Prospective Evaluation of the Clinical Prediction of Electrodiagnostic Results in Carpal Tunnel Syndrome

Lo, Julian K.; Finestone, Hillel M.; Gilbert, Ken

doi:10.1016/j.pmrj.2009.05.004

« Previous Next »PM&R
Volume 1, Issue 7 , Pages 612-619, July 2009

Prospective Evaluation of the Clinical Prediction of Electrodiagnostic Results in Carpal Tunnel Syndrome

Presented, in part, as a poster presentation at the Annual Assembly of the American Academy of Physical Medicine and Rehabilitation 2005.

Received 30 December 2008; accepted 12 May 2009.

Objective

To develop a unique clinical prediction point-score system for the diagnosis of carpal tunnel syndrome (CTS) and to prospectively evaluate this clinical rule to determine its ability to predict electrodiagnostic results.

Design

Retrospective case control to develop the clinical point-score system and then prospective diagnostic study with blind comparison to electrodiagnostic results.

Setting

Community-based electrodiagnostic laboratory.

Patients

A total of 348 subjects referred to the electrodiagnostic laboratory with a clinical suspicion of CTS to develop the clinical point-score system and then a different population of 278 subjects referred to the electrodiagnostic laboratory over a 1-year period with a clinical suspicion of CTS to prospectively evaluate the ability of the clinical rule to predict electrodiagnostic results.

Main Outcome Measures

Clinical point-score system results were compared with the electrodiagnostic results for CTS, which served as the primary outcome measure.

Results

A “weighted” point-score system was developed from 9 clinical variables including: gender, duration of symptoms, presence of wrist pain (negative predictor), presence of neck pain (negative predictor), nocturnal symptoms, presence of thenar atrophy, abductor pollicis brevis weakness, median sensory symptoms, and results of pinprick sensation examination. The prospective evaluation indicated that the clinical point-score rule, overall, performed moderately in predicting electrodiagnostic results (receiver operator characteristic curve = 0.80). The clinical point score was most predictive in a small subset of subjects with very high scores. However, in most subjects, the clinical point score could not accurately predict electrodiagnostic outcome in CTS, particularly in subjects with middle to low clinical point scores.

Conclusions

The developed clinical point-score rule was most predictive in subjects with high point score; however, the majority of the referrals to the electrodiagnostic laboratory indicated that the electrodiagnostic studies provided additional, independent information beyond what could be obtained by the clinical point score.