In the past 10-15 years shoulder surgeons have become much more aware of the significance of glenoid bone loss in the setting of shoulder instability. Burkhart and DeBeer et al demonstrated the association between much higher failure rates for arthroscopic shoulder labral repair in the setting of glenoid bone loss (1). This has lead many surgeons to consider open labral repair or bone augmentation for shoulder instability when glenoid bone loss is present (2,3). What has been less clear is determining how much glenoid bone loss is significant and what is the best modality and measurement technique for evaluation of bone loss. Multiple measuring techniques have been developed (4, 5). The majority of techniques require a 3-D reconstruction of the glenoid with humeral head substraction for accurate measurement. The literature has shown CT scan to be more accurate than MRI for assessment of bone (6, 7). Therefore, most surgeons require both an MRI for evaluation of the torn labrum and then a subsequent CT scan if concerned for glenoid bone loss. In addition, because of the inherent challenge of trying to measuring something that isn't present, some suggest imaging of the contralateral shoulder when determining bone loss (8). Recently, Owens et al developed a predictive formula* for determining glenoid width based on the MRI scans of more than 1200 shoulders (9). Although, a similar predictive formula** has been developed for CT scans, there has yet to be a comparison between MRI and CT scan for accuracy and reliability of measuring bone loss when using this novel predictive formula (10). If MRI can demonstrate accuracy and reliability similar to CT scan this will save patients the time, cost, and radiation exposure of an undergoing an additional CT scan. * Male: Glenoid width = (1/3 height) + 15mm; Female: Glenoid width = (1/3 height) + 13mm ** Male: Glenoid width = (2/3 height) + 5mm; Female: Glenoid width = (2/3 height) + 3mm

1) Can MRI accurately assess glenoid bone loss using a validated predictive formula in comparison to 3-D CT scan

2) Does MRI have high inter- and intrareader reliability for assessment of glenoid bone loss when using a validated predictive formula?

We aim to retrospectively evaluate the use of MR imagining in conjunction with a validated predictive formula for assessment of glenoid bone loss in comparison to the gold standard of CT with 3-D reconstruction. If MR imaging with use of a predictive formula is accurate with good inter- and intrareader reliability it may replace the need for subsequent CT with 3-D reconstruction when suspicious of bone loss. This will prevent the additional associated radiation exposure to the patient of a CT scan as well as prevent the associated increased cost and time.

We plan to review the PACS web based Centricity radiology system for all patients having undergone a CT of the shoulder with 3-D reconstruction and associated MRI of the ipsilateral shoulder with 90 days. Those patients with implants or hardware in the area of the glenoid or humeral head will be excluded due to concern for artifact affecting the quality of the imaging study. Glenoid width and glenoid bone loss will be calculated using a validated predictive formula specific to MRI* or CT**. The Student test will be used to test for difference in mean glenoid width measured with CT and MR imaging. The Pearson correlation will be used to examine the correlation between percentage of glenoid bone loss measured with CT and MR imaging. Inter- and intrareader reliability for the MR imaging and CT imaging assessment of glenoid bone loss will be determined with the interclass coefficient, with R <0.40 indicative of poor agreement, R>0.40 fair agreement, and R>0.75 excellent agreement. For all test, P<0.05 will be regarded as indicative of statistically significant difference

* Male: Glenoid width = (1/3 height) + 15mm; Female: Glenoid width = (1/3 height) + 13mm

** Male: Glenoid width = (2/3 height) + 5mm; Female: Glenoid width = (2/3 height) + 3mm

Inclusion criteria:

1) CT scan of shoulder with 3-D reconstruction and humeral head subtraction

2) MRI of ipsilateral shoulder with 90 days of the CT scan

3) Age >18 and <60

Exclusion criteria include:

1) Implants or hardware in the area of the glenoid or humeral head

2) Age <18 or >60

SPSS, Version 17

American Academy of Orthopedic Surgeons (AAOS) Annual Meeting Orthopaedic weekly grand rounds; orthopaedic daily morning didactics and radiology conferences; shoulder/upper extremity weekly conferences; sports medicine weekly conferences.