Diagnostic utility of ultrasonography versus nerve conduction studies in mild carpal tunnel syndrome

OBJECTIVE: To prospectively compare high-resolution ultrasonography (US) and nerve conduction velocity (NCV) in clinically diagnosed mild carpal tunnel syndrome (CTS). METHODS: Eighty-five patients (70 women and 15 men, mean age 46.8 years) reported symptoms compatible with classic/probable CTS. The protocol included NCV of the median and ulnar nerves (distal motor latency DML], sensory conduction velocity SCV] from the third M3 SCV] and fourth fingers M4 SCV] to the wrist for the median nerve); electrophysiologic severity scale; self-administered Levine/Boston questionnaire (BQ); and cross-sectional area (CSA) measurement of the nerve at the tunnel inlet (CSA-I), at the middle (CSA-M), and at the outlet (CSA-O). Relationship between age, body mass index, duration of symptoms, CSAs, NCV, electrophysiologic severity scale, and BQ scores was calculated. Concordance between CSAs and NCV, sensitivity of NCV and US was also evaluated. RESULTS: The mean values of CSA-I, CSA-M, and CSA-O were 10.3, 9.8, and 8.7 mm2, respectively. Relationships were found between CSA-I and M3 SCV (r = -0.45), M4 SCV (r = -0.56), and median nerve DML (r = 0.29). Anomalous CSA-I, CSA-M, and CSA-O were found in 48, 25, and 26 patients, respectively; 55 (64.7%) had > or =1 abnormal CSA. NCV abnormalities were found in 67%. The sensitivity increased to 76.5% if US and NCV were considered together. The highest concordance to detect absence/presence of abnormalities was between CSA-I and NCV (77.6%; kappa = 0.52). CONCLUSION: In mild cases of CTS, US did not detect more anomalies than NCV and vice versa, and no anomalies were detected with either diagnostic instrument in 23.5% of mild cases.