The purpose of this study was to assess the accuracy of a modified version of the pivot shift test in detecting ruptures of the anterior cruciate (ACL) ligament.

Methods: Two groups of patients aged 18 to 50 years were recruited from operating theatre lists examined at a check and consent clinic. One of the groups had ACL deficient knees and the control group had intact ACL, later proven at arthroscopy. A total of 48 independent examinations, 26 with ACL rupture and 22 without, were undertaken by a consultant or registrar proficient in the modified version of the pivot shift test. At the start of the consultation the examiner, who was blinded to the operation that the patient was listed for and to the patient’s history, performed the modified pivot shift test only. The result of the test was reported as positive or negative for ACL rupture to the study co-ordinator before the examiner continued with the consultation.

Results: Of the 26 examinations of ACL deficient knees, 22 were reported as positive and four negative for ACL rupture using the pivot shift test only. Of the 22 examinations of ACL intact knees one was reported as positive for ACL rupture and 21 as negative. This gives a specificity and sensitivity of the modified version of the pivot shift test as 95.5% (95% CI 75.1–99.8%) and 84.6% (95%CI 64.3–95.0%) respectively.

Conclusion: The modified pivot shift test is an accurate test for the detection of ACL rupture. Our modified pivot shift test compares favourably with data from previous accuracy studies of previously described versions of the pivot shift test. We now plan to assess the accuracy of the modified pivot shift test when carried out by medical students and junior doctors.