abstract
- Both hands of 39 patients who had symptoms of pain and/or numbness in one or both hands were tested by two hand therapists using the full kit of Semmes-Weinstein monofilaments (SWMFs). The SWMF thresholds were obtained for the thumb, the index finger, and the long and small fingers. These thresholds were classified as normal or abnormal based on four decision rules and two criterion measures. Decision rules were based on whether SWMF 2.83 or 3.22 would be the best limit of normality, and whether the small finger should be used for within-subject comparisons. The criterion measures were the highest threshold of all three radial digits and the highest threshold of the long finger alone. Intertherapist agreement on normality was fair to moderate (kappa = 0.22-0.51), varying according to decision rules and criterion measures. Reliability was higher when the additional comparison with the small finger was omitted. High accuracy in identifying cases of carpal tunnel syndrome (CTS) was possible, but accuracy varied moderately between testers and greatly according to decision rules and criterion measurements. The best overall accuracy (81%-82% sensitivity and 57%-86% specificity) was achieved when SWMF 2.83 was used as the upper limit of normality and the small finger was used for within-subject comparison, and when data from the long finger alone were used for decision making.