Letters

Clinical trials and rare diseases

BMJ 1996; 312 doi: https://doi.org/10.1136/bmj.312.7035.910a (Published 06 April 1996) Cite this as: BMJ 1996;312:910

Statistical results should be expressed in different ways, depending on circumstances

  1. C K Connolly
  1. Consultant physician Darlington Memorial Hospital NHS Trust, Darlington DL3 6HX

    EDITOR,—Since the BMJ requested that, when possible, the statistical analysis of results should give confidence intervals, the relevance of many studies has become clearer. In recommending a bayesian approach to clinical trials in rare diseases Richard J Lilford and colleagues point out that power …

    View Full Text

    Sign in

    Log in through your institution

    Subscribe