F E Musiek1, W W Lee. 1. Department of Surgery, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire, USA.
Abstract
OBJECTIVE: The purpose of this study was to determine the value, based on true-positive and false-positive rates, of various auditory brain stem response (ABR) indices in discriminating patients with brain stem lesions from patients with cochlear lesions. DESIGN: A factorial design was used in this retrospective study. Receiver operating characteristics (ROC) curves were constructed to analyze the performance of 32 patients with brain stem lesions and 33 patients with cochlear involvement for the absolute latency of wave V; the I-III, III-V, and I-V interwave intervals; the interaural latency difference and the V/I amplitude ratio. Analyzed in the same manner were four combinations of ABR indices. RESULTS: Based on the ROC curves, the clinically most valuable index was the I-V interval, but overall the individual indices showed only moderate sensitivity to brain stem pathology. By combining several indices, the sensitivity to brain stem involvement was improved. However, as expected, the overall false-positive rate also increased. CONCLUSIONS: There is a diagnostic advantage to using more than one ABR index in evaluating patients suspected of having brain stem involvement.
OBJECTIVE: The purpose of this study was to determine the value, based on true-positive and false-positive rates, of various auditory brain stem response (ABR) indices in discriminating patients with brain stem lesions from patients with cochlear lesions. DESIGN: A factorial design was used in this retrospective study. Receiver operating characteristics (ROC) curves were constructed to analyze the performance of 32 patients with brain stem lesions and 33 patients with cochlear involvement for the absolute latency of wave V; the I-III, III-V, and I-V interwave intervals; the interaural latency difference and the V/I amplitude ratio. Analyzed in the same manner were four combinations of ABR indices. RESULTS: Based on the ROC curves, the clinically most valuable index was the I-V interval, but overall the individual indices showed only moderate sensitivity to brain stem pathology. By combining several indices, the sensitivity to brain stem involvement was improved. However, as expected, the overall false-positive rate also increased. CONCLUSIONS: There is a diagnostic advantage to using more than one ABR index in evaluating patients suspected of having brain stem involvement.