Endgames Statistical Question

Double dummy trials

BMJ 2011; 343 doi: https://doi.org/10.1136/bmj.d7294 (Published 16 November 2011) Cite this as: BMJ 2011;343:d7294
  1. Philip Sedgwick, senior lecturer in medical statistics
  1. 1Centre for Medical and Healthcare Education, St George’s, University of London, Tooting, London, UK
  1. p.sedgwick{at}sgul.ac.uk

Researchers evaluated the efficacy of quetiapine and of rivastigmine in the treatment of agitation in people with dementia in institutional care. Little research had been undertaken in the use of these drugs to treat this condition. A randomised, double blind, double dummy, placebo controlled trial was performed.1 The outcome measures included agitation and cognitive performance at baseline, at six weeks, and at 26 weeks. A total of 93 patients were randomised to quetiapine, rivastigmine, or placebo.

The researchers reported that, when compared with placebo, neither quetiapine nor rivastigmine was effective in the treatment of agitation in people with dementia in institutional care. When compared with placebo, quetiapine was associated with significantly greater cognitive decline.

Which of the following statements, if any, are true?

  • a) The trial was active controlled.

  • b) All trial participants received a placebo.

  • c) All trial participants received an active drug.

  • d) Patients in the placebo group received two different placebos.


Answers b and d are true, while a and c are false.

The trial was performed because there was limited research on the respective efficacy of quetiapine and rivastigmine in the treatment of agitation in patients with dementia. Quetiapine is an atypical antipsychotic, while rivastigmine is a cholinesterase inhibitor, drugs not originally developed to treat agitation in dementia. A placebo group was therefore included in the trial as a control, against which both quetiapine and rivastigmine were compared to determine whether either drug had any treatment benefits. Hence the trial was placebo controlled. If a trial compares a new treatment against the standard existing treatment rather than placebo, then the trial is referred to as active controlled (a is false).

Blinding of the recruiting clinicians, patients, and outcome assessors to treatment allocation was essential to minimise the potential of assessor and reporting biases.2 However, it was not possible to make quetiapine and rivastigmine indistinguishable from each other, and therefore blinding could not be easily achieved. Blinding was only possible in this trial by the method of double dummy. This involved the quetiapine and rivastigmine treatment groups receiving a placebo, and so therefore all participants received a placebo (b is true). Patients allocated to quetiapine also received a placebo that was indistinguishable from rivastigmine (placebo rivastigmine), while patients allocated to the rivastigmine treatment group also received a placebo indistinguishable from quetiapine (placebo quetiapine).

Those patients allocated to the placebo group also needed to be blind to group allocation. Because the participants in the quetiapine and rivastigmine treatment groups took an active drug and a placebo, it was necessary for the placebo group to take two placebo pills—that is, placebo rivastigmine and placebo quetiapine (c is false, and d is true).


Cite this as: BMJ 2011;343:d7294


  • Competing interests: None declared.


View Abstract

Log in

Log in through your institution


* For online subscription