Clinical methods for the objective estimation of loudness discomfort level (LDL) using auditory brainstem responses in patients.

Our previous work showed that the slope of the JV latency/intensity function can be used to estimate the loudness discomfort level (LDL). In this study, the previous work was repeated using normal subjects and patients whose LDL was unknown at the time of test. The results showed that the original model still applied. Further measures were taken to evaluate clinical methods of applying this technique. Four basic approaches were used, a 'correction factor' model and a 'curve-fitting' model and these were applied to data obtained from measurements taken in both 5 dB and 10 dB increments. The results showed that the 'correction factor' models were better than the 'curve-fit' approach. The prediction of LDL based upon the 5 dB increment data gave the greater accuracy but the 10 dB increment data gave predictions that were sufficiently accurate for clinical use (95% confidence limits of +/- 8 dB). Thus this ABR estimator has wide application and good accuracy in estimating subjective LDL. A clinical protocol for applying this technique is described.