Hill-Sachs Lesion Classification by the Glenoid Track Paradigm in Shoulder Instability: Poor Agreement Between 3-Dimensional Computed Tomographic and Arthroscopic Methods

Abstract

The purpose of this study was to determine the amount of agreement between preoperative 3-dimensional computed tomographic (3D-CT) and intraoperative arthroscopic classification of Hill-Sachs lesions (HSLs) according to the glenoid track (GT) paradigm. Records for patients treated surgically for anterior shoulder instability from a single surgeon's practice from August 2013 until March 2016 were retrospectively reviewed. Inclusion criteria were presence of an HSL, < 25% glenoid bone loss, bilateral 3D-CT, and arthroscopically recorded bone loss measurements. Records for patients with chronic dislocations or prior operations were excluded. Calculations by 3D-CT and arthroscopy were performed as follows: Hill-Sachs interval (HSI) was the distance from rotator cuff insertion to medial edge of the HSL; GT was 83% of the normal glenoid width minus any glenoid defect; on-track was HSI less than GT; off-track was HSI greater than GT. Sixteen shoulders with HSL status determined as on- or off-track demonstrated agreement between the 2 methods in 10 of 16 cases (63%, Cohen's κ= 0.16). All 6 cases with disagreement were calculated as on-track by 3D-CT and off-track by arthroscopic measurement. The GT was larger as determined by 3D-CT measurement (22 ± 1mm [21-24]) compared with arthroscopy (18 ± 1mm [17-20], P= .002). Preoperative 3D-CT showed slight agreement compared with intraoperative arthroscopic measurements in classifying HSL as on-track versus off-track in the GT paradigm; larger GT size by 3D-CT versus arthroscopy accounted for all discrepancies. Determination of off-track status based on preoperative 3D-CT versus determination with the arthroscopic method would result in fewer HSLs treated with remplissage if the GT treatment paradigm were followed. Surgeons using the GT paradigm to determine treatment of HSL by remplissage should recognize the potential for discordance between arthroscopic and radiographic measurements.