ACCURACY OF PHYSICAL EXAMINATION IN THE EVALUATION OF ROTATOR CUFF LESIONS DIAGNOSED BY MRI

Background: Shoulder pain due to rotator cuff injuries is a common presenting complaint in rheumatology units. There are a number of physical examination (PE) manoeuvres for the evaluation of specific tendons of the rotator cuff. MRI provides an accurate anatomic image of the different tendinous structures of the shoulder and it is considered the gold standard imaging technique in the diagnosis of rotator cuff pathology.Objectives: The aim of this study is to evaluate the accuracy of the PE in the diagnosis of rotator cuff injuries confirmed by MRI.Methods: Consecutive outpatients from Emergency and Rheumatology Units with a new-onset episode of shoulder pain were prospectively included in the study. After obtaining an informed consent, a pre-defined physical examination was done separately by two rheumatologists, including ten manoeuvres aimed to explore supraespinatus and infraespinatus tendons (Jobe, passive and resisted abduction, Patte, passive and resisted external rotation, Neer, Hawkins, Yocum and Impingement), and two manoeuvres that explored the subescapularis tendon (Gerber, resisted internal rotation). Within 3 days, a MRI of the painful shoulder was done (Siemens Avanto 1,5 TESLA Shoulder Array Coil). Sensitivity, specificity, positive and negative predictive values, and precision of each manoeuvre for the diagnosis of rotator cuff lesions were calculated.Results: Twenty-nine patients were included, with a mean age of 55±14 years and a median duration of symptoms of 83 days (range: 9-738). With regard to total rotator cuff tears, Jobe, Yocum and the impingement manoeuvre obtained a sensitivity of 100%, although none of them exceeded 45% of accuracy. As far as partial rotator cuff tears is concerned, Jobe manoeuvre was the most sensitive test, followed by passive abduction, Neer and Yocum manoeuvres. For the diagnosis of rotator cuff tendinosis, the most sensitive tests were the impingement and passive abduction manoeuvres. All of them achieved positive predictive values over 75% for the detection of tendinosis. The accuracy to detect rotator cuff calcifications was relatively low for all the tests, obtaining Jobe manoeuvre the highest precision (62%). Regarding partial subescapularis tear, resisted internal rotation resulted to be an accurate but not sensitive manoeuvre. Neither Gerber manoeuvre nor resisted internal rotation test obtained a good precision to detect tendon calcifications.Conclusion: Accuracy of PE manoeuvres in the evaluation of painful shoulder is good to localize the injured structure, but very low to define the nature of the lesion. The best manoeuvre to identify a rotator cuff tear is Jobe manoeuvre. The best tests to detect rotator cuff tendinosis are the impingement and the passive abduction manoeuvres. Resisted internal rotation is an accurate manoeuvre to detect partial subescapularis tears.Citation: Ann Rheum Dis, volume 67, supplement II, year 2008, page 559Session: Back pain, mechanical musculoskeletal problems, local soft tissue disorders