Similar revision rates in clinical studies and arthroplasty registers and no bias for developer publications in unicompartmental knee arthroplasty.

Our aim was to assess the outcome with respect to cumulative revision rates of unicompartmental knee arthroplasty (UKA) by comparing published literature and arthroplasty registry data. Our hypothesis was that there is a superior outcome of UKA described in dependent clinical studies compared to independent studies or arthroplasty registers. A systematic review of all clinical studies on UKA in the past decade was conducted with the main endpoint revision rate. Revision rate was calculated as "revision per 100 component years (CY)". The respective data were analysed with regard to a potential difference of the percentage of performed revision surgeries as described in dependent and independent clinical studies. Clinical data were further compared to arthroplasty registers in a systematic search algorithm. In total, 48 study cohorts fulfilled our inclusion criteria and revealed 1.11 revisions per 100 CY. This corresponds to a revision rate of 11.1% after 10 years. No deviations with regard to revision rates for UKA among dependent and independent clinical literature were detected. Data from four arthroplasty registers showed lower survival rates after 10 years compared to published literature without being significant. The outcomes of UKA in dependent and independent clinical studies do not differ significantly and are in line with arthroplasty register datasets. We cannot confirm biased results and the authors recommend the use of UKAs in properly selected patients by experts in their field.