ABSTRACT
OBJECTIVES: To evaluate the sensitivity and specificity of the latency difference (DLat) between ulnar and median nerves of the arm after stimulation at the wrist; one of the easiest techniques proposed for recognizing ulnar neuropathy at the elbow (UNE). As latency difference is not a standardized technique, we set up a multicenter study to recruit large numbers of normal subjects and patients with UNE or generalized neuropathy. METHODS: Six centers participated in the study with data obtained from three groups of participants, controls (CTRLs), patients with UNE and patients with generalized neuropathy (GNP). We first verified the anatomical superposition of the ulnar and median nerves in cadaver examination. The optimal recording site for these two nerves was found to be 10 cm above the medial epicondyle. We then standardized the position of the arm with full extension of the elbow and stimulated first the median and then the ulnar nerves at the wrist. CTRLs were examined on both arms at two consecutive visits. RESULTS: We recorded 32 idiopathic UNE cases, 44 GNP patients and 62 controls. We demonstrated that a DLat cut-off value of 0.69 ms brings a sensitivity of 0.86 and specificity of 0.89 to discriminate CTRLs from UNE. We also validated that intra-examiner reproducibility was good. CONCLUSION: We report a lower normal value for DLat than reported in several non-standardized studies and CTRL and UNE groups have clearly separated DLat values. SIGNIFICANCE: Due to its high sensitivity, our standardized technique could be used as a first-line diagnostic tool when UNE is suspected.