Ultrasound is more reliable than clinical tests to both confirm and rule out pathologies of the long head of the biceps: a systematic review and meta-analysis.

ABSTRACT

PURPOSE: To synthesize the literature and critically appraise current evidence to determine the most accurate physical examination (clinical test or ultrasound) to detect pathologies of the long head of the biceps tendon (LHBT). METHODS: A search was performed on PubMed, Embase®, and Cochrane. Studies that compared the diagnostic accuracy of clinical tests or ultrasound versus arthroscopy for the assessment of LHBT pathologies were included. RESULTS: Seven studies were included reporting on a total of 448 patients. One study on instability using ultrasound reported sensitivity and specificity of 1.00 and 0.96, respectively. Two studies on full-thickness tears using ultrasound reported sensitivity and specificity of 0.88-0.95 and 0.71-0.98, respectively. Four studies on partial-thickness LHBT tears reported sensitivity and specificity of 0.17-0.68 and 0.38-0.92, respectively, for clinical tests, versus 0.27-0.71 and 0.71-1.00, respectively, for ultrasound. Three studies on other LHBT pathologies reported sensitivity and specificity of 0.18-0.79 and 0.53-0.85, respectively, for clinical tests, versus 0.50 and 1.00, respectively, for ultrasound. CONCLUSION: To detect LHBT pathologies, sensitivity is high-to-excellent using ultrasound, and moderate using Neer's sign and Speed's test, while specificity is high-to-excellent also using ultrasound, as well as the belly press, lift-off and Kibler's tests. The clinical relevance of these findings is that clinical tests are only reliable either to confirm or rule out LHBT pathologies, whereas ultrasound is reliable both to confirm and rule out LHBT pathologies. While diagnostic imaging cannot substitute for patient history and physical examination, the reliability and accessibility of ultrasound render it practical for routine use, particularly if clinical tests render unclear or contradictory findings. LEVEL OF EVIDENCE Level III.