruitment—17 (50%) trials recruited to time or overran by less than 50%. One variable was statistically associated with recruitment duration. If GPs were responsible for gaining patient consent, only 12.5% of trials recruited within 50% of the planned time, compared with 61.5%, where the GP was not responsible (
2 = 5.89, d.f. = 1, P = 0.04).
- (b) Responses to recruitment problems—these included extending the recruitment period (56%); seeking additional funds (31%); introducing other recruitment methods (18%); increasing the number of sites (44%); recalculating power (21%) and finishing with insufficient patients (18%).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Our restriction to published studies means that the results cannot be representative of all trials and are likely to overestimate the effectiveness of recruitment. The response rate was poor, and limiting the inclusion of more prolific and successful trial authors may overestimate recruitment difficulties. Cross-sectional associations cannot determine cause–effect relationships, and the data were restricted to self-report. The analysis of factors associated with recruitment had low power, and some important factors (e.g. the type of disorder or intervention) were not tested. Furthermore, it is difficult to distinguish cause from effect, as trials may adopt methods because of poor recruitment, rather than those methods causing recruitment problems. However, the finding that GPs gaining patient consent is associated with recruitment problems supports the previous findings.1 Clearly, a larger study using prospective data collection from a trial register is indicated.
Despite these limitations, the results do provide an estimate of the range of delays in primary care trials. Clearly, trials running past their planned recruitment timetable are the norm, and one-third of the published trials were forced to seek additional funds. This has clear implications for both funders and grant applicants.
There is a small but emerging literature on determinants of recruitment.2,4 If the potential of UK primary care as a platform for high-quality trials is to be realized, there is a need to consider a range of potential interventions and begin a programme of work to test and disseminate different recruitment methods.
|
Declaration |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Ethical approval: None.
Conflicts of interest: None.
|
Acknowledgments |
|---|
The work was carried out on behalf of The Royal College of General Practitioners Research Group.
|
Notes |
|---|
Bower P, Wilson S and Mathers N. How often do UK primary care trials face recruitment delays? Family Practice 2007; 24: 601–603.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
1 Fairhurst K, Dowrick C. Problems with recruitment in a randomised controlled trial of counselling in general practice: causes and implications. J Health Serv Res Policy (1996) 1:77–80.[Medline]
2 McDonald A, Knight R, Campbell M, et al. What influences recruitment to randomised controlled trials? A review of trials funded by two UK funding agencies. Trials (2006) 7:9.[CrossRef][Medline]
3 Foy R, Parry J, Duggan A, et al. How evidence based are recruitment strategies to randomized controlled trials in primary care? Experience from seven studies. Fam Pract (2003) 20:83–92.
4 Watson J, Torgerson D. Increasing recruitment to randomised trials: a review of randomised controlled trials. BMC Med Res Methodol (2006) 6. doi:10.1186/1471-2288-6-34.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. Bower, P. Wallace, E. Ward, J. Graffy, J. Miller, B. Delaney, and A. L. Kinmonth Improving recruitment to health research in primary care Fam. Pract., October 1, 2009; 26(5): 391 - 397. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Draper, S. Wilson, S. Flanagan, and J. Ives Offering payments, reimbursement and incentives to patients and family doctors to encourage participation in research Fam. Pract., June 1, 2009; 26(3): 231 - 238. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Goodyear-Smith, D. York, H. Petousis-Harris, N. Turner, J. Copp, N. Kerse, and C. Grant Recruitment of practices in primary care research: the long and the short of it Fam. Pract., April 1, 2009; 26(2): 128 - 136. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Graffy, J. Grant, S. Boase, E. Ward, P. Wallace, J. Miller, and A. L. Kinmonth UK research staff perspectives on improving recruitment and retention to primary care research; nominal group exercise Fam. Pract., February 1, 2009; 26(1): 48 - 55. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||