Abstract
Objectives:
The glenoid track concept was developed for management of anterior shoulder instability originally via CT-based measurements of the glenoid track (GT) and Hill-Sachs interval (HSI). These methods are used to calculate whether the lesion is considered on, off, or near track and can guide surgical treatment. The methodology has been extended to MRI-based measurements. The objective of this study was to assess the reliability of glenoid track measurements by CT and MRI. We hypothesized that CT and MRI would have equivalent reliability.
Methods:
A cohort of 46 patients with Hill-Sachs lesions following anterior shoulder dislocations were collected. Inclusion criteria included both CT and MRI preoperative scans. Exclusion criteria included imaging of prior surgery, non-reduced dislocations, or those without Hill-Sachs lesions. Glenoid circumference using a best-fit circle, anterior bone loss, and Hill-Sachs interval were assessed by a sport medicine fellow and an orthopaedic surgeon. These measurements were used to determine the glenoid track which is defined as GT = 0.83*D-d, with D representing the diameter of the glenoid best-fit circle and d representing the width of anterior glenoid bone loss. Distance to dislocation (DTD), defined as DTD = GT-HSI, was determined and thus whether the lesion was on track (DTD ≥ 10 mm), near track (0 mm < DTD < 10 mm), or off track (DTD ≤ 0 mm). Statistical analysis was carried out with Cohens Kappa and interclass correlation coefficient (ICC).
Results:
A total of 44 patients met the inclusion and exclusion criteria. The mean age of patients included was 25.5 years old (range 16 – 40). Assessment of whether the lesion was on, near, or off track via MRI showed moderate agreement between raters (70%, k = 0.47), while assessment via CT showed fair agreement between raters (56%, k=0.28). On, near, and off-track agreement between CT and MRI ranged from fair to almost perfect (59-86%, k = 0.29 - 0.72). The intraclass correlation coefficient for glenoid track width via MRI was moderate (ICC = 0.51), and via CT was poor (ICC = 0.22). The intraclass correlation coefficient for Hill-Sachs interval via MRI was moderate (ICC = 0.70), and via CT was moderate (ICC = 0.53). The intraclass correlation coefficient for distance to dislocation via MRI was moderate (ICC = 0.57), and via CT was moderate (ICC = 0.50).
Conclusions:
CT and MRI do not demonstrate similar reliability regarding measurements of the glenoid track. Based on the results of our study, MRI showed a higher interrater reliability across multiple measurements. This may be due to the lower number of slices per MRI for raters to assess. The larger number of slices available for analysis in CT may lead to lower reliability between raters. The additional measurement of HSI may have contributed to decreased reliability when compared to current literature on measurement reliability of anterior glenoid bone loss. Poor reliability between CT measurements of the glenoid track may potentially affect clinical decisions. While MRI demonstrated better interrater reliability, the discrepancy between CT and MRI in the same individual implies that the modalities are not equivalent for calculating the glenoid track.