Article Text

Download PDFPDF
General medicine
Assessing assumptions for statistical analyses in randomised clinical trials
  1. Emil Eik Nielsen1,2,
  2. Anders Kehlet Nørskov1,
  3. Theis Lange3,4,
  4. Lehana Thabane5,
  5. Jørn Wetterslev1,6,
  6. Jan Beyersmann7,
  7. Jacobo de Uña-Álvarez8,
  8. Valter Torri9,
  9. Laurent Billot10,
  10. Hein Putter11,
  11. Per Winkel1,
  12. Christian Gluud1,
  13. Janus Christian Jakobsen1,12
  1. 1 Copenhagen Trial Unit, Centre for Clinical Intervention Research, Rigshospitalet, Kobenhavn, Denmark
  2. 2 Department of Internal Medicine, Holbaek Sygehus, Holbaek, Denmark
  3. 3 Section of Biostatistics, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark
  4. 4 Center for Statistical Science, Peking University, Beijing, China
  5. 5 Health Research Methods, Evidence, and Impact, Mcmaster University, Hamilton, Canada
  6. 6 Centre for Research in Intensive Care (CRIC), Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark
  7. 7 Institute of Statistics, University of Ulm, Ulm, Germany
  8. 8 Department of Statistics and Operations Research and Biomedical Research Center (CINBIO), Universidad de Vigo, Vigo, Spain
  9. 9 IRCCS - Istituto Di Ricerche Farmacologiche "Mario Negri", Milano, Italy
  10. 10 The George Institute for Global Health, University of New South Wales, Sydney, Australia
  11. 11 Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, The Netherlands
  12. 12 Department of Regional Health Research, Faculty of Health Sciences, University of Southern Denmark, Denmark
  1. Correspondence to Dr Emil Eik Nielsen, Copenhagen Trial Unit, Centre for Clinical Intervention Research, Rigshospitalet, Kobenhavn DK-2100, Denmark; emil.eik.nielsen{at}gmail.com

Abstract

In order to ensure the validity of results of randomised clinical trials and under some circumstances to optimise statistical power, most statistical methods require validation of underlying statistical assumptions. The present paper describes how trialists in major medical journals report tests of underlying statistical assumptions when analysing results of randomised clinical trials. We also consider possible solutions how to improve current practice by adequate reporting of tests of underlying statistical assumptions. We conclude that there is a need to reach consensus on which underlying assumptions should be assessed, how these underlying assumptions should be assessed and what should be done if the underlying assumptions are violated.

View Full Text

Statistics from Altmetric.com

Footnotes

  • Contributors All authors developed the framework and participated in writing the article. EEN is the guarantor of the article. All remaining authors are selected statistical experts, methodologists and trialists from different parts of the world.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Patient consent for publication Not required.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.