Biceps sheath fluid on shoulder ultrasound as a predictor of rotator cuff tear: analysis of a consecutive cohort.

BACKGROUND: Ultrasound provides evaluation of rotator cuff disease with accuracy comparable to that of magnetic resonance imaging. Fluid in the sheath of the long head of the biceps tendon (LHB), identified on ultrasound scan, has been associated with disease of the rotator cuff, LHB, and glenohumeral joint. Prior literature has compared ultrasound findings only with arthrography, and results have been conflicting. Arthroscopy remains the reference standard in assessing accuracy of imaging modalities. We present the first study investigating the significance of fluid in the LHB on ultrasound in predicting subsequent rotator cuff disease identified on arthroscopy.

METHODS: Records were reviewed of 175 patients undergoing ultrasound and subsequent arthroscopy under 1 shoulder surgeon. Experienced musculoskeletal radiologists and sonographers performed ultrasound. Ultrasound examination and operating records were collected and analyzed. Data were analyzed using descriptive statistics, correlation, and logistic regression modeling.

RESULTS: Highly significant correlation (P < .001; ρ = 0.354) was found between fluid in the LHB sheath and rotator cuff tears on arthroscopy. Statistically significant but weak correlation (P < .05; rho = 0.187) was found between fluid in the LHB sheath and both biceps tendon disease and glenohumeral joint disease. Fluid around the LHB was shown to increase the likelihood of having rotator cuff tear (odds ratio, 2.641; 95% confidence interval, 1.229-5.674) and biceps tendon disease (odds ratio, 2.698; 95% confidence interval, 1.216-5.987).

CONCLUSION: This is the first report identifying significant correlation between fluid in the LHB sheath identified on ultrasound and subsequent rotator cuff disease identified at arthroscopy. We recommend routine reporting of fluid in the LHB sheath as it is likely to improve the accuracy of detecting rotator cuff and biceps tendon diseases.