Bias and unicompartmental knee arthroplasty

D.W. Murray, A.D. Liddle, Andy Judge, H. Pandit

Research output: Contribution to journalArticle (Academic Journal)peer-review

27 Citations (Scopus)
206 Downloads (Pure)


We recently published a paper comparing the incidence of adverse outcomes after unicompartmental and total knee arthroplasty (UKA and TKA). The conclusion of this study, which was in favour of UKA, was dismissed as "biased" in a review in Bone & Joint 360 Although this study is one of the least biased comparisons of UKA and TKA, this episode highlights the biases that exist both for and against UKA. In this review, we explore the different types of bias, particularly selection, reporting and measurement. We conclude that comparisons between UKA and TKA are open to bias. These biases can be so marked, particularly in comparisons based just on national registry data, that the conclusions can be misleading. For a fair comparison, data from randomised studies or well-matched, prospective observational cohort studies, which include registry data, are required, and multiple outcome measures should be used. The data of this type that already exist suggest that if UKA is used appropriately, compared with TKA, its advantages outweigh its disadvantages. Cite this article: Bone Joint J 2017;99-B:12-15.
Original languageEnglish
Pages (from-to)12-15
Number of pages4
JournalBone and Joint Journal
Issue number1
Early online date1 Jan 2017
Publication statusPublished - Jan 2017

Bibliographical note

M1 - Print


  • Humans
  • Postoperative Complications
  • Arthroplasty, Replacement, Knee
  • Registries
  • Bias (Epidemiology)
  • Meta-Analysis as Topic


Dive into the research topics of 'Bias and unicompartmental knee arthroplasty'. Together they form a unique fingerprint.

Cite this