Objective: To investigate the diagnostic value of sonography (SG) and magnetic resonance imaging (MRI) in the assessment of full-thickness rotator cuff tears (RCTs).
Methods: Twenty-one consecutive, otherwise healthy patients with noninflammatory unilateral chronic (>3 months) shoulder complaints due to a possible full-thickness RCT were studied (9 women and 12 men, mean +/- SD age 56 +/- 12). According to standardized procedures, SG was performed by both a radiologist and a rheumatologist, and MRI was evaluated by 2 radiologists. All assessors were blinded to the patient's diagnosis. Within 3 weeks after SG and MRI, arthroscopy was performed. SG, MRI, and arthroscopy results were scored as negative or positive for the presence of a full-thickness RCT. The result of surgical inspection was used as the "gold standard."
Results: For full-thickness RCTs, the sensitivity was 0.81 for SG and 0.81 for MRI. The specificity was 0.94 for SG and 0.88 for MRI. The positive predictive value was 0.96 for SG and 0.91 for MRI. The negative predictive value was 0.77 for SG and 0.74 for MRI. Accuracy was 0.86 for SG and 0.83 for MRI.
Conclusion: Full-thickness RCTs can be identified accurately by both SG and MRI. Because of its low cost and because it can be performed in the rheumatology unit, SG seems to be a promising diagnostic tool for use by the rheumatologist.