Severity-correlated enlargement of the median nerve occurs in idiopathic carpal tunnel syndrome (CTS). We determined whether measurement of the nerve cross-sectional area was useful in making the diagnosis of CTS. In 414 wrists of 275 patients with clinically diagnosed idiopathic CTS and 408 wrists of 408 controls, we made ultrasonographic measurements of the nerve area at the distal (distal edge of the flexor retinaculum), mid (hook of the hamate), and proximal carpal tunnel (wrist crease). Criteria based on the area at a single level yielded sensitivities of 43-57% and specificities of 96-97%. Use of the mean carpal nerve area (average of the areas at the three levels) improved the results (sensitivity, 67%; specificity, 97%), which was as sensitive as the nerve conduction studies (NCS). In the NCS, we obtained sensitivities of 66% for the distal motor nerve latency and 67% for the distal sensory nerve latency, with specificities of 97% for each. A combination of the mean carpal nerve area and NCS criteria yielded a sensitivity of 84% and a specificity of 94%. The clinical implications of these findings for the diagnosis of CTS are discussed.
Copyright 2002 Wiley Periodicals, Inc. Muscle Nerve 26: 798-803, 2002