Endgames Statistical Question

Allocation concealment

BMJ 2010; 340 doi: https://doi.org/10.1136/bmj.c537 (Published 04 February 2010) Cite this as: BMJ 2010;340:c537
  1. Philip Sedgwick, senior lecturer in medical statistics
  1. 1Centre for Medical and Healthcare Education, St George’s, University of London, Tooting, London SW17 0RE
  1. p.sedgwick{at}sgul.ac.uk

    A randomised controlled trial investigated whether an exercise programme increased activity levels in women. Intervention consisted of brief physical activity with regular support that was led by nurses. Control participants received usual care. The order in which participants were allocated to treatment was contained in sequentially numbered opaque envelopes. A researcher not involved in recruitment generated the sequence by computer. Following written informed consent and baseline measurements, treatment allocation was revealed by a nurse not involved in future assessments. Participants were assessed at follow-up by nurses unaware of group allocation; assessors and participants were asked not to discuss group allocation.1

    Which of the following, if any, are true?

    • a) Allocation concealment was achieved

    • b) The trial was single blind

    • c) The trial was double blind

    • d) Selection bias was minimised

    Answers

    a, b, and d are true; c is false.

    Women were allocated to treatment using a random sequence constructed before recruitment. This sequence was concealed from the nurse who recruited the participants, thereby ensuring allocation concealment (answer a). To ensure allocation concealment, the allocation sequence was constructed by someone not involved in recruitment. The allocation sequence was then placed in sequentially numbered opaque envelopes. Treatment allocation was revealed only after women had been recruited to the trial.

    Allocation concealment was essential to minimise selection bias (answer d). If the recruiting nurse knew the allocation sequence they may have influenced, unconsciously or otherwise, those women recruited to the trial. The nurse may have believed that some women would not have accepted, or been unsuitable for, their allocated treatment. This would have influenced which women were allocated to the intervention or control and caused systematic differences in unmeasured or measured baseline characteristics between groups. Therefore, any differences between treatment groups at the end of the study may not have been the result of differences in treatment but confounding at baseline.

    It was not possible to blind participants to their allocated treatment although the nurse assessor was not told. Therefore, the trial was single blind (answer b is true, and c is false). Blinding, discussed in a previous question,2 should be distinguished from allocation concealment. Allocation concealment involves concealing the allocation sequence before participant recruitment to minimise selection bias. Blinding is the masking of allocated treatment after recruitment, which minimises response and assessment biases.3

    Notes

    Cite this as: BMJ 2010;340:c537

    Footnotes

    • Competing interests: None declared.

    References

    View Abstract

    Sign in

    Log in through your institution

    Subscribe