No bias for developer publications and no difference between first-generation trochlear-resurfacing versus trochlear-cutting implants in 15,306 cases of patellofemoral joint arthroplasty

Abstract

Meta-analysis of Level IV case series.

Revision Rates After Total Ankle Replacement: A Comparison of Clinical Studies and Arthroplasty Registers

“…As already frequently applied in the literature, an exceedance of the factor 3 (deviation of 33%-300%) regarding the revision rate was seen as significant. 25,26,57,62,63 The level of significance was deliberately generous as clinical outcomes are highly dependent on individual factors not connected to the implant, such as patient selection, postoperative care, expertise, or general variability among hospitals. 62…”

“…This method has been commonly applied in the literature. [24][25][26]57,63,65 The exact formula for calculation is number of revisions divided by observed CYs multiplied by 100. A value of 1 revision per 100 CYs equals a 7-year revision rate of 7%, and a 10 year revision rate of 10%.…”

“…26,42 The usual approach used to draw conclusions on the patient population is derived from the information of clinical case studies. 63 However, national joint registries report that TARs contribute fewer than 1% of the total amount of arthroplasties performed. 2, 51,52 With a surgery being carried out that rarely and the fact that case studies are often conducted in centers of competence using a single type of ankle prosthesis, one might get a restricted view on the realistic situation.…”

“…2, 51,52 With a surgery being carried out that rarely and the fact that case studies are often conducted in centers of competence using a single type of ankle prosthesis, one might get a restricted view on the realistic situation. 63 Furthermore, publications exist, stating a more favorable outcome in studies conducted by implant developers. 38,40 This publication bias has raised doubts about the low revision rates in developer studies.…”

“…However, these studies were published a decade ago and this bias has not yet been observed in other joint prostheses in recent years. 24,26,63 Therefore, the primary aim of this study was to compare current revision rates of TARs published in clinical studies of the past decade to those described in arthroplasty registers. The secondary aim was to compare revision rates between sponsored/dependent and independent studies and between mobile bearing TARs (MB-TARs) and fixed bearing TARs (FB-TARs).…”