Journal of Experimental Orthopaedics (Jul 2024)

Fragility analysis and systematic review of patellar resurfacing versus non‐patellar resurfacing in total knee arthroplasty

  • Teja Polisetty,
  • Alexandra L. Hohmann,
  • Eric DiDomenico,
  • Jess H. Lonner

DOI
https://doi.org/10.1002/jeo2.12113
Journal volume & issue
Vol. 11, no. 3
pp. n/a – n/a

Abstract

Read online

Abstract Introduction Fragility analysis is a method of further characterising the robustness of statistical outcomes. This study evaluates the statistical fragility of randomised controlled trials (RCTs) comparing patellar resurfacing versus non‐patellar surfacing in total knee arthroplasty (TKA). Methods PubMed, MEDLINE and EMBASE were searched for RCTs comparing outcomes in TKA based on patellar resurfacing. Fragility index (FI) and reverse FI (collectively, “FI”) were calculated for dichotomous outcomes as the number of outcome reversals needed to change statistical significance. Fragility quotient (FQ) was calculated by dividing the FI by the sample size for that outcome. Median FI and FQ were calculated for each individual outcome and for the overall study. Subanalyses were performed to assess FI and FQ based on outcome type, statistical significance and loss to follow‐up. Results Twenty‐one RCTs were included in the analysis, capturing 3910 subjects. The overall median FI was 5.0 (interquartile range, [IQR] 4.0−6.0), and the overall median FQ was 0.048 (IQR 0.022−0.065). The outcome of anterior knee pain has a median FI of 6.0 (IQR 4.0−6.0) and a median FQ of 0.057 (IQR 0.025−0.065). Only five (7%) outcomes were significant. The loss to follow‐up was greater than the FI in 12 of 19 studies (63%) with available data. Conclusion RCTs comparing patellar resurfacing in TKAs show significant statistical fragility; a few outcome reversals can alter findings. The majority of outcomes were nonsignificant, indicating that the choice to resurface the patella may not affect most clinical outcomes; however, clinical conclusions are limited by the statistical fragility of the analysed outcomes. Larger RCTs for this comparison are necessary, and we suggest adding FI and FQ to RCT reports with p values to improve the interpretability of results. Level of Evidence Level II.

Keywords