The rotator cuff functional index

BACKGROUND: Despite improvements in the clinical assessment of rotator cuff injuries, shortcomings exist in diagnosing rotator cuff tears. PURPOSE: To formulate a clinical rotator cuff functional index for predicting rotator cuff tears based on handheld dynamometer measurements for shoulder strength testing and to assess its ability to diagnose a rotator cuff tear. STUDY DESIGN: Cohort study (diagnosis); Level of evidence, 2. METHODS: Preoperative handheld dynamometer measurements, including supraspinatus, external rotation, internal rotation, adduction, and lift-off force, were compared in 100 patients with full-thickness rotator cuff tears (> 1.0 cm) and 100 patients with no rotator cuff injury (confirmed by arthroscopy). Using multiple logistic regression analysis, the rotator cuff functional index and index for rotator cuff tear size were formulated in which 100 represented excellent rotator cuff function and less than 0 indicated a full-thickness supraspinatus tear. Each patient's index was calculated to determine its clinical value in predicting a rotator cuff tear; the Pearson correlation coefficient was established for the rotator cuff functional index for rotator cuff size. RESULTS: Two handheld dynamometer measurements predicted a rotator cuff tear (P < .001, r = 0.53), with rotator cuff functional index = 1.27 * supraspinatus (N) - 0.72 * adduction (N) - 11.6, but not tear size (r = 0.38). In predicting a rotator cuff tear (rotator cuff functional index < 0 = rotator cuff tear; rotator cuff functional index > 0 = no tear), the rotator cuff functional index had 83% sensitivity, 79% specificity, 80% positive predictive value, 82% negative predictive value, likelihood ratio of 4.0 for a positive test result and 0.2 for a negative test result and an accuracy of 81%. CONCLUSION: The 2 most useful tests for determining a rotator cuff tear with a handheld dynamometer were supraspinatus force and adduction force (as negative control). A functional index based on these measurements was able to rule out a rotator cuff tear and predict the presence but not the size of the tear.