Related Experiment Video
Updated: Apr 28, 2026
Novel Diagnostics in Revision Arthroplasty: Implant Sonication and Multiplex Polymerase Chain Reaction
Published on: December 3, 2017
[Early detection of systematic defects by endoprostheses registries]
T R Liebs1, O Melsheimer, J Hassenpflug
1Orthopädische Klinik, Universitätsklinikum Schleswig-Holstein, Campus Kiel, Michaelisstr. 1, 24105, Kiel, Deutschland, Liebs@Liebs.eu.
Background:
The burden of revision of the 210,000 total hip and the 146,000 total knee arthroplasties performed in Germany every year is 10.9% and 7.9%, respectively. Since the probability of revisions could be cut in half by the implementation of a national joint registry in Sweden, several other countries have also established national joint registries. Nevertheless, these registries were not able to detect the failure of some implant systems, which resulted in thousands of patients who had received implants being recalled later on.
Objectives:
We aimed to identify the methods that are currently used for statistical analyses in joint registries and which could be used to implement an early warning system.
Materials And Methods:
We analyzed the available reports of national joint registries regarding the methods used for the early detection of implants that perform less efficiently than expected. The methods identified are described and evaluated if they are suitable for being used in a national joint registry.
Results:
While most of the identified national joint registries use the Kaplan-Meier method for analyzing implant survival, there was little agreement regarding the methods used for the early detection of implants that perform less efficiently than expected. Methods identified included revision rate per 100 component years, variants of statistical process control (SPC), cumulative sums (CUSUM) and funnel plots.
Conclusion:
Currently, there are no standardized methods for the early detection of inferior implants that are used by the established national joint registries. Most of the statistical methods that could be identified were either not used at all or only in a very limited number of registries. The value of these methods still needs to be established. It appears probable that the statistical methods for early detection have to be developed further in order to identify a possible inferior performance of implants at an earlier point in time. This is necessary in order to avoid that thousands of patients are again treated with implants which are associated with an increased revision rate.
