Letters

Case selection bias affected results

BMJ 1996; 312 doi: https://doi.org/10.1136/bmj.312.7039.1156a (Published 04 May 1996) Cite this as: BMJ 1996;312:1156
  1. Ian Campbell
  1. Statistical consultant I C Statistical Services, Merseyside L63 0NL

    EDITOR,—Charles R Gillis and David J Hole provide clear evidence that patients of specialist surgeons have a higher survival rate from breast cancer than those of non-specialist surgeons.1 However, they go on to conclude that this association is one of cause and effect, rather than due to differences in the cases referred to the two groups of surgeons (case selection bias). Their reason is that the difference in outcome persisted after adjustment for some prognostic factors; but statistical techniques can correct …

    View Full Text

    Sign in

    Log in through your institution

    Subscribe