Patient relevant outcomes of unicompartmental versus total knee replacement: systematic review and meta-analysis.

In this research article by Wilson and colleagues (BMJ 2019;364:l352, doi:10.1136/bmj.l352, published 21 February), coauthor Stephanie Smith should have been listed as a postdoctoral research fellow rather than research student. The search strategy was also missing, and incorrectly cited to appendix 1a rather than appendix 1.

Furthermore, several studies were included twice by mistake in the same meta-analysis where more than one outcome measure was used to measure pain and function (figure 5, and supplementary figures 10-12), which resulted in the double counting of some participants. The authors have since removed the duplicate study entries, and if more than one outcome measure was included for the same study, the authors included the outcome measure that had the greatest internal and external validity. The effect estimates change slightly and the confidence intervals are slightly wider, but there is no change in the overall direction of the effect or to the conclusions of the systematic review.

The article text, table 1, figure 5, web appendix 2, and online PDF have since been corrected to reflect these changes.