[Reliability of neuroconduction studies in carpal tunnel syndrome]

To determine the reliability of neuroconduction studies by comparing two observers and detecting minimum changes when diagnosing carpal tunnel syndrome.

Sixty-nine patients referred for electrophysiological study due to suspected carpal tunnel syndrome were studied. The patients underwent two examinations, performed by two evaluators, on two different days. Sensory and motor latencies of the median and ulnar nerves were evaluated; all were classified as negative, incipient, mild, moderate, severe or extreme. Relative interval variation, intraclass correlation coefficient, kappa index, limit of agreement and minimum detectable change were estimated.

The relative variation of motor nerve latency of the median nerve was -6.8% to 15.9%, with intraclass correlation coefficient of 0.98 for the difference of median-ulnar nerve latency. The minimum detected change was 0.4ms, while the relative interval variation of sensory latency of the median nerve was -5.0% to 11%, with intraclass correlation coefficient of 0.95 for difference with the ulnar nerve. The minimum detectable change was 0.2ms. Electrophysiological classification agreed in 93% of the cases, with a kappa index of 0.89.

Sensory and motor latencies of the median nerve, as well as the difference between them and the ulnar nerve, are reliable measures. The minimum detectable change obtained in our study helps clinicians to establish whether changes in latencies in consecutive or post-treatment studies are significant.