Original ArticleBlinding decreased recruitment in a prevention trial of postmenopausal hormone therapy
Introduction
There is extensive literature on barriers discouraging physicians and patients from participating in randomized trials (e.g., [1], [2], [3], [4], [5], [6]). Barriers for physicians include time constraints, worry over the doctor–patient relationship, loss of professional autonomy, and difficulties with informed consent procedures. Barriers for patients include preferences for assured treatment, worry caused by uncertainty, and informed consent. We have found no previous empirical research on the influence of blinding specifically on recruitment, but preliminary results have been presented of a secondary prevention trial of osteoporotic fractures (McPherson G, Avenell A, Grant AM, McGee M, Campbell MK, McGee MA. Is recruitment easier with an open rather than a blinded, placebocontrolled design? A randomised controlled trial. Paper presented at the International Society of Technology Assessment in Health Care [ISTAHC] conference in Canmore, Canada, June 22–25, 2003). A trial in homeopathy studying the influence on blinding on outcome is being planned [7].
Blinding in trials is used for many reasons. These include the wish to minimize the bias involved in observing and recording outcomes and to eliminate outcomes resulting from beliefs, expectations, and changes in caregiver and patient behavior due to the therapy [8]. The first aim is noncontroversial—but research aimed at providing information on effectiveness in real-life situations and on changes resulting from beliefs, expectations, and behavior might usefully be included in evaluation [9]. Blinding, especially if it includes use of placebo, may negatively influence the process and feasibility of trials. The effects on feasibility may be especially important in long-term preventive drug trials. In such trials, participants are usually healthy and may be more difficult to motivate than are patients looking for a treatment. Furthermore, the use of the trial drug may influence other (preventive) behaviors. Compared to treatment trials, in preventive trials physicians may also be less enthusiastic about including study participants.
In a primary prevention trial with postmenopausal hormone therapy (PHT; also known as hormone replacement therapy, HRT), we wanted to study the effect on numbers recruited and the process of recruitment when using blinding (the blind group) as compared to the situation when both the caregiver and the woman will know in which arm the woman is (the nonblind group). Our hypothesis was that more women would be recruited in the nonblind group. The influence on treatment compliance and outcomes will be studied separately after the data are available.
Section snippets
Methods
The data come from the recruitment process for the Estonian Postmenopausal Hormone Therapy Trial and from the one-year follow-up. It is a randomized trial concerning the health, social, and health-services effects of PHT. Recruitment for the trial was conducted by means of a questionnaire sent to all 50- to 64-year-old (45–64, in the pilot) Estonian-speaking women in two areas, Tallinn and Tartu and their surrounding counties, in 1998 and 1999 (n = 39,713) (Fig. 1). Names and addresses were
Numbers recruited
Figure 1 shows the recruitment into the trial. Overall, the proportion of women interested in joining the trial was relatively high: 6,605 women expressed interest (17% of those to whom an invitation was sent and 44% of those responding). On the basis of the questionnaire data, one third (35%) of the 6,605 women interested were judged ineligible, mainly on the basis of health or because they had menstruated within the past 12 months. This left 4,295 willing and eligible women (11% of the 39,713
Discussion
As expected, compared to the blind group, more women joined our preventive randomized trial when they were going to know, after inclusion, to which treatment arm they belonged. Most of this difference was explained by the women's own choice; the varying exclusions by physicians explained less. For every 100 women entered into the nonblind trial, we may predict that 20 would not have joined a blinded trial. Of these, 17 would not have attended for physician assessment, 2 would have been found
Acknowledgments
The trial was financially supported by The Academy of Finland, STAKES, and the Finnish Ministry of Education (Doctoral Programs in Public Health). The drugs were donated by Wyeth-Ayerst via the WISDOM trial (Women's International Study of Long Duration Oestrogen after Menopause), coordinated by Dr. Madge Vickers, London, UK. We thank the trial team, especially Mare Tekkel and Tatjana Veideman (TAI) for their help in field work and data analysis. Patients were recruited by physicians and
References (24)
- et al.
Informed versus randomised consent to clinical trials
Lancet
(1995) - et al.
Recruitment for controlled clinical trials: literature summary and annotated bibliography
Control Clin Trials
(1997) - et al.
Barriers to participation in randomised controlled trials: a systematic review
J Clin Epidemiol
(1999) What are the main methodological problems in the estimation of placebo effects?
J Clin Epidemiol
(2002)- et al.
Why use placebos in clinical trials? A narrative review of the methodological literature
J Clin Epidemiol
(2000) - et al.
Meta-analysis of satisfaction with medical care: description of research domain and analysis of overall satisfaction levels
Soc Sci Med
(1988) - et al.
Patient sociodemographic characteristics as predictors of satisfaction with medical care: A meta-analysis
Soc Sci Med
(1990) - et al.
Iron supplementation during pregnancy: experiences of a randomized trial relying on health service personnel
Control Clin Trials
(1989) - et al.
Determinants of long-term hormone replacement therapy and reasons for early discontinuation
Obstet Gynecol
(2000) - et al.
Estrogen replacement therapy: determinants of persistence with treatment
Obstet Gynecol
(2001)