Radiographic Assessment of a Medullary Total Ankle Prosthesis: A Test of Agreement and Reliability

Total ankle replacement (TAR) is a viable alternative to ankle fusion in certain patients with end-stage ankle arthritis. Despite the importance of understanding alignment and movement of the prosthesis, there is no standardized radiographic method for evaluating the position and movement of the INBONE 2 prosthesis. The aims of this study were to describe a radiographic measurement protocol for INBONE 2 for clinical practice and research while determining the interobserver and intraobserver reliability using standard weightbearing radiographs. Fifteen patients were randomly selected with operative dates from January 2011 to January 2014 who underwent primary TAR using the INBONE 2 prosthesis. Most recent preoperative and first postoperative weightbearing anteroposterior and lateral radiographs were pulled and deidentified. Three foot and ankle surgeons blinded from the patient selection and deidentification, measured the described measurements on separate occasions. Intraobserver reliability: surgeon 1 had acceptable reliability for 9 of 13 continuous radiographic measurements (69.2%), surgeon 2 had acceptable reliability for 8 of 13 measurements (61.5%), and surgeon 3 had acceptable reliability for 12 of 13 measurements (92.3%). Interobserver reliability: among the first measurements, 6 of 13 continuous radiographic measurements (46.2%) had acceptable reliability. Among the second measurements, 7 of 13 measurements (53.8%) had acceptable reliability. Among the first and second measurements combined, 7 of 13 measurements (53.8%) had acceptable reliability. This study promotes the need for meticulous evaluation of annual radiographic findings following TAR in an effort to avoid catastrophic failure and represents moderate agreement can be obtained by employing the proposed measurements for surveillance of INBONE 2 TAR at annual postoperative visits. Measurements on the anteroposterior radiograph appear to demonstrate more consistent results for surveillance than lateral measurements. The intraobserver reliability results were somewhat superior to the interobserver reliability, implying more relevance for a single surgeon applying these measurements annually for postoperative surveillance.

Diagnostic, Level III.