The Thessaly test for detection of meniscal tears: validation of a new physical examination technique for primary care medicine.

OBJECTIVE: To assess the validity of a new clinical test (Thessaly) as a means of detecting meniscal tears of the knee by comparing arthroscopic findings to a clinical examination finding.

DESIGN: Retrospective cohort study.

SETTING: All preoperative examinations were performed in the Department of Orthopedic Surgery, a secondary care center, Dwight David Eisenhower Army Medical Center, Fort Gordon, Georgia.

PARTICIPANTS: 116 consecutive patients undergoing knee arthroscopy for suspected meniscal pathology.

INTERVENTION: The Thessaly test was performed during the preoperative examination as previously described. The clinician supports the patient by holding their outstretched hands while the patient stands flatfooted. The patient then rotates their knee and body, internally and externally, three times, keeping the knee flexed at 20 degrees. Patients with suspected meniscal tears will experience joint-line discomfort.

MAIN OUTCOME MEASURES: The Thessaly test performed at 20 degrees of flexion and arthroscopic surgical diagnosis.

RESULTS: Of the 66 patients with a positive Thessaly test, 65 had an arthroscopically verified meniscal tear. The Thessaly test revealed a sensitivity of 90.3%, specificity of 97.7%, positive predictive value of 98.5%, negative predictive value of 86.0%, likelihood ratio for a positive test of 39.3, likelihood ratio for a negative test of 0.09, and diagnostic accuracy of 88.8%. The resulting kappa coefficient revealed a statistically significant level of agreement (P < 0.001) for the surgical diagnosis and the Thessaly test.

CONCLUSIONS: The Thessaly test is a valid and reproducible physical examination technique for predicting meniscal tears. The Thessaly test shows promise as an easily performed maneuver that may have better diagnostic accuracy than traditional tests. However, this study was performed at a referral center; therefore, the diagnostic relevance cannot be appropriately applied to a more generalized population.