Consistency of commonly used orthopedic special tests of the shoulder when used with the McKenzie system of mechanical diagnosis and therapy

Background: Shoulder Orthopedic Special Tests (OSTs) are used to assist with diagnosis in shoulder disorders. Issues with reliability and validity exist, making their interpretation challenging. Exploring OST results on repeated testing within Mechanical Diagnosis and Therapy (MDT) shoulder classifications may offer insight into the poor performance of these tests. Objectives: To investigate in patients with shoulder complaints, whether MDT classifications affect the agreement of OST results over the course of treatment. Methods: An international group of MDT clinicians recruited 105 patients with shoulder problems. Three commonly used OSTs (Empty Can, Hawkins-Kennedy, and Speed's tests) were utilized. Results of the OSTs were collected at sessions 1, 3, 5 and 8, or at discharge from an MDT classification-based treatment. The Kappa statistic was utilized to determine the agreement of the OST results over time for each of the MDT classifications. Results: The overall Kappa values for Empty Can, Hawkins-Kennedy and Speed's tests were 0.28 (SE = 0.07), 0.28 (SE = 0.07) and 0.29 (SE = 0.07), respectively. The highest level of agreement was for Articular Dysfunction for the Empty Can test (0.84, SE = 0.19). For shoulder Derangements, there was no agreement for any of the OSTs (P values > 0.05). Conclusion: The lack of agreement when the OSTs were consecutively tested in the presence of the MDT Derangement classification contrasted with the other MDT classifications. The presence of Derangement was responsible for reducing the overall agreement of commonly used OSTs and may explain the poor consistency for OSTs.