Arthroscopy in acute knee injuries: a prospective controlled trial.
Eighty-two patients took part in a prospective trial to assess the need and timing for arthroscopy in acute knee injuries. Patients with a suspected ligament injury, a suspected meniscal tear, or a haemarthrosis in the absence of fracture were included in the trial. Patients were entered randomly into two groups. In group I early arthroscopy was performed (within 48 h), in group II arthroscopy was performed within 21 days, but only when it was felt to be clinically indicated. Patients were assessed using the OAK knee score 1 year after the injury. There was no difference in the final result between the two groups. Early arthroscopy (within 48 h) showed no beneficial effect when compared with delayed arthroscopy (3 to 21 days after the injury). Indeed, an important diagnosis was missed at early arthroscopy in three cases, which led to delay in subsequent diagnosis and treatment. Arthroscopy raised the diagnostic accuracy from 61 per cent to 93 per cent. Considerable morbidity arises as a result of acute knee injuries. Clinical judgment will correctly define the need for arthroscopy in most cases at the time of injury, and arthroscopy is not necessary in all cases of acute haemarthrosis.