Original ArticleIndications and requirements for the use of prerandomization
Introduction
Currently, evidence-based medicine is widely propagated as the reference for good clinical practice. This development involves a critical appraisal of a widening range of public health or clinical interventions. As a consequence, the need for patient-related research is growing, as is demonstrated by the increasing number of such scientific publications. In effectiveness studies, the conventional randomized trial will always be the design of first choice. However, more and more investigators are becoming aware that the conventional design, in which informed consent is obtained in all participants before randomization, is not the panacea for all research questions, because in certain situations, information in the reference group about the experimental intervention could lead to serious protocol violations. Therefore, in the last two decades, several investigators applied prerandomization designs to counter this contamination problem, questioning the absolutism of the conventional design [1], [2], [3], [4]. The term prerandomization implies that the randomization takes place before seeking informed consent.
Because of this deviant informed consent procedure, the position on prerandomization of medical ethicists, health lawyers, methodologists, and clinicians remains controversial, and varies from total unacceptability to acceptance under specific conditions [5], [6], [7], [8], [9], [10], [11], [12]. Unsurprisingly, there is controversy both between and within research ethics committees (REC) about when, or even whether, these designs are admissible [13].
Prerandomization is also applied in apparently less controversial designs in which clusters, such as primary care practices or schools, rather than individuals, are randomized, and obtaining informed consent from the participants after randomizing the clusters is inevitable. Another situation in which prerandomization is used is in patients who are unable to give informed consent because of a reversible process of disease, such as comatose or psychiatric patients who may be able to give consent after a period of convalescence [14], [15]. Whether this is justified and in which circumstances a temporarily proxy consent may be used is a complex issue, and the ethical and judicial ramifications are discussed elsewhere [16], [17], [18], [19]. This application of the prerandomization design is beyond the scope of this article.
In this article, we derive guidelines for the optimal application of prerandomization designs.
Preceding this, different types of prerandomization designs are outlined, and an overview of methodological differences between conventional designs and prerandomization designs is given. In the last section, ethical and judicial considerations concerning trials and informed consent are described.
Section snippets
Prerandomization designs
Box 1 shows the definitions of key concepts use in this article. In 1977, Zelen introduced the single-consent design [20], [21]. The key characteristic of this design is that, patients are randomized before informed consent is sought. The patients allocated to the experimental treatment group are approached by the investigator or clinician for informed consent. However, the patients randomized to the reference group are given the (best) standard treatment but are not approached for informed
General issues
Clinical trials should provide information about the relative magnitude of effect(s) of an experimental intervention on (disease) outcome as compared with some reference intervention(s) or policies. This information is ultimately obtained by comparing the outcomes of interest between the study groups. The difference thus calculated may be explainable by different mechanisms. First, it could be the result of the experimental intervention. Second, it could be the result of bias. Briefly, bias may
Ethical and judicial considerations
In a clinical trial, there is a balance between individual and collective ethics [33], [34]. “Individual ethics” implies that each patient should receive the treatment which is thought to be most beneficial for his condition. This is the clear aim of good clinical practice, in which the patient and his clinician decide together on the best course of action. Usually, the clinician proposes the therapy on the basis of his knowledge and experience, which is followed by the appropriate
Conclusion
In general, the use of prerandomization in study designs is underestimated. However, based on well-defined indications and requirements, prerandomization designs have an essential contribution to evidence-based medicine. Banning prerandomization a priori implies that information about the effectiveness of numerous public health and medical intervention will not be forthcoming. A number of effective interventions will not be introduced because of the lack of evidence. Furthermore, the use of
References (47)
- et al.
The Zelen design may be the best choice for a heroin-provision experiment
J Clin Epidemiol
(1999) - et al.
Informed versus randomised consent to clinical trials
Lancet
(1995) Informed consent: protection or obstacle? Some emerging issues
Controlled Clin Trials
(1997)- et al.
Members of research ethics committees accepted a modification of the randomized consent design
J Clin Epidemiol
(2005) - et al.
Randomised consent designs in cancer clinical trials
Eur J Cancer
(1995) - et al.
Ethics and practice: alternative designs for phase III randomized clinical trials
Controlled Clin Trials
(1999) Health care research in ambulatory care
Ethiek en recht in de gezondheidszorg
(1997)- Schellings R, Kessels AG, ter Riet G, Knottnerus JA, Sturmans F. Randomized consent designs in randomized controlled...
Randomization designs in comparative clinical trials
New Engl J Med
(1984)- et al.
Ethical guidelines for epidemiologists
J Clin Epidemiol
(1991)
Informed consent en prerandomisatie
Using the Zelen design in randomized controlled trials: debates and controversies
J Adv Nurs
Ethics, data-dependent designs, and the strategy of clinical trials: time to start learning-as-we-go?
Stat Methods Med Res
An argument that all prerandomized clinical trials are unethical
J Med Philos
Research ethics and consent in the intensive care unit
Curr Opin Crit Care
Lithium carbonate in chronic schizophrenia: a brief trial of lithium carbonate added to neuroleptics for treatment of resistant schizophrenic patients
Acta Psychiatr Scand
Novel consent process for research in dying patients unable to give consent
BMJ
Protecting subjects with decisional impairment in research: the need for a multifaceted approarch
Am J Crit Care Med
Consent and the mentally disordered detained patient
Br J Nurs
Research on subjects incapable of giving informed consent: the situation in Dutch intensive care departments
Intensive Care Med
Statistical options in clinical trials
Semin Oncol
A new design for randomized clinical trials
N Engl J Med
Cited by (35)
Should the cost perspective be integrated in clinical guidelines?: On individual and societal perspectives in health research
2019, Journal of Clinical EpidemiologyInvestigation of the “m” in the cmRCT (cohort multiple randomized controlled trial) design revealed dependence between trial results
2018, Journal of Clinical EpidemiologyA review of cluster randomized trials found statistical evidence of selection bias
2018, Journal of Clinical EpidemiologyCitation Excerpt :Other studies have highlighted problems of cluster trials with a particular reference to recruitment bias [4], and recent books on cluster trials [5–7] alert the reader to this problem. Recruitment bias can be a particular problem with some cluster trials when they adopt a prerandomization (randomization before recruitment) design, which is similar in concept to Zelen's method (i.e., randomizing without consent and only obtaining consent from those assigned to the intervention group) [8]. Previous reviews on the methodological issues of cluster trials have shown that a significant proportion of cluster trials recruit participants after randomization, even when it is possible to recruit them beforehand [2].
The evaluation of pharmacological drugs, medical devices, and non-pharmacological or public health interventions: Experimental design limitations. Moving towards new methods?
2016, Revue d'Epidemiologie et de Sante PubliqueThe ethics of cluster-randomized trials requires further evaluation: A refinement of the Ottawa Statement
2015, Journal of Clinical EpidemiologyCluster randomized trials must be better designed and reported
2015, Journal of Clinical Epidemiology