Abstract
In 228 knees submitted to arthrotomy following arthroscopy, there were eleven diagnostic errors in those knees examined under general anesthesia and five diagnostic errors in those knees examined under local anesthesia with a diagnostic accuracy of 91.1 and 95.2 per cent, respectively. Therefore, we concluded that accuracy is not diminished by performing arthroscopic examinations under local anesthesia on ambulatory out-patients. In 297 knees examined under local anesthesia, no complications were encountered. In 282 knees examined under general or spinal anesthesia, there was one pulmonary embolus and two significant hemarthroses. Except when there are specific overriding considerations, local anesthesia is preferable in the performance of diagnostic arthroscopy.