You are here

Article Text

Article menu
Download PDFPDF
EBM learning
Deceptive shifts in cancer stage distribution
  1. Jason L Oke1,
  2. H Gilbert Welch2
  1. 1 Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford, UK
  2. 2 Center for Surgery and Public Health, Department of Surgery, Brigham and Women's Hospital, Boston, Massachusetts, USA
  1. Correspondence to Dr Jason L Oke, Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford, OX2 6GG, UK; jason.oke{at}phc.ox.ac.uk

https://doi.org/10.1136/bmjebm-2023-112238

Statistics from Altmetric.com

    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.

    A change in cancer stage distribution, also known as a ‘stage shift’, is a common outcome measure in cancer epidemiology. A favourable stage shift—an increased proportion of early-stage cancer—is often the goal of early cancer detection efforts. An unfavourable stage shift—an increased proportion of late-stage cancer—is typically seen as evidence of the harm of diminished screening. Shifting cancer stage distributions, however, can be deceptive.

    The problem

    Consider the aspirational goal of the UK’s National Health Service (NHS): to increase the proportion of all cancers diagnosed at an early stage from approximately 50%–75%.1 The hope is that, by finding more people with early-stage cancer, the number presenting with late-stage cancer will fall. But even if only more early-stage cancers are detected, with no change in the number of late-stage cancers—a signature highly suggestive of cancer overdiagnosis—the proportion of all cancers diagnosed at an early stage will nonetheless rise.

    The problem is in the measure. Stage distribution is a proportion; thus, changes can be driven by changes in the numerator, denominator or both. That makes changes difficult to interpret.

    Alternatively, consider the interest in measuring the effect of COVID-19-related reductions to cancer screening. Here, the concern is whether diminished screening is associated with cancers being diagnosed later, leading to an increased proportion of late-stage cancer. But, of course the stage distribution will appear to be less favourable and a higher proportion will be late-stage cancer—not necessarily …

    View Full Text

    Footnotes

    • Twitter @Oke_jason

    • Contributors JO and HGW conceived the idea, drafted the first version of the manuscript and equally contributed to revisions. JO is the guarantor.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

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