Carpal tunnel syndrome--course and prognosis.

In a retrospective study 157 patients with electrophysiologically proven carpal tunnel syndrome (CTS) were followed up. Of the 85 operatively treated patients, 86% showed clear improvement or cure, as did 32% of the patients not operated upon. Multivariate analysis of variance (MANOVA) and linear discriminant analyses indicated atrophy and length of history as effective predictors of clear improvement. The optimized discriminant point was ascertained in the receiver operator characteristics (ROC) curve. A predictive value of positive test of 0.89 was found for the course form "not cured" and a predictive value of negative test of 0.75 for the course form "cured" with a prevalence of one-third for the good course form. The prognosis was bad if atrophy occurred or the history exceeded 7 months. The determination of distal motor latency is indispensable for diagnosing CTS. For prognosis, however, length of history and clinical findings have been more helpful.