We reviewed 38 patients with arthroscopically-proven complete ACL tears operated on less than three weeks after injury. Their average age was 26 years (16 to 43), with 27 males and 11 females. All patients had had MRI preoperatively. The same examiner performed the Lachman, anterior-drawer, and pivot-shift tests without anaesthesia, recording the differences between the injured and the normal knees. The patients then had KT-1000 arthrometry by the same examiner at 15 lb (6.8 kg) and 20 lb (9 kg), with active displacement, and with maximum manual displacement. All scored positive for differences greater than 3 mm. The results of physical examination, KT-1000 tests and MRI were analysed using McNemar's test for matched data with continuity correction and a 95% confidence interval for each test. The sensitivity of the KT-1000 manual maximum test was 97% for 3 mm and 100% for 2 mm; this was the most useful arthrometric result. The Lachman test gave 95% sensitivity, providing the best simple clinical assessment. MRI was 97% sensitive for the detection of all ACL injuries, but this fell to 82% with respect to complete rupture. We found no significant differences between the results of the Lachman test and the KT-1000 manual maximum test, but these were significantly better than all other tests. In an era of cost-containment, we have shown that inexpensive tests in the clinic can allow treatment to proceed rapidly and in the most economical manner without the routine use of MRI.
