It’s difficult to answer the question about how close you can get to threshold using eABR. eABR testing allows us to record a threshold on the equipment, and this is then used to set the T level on the CI device. There’s a few interesting points I’d raise about this:
- To assess how close this gets to “threshold” depends on the measure you’d use to define threshold. For CI users we don’t have a gold standard against which to reference/compare the eABR results.
- It’s worth noting that even in traditional electrophysiological responses such as ABR/ASSR we don’t reach exact threshold and have to apply correction factors. As it stands there aren’t any correction factors for eABR yet because the stimuli from the implant are so different to traditional acoustic stimuli; this makes it very challenging to establish a common ground between different tests in order to establish these correction factors.
- Typically in most audiology we’d use a comparison with behavioural/subjective assessment, but in this instance we need to remember that subjective perception of sound can be challenging in many CI device users, because many will never have heard sound “properly” before, so how do we know how good/correct/accurate their perception of threshold actually is? The risk is that in comparing eABR thresholds to behavioural/subjective thresholds, we are comparing the eABR results to a possibly inaccurate measure. However, it’s probably the best we have. But we can see some variability between subjects as a result of their ability to “perceive sound” – this will depend on auditory deprivation, length of and nature of hearing loss, previous amplification etc.
- The literature has really demonstrated that NRT is inaccurate. There are studies that have compared eABR to NRT (also known as eCAP) and eABR elicits lower thresholds than NRT – this suggests that eABR is more accurate than NRT, but doesn’t answer the question specifically about how close to threshold you can get, because again the flawed NRT shouldn’t be used as the “gold standard” against which we should benchmark eABR.
And so, our implementation of eABR is that it’s an efficient starting point as there is a very good link between eABR levels and T levels (far more so than NRT/eCAP) and the fact that we can record lower thresholds, and from further along the nerve than NRT/eCAP, with eABR really puts it in the strongest position for use in CI programming. However, adjustments may be needed based on other tests and patient/parental reporting. Truly, a test battery approach is required to create the optimum CI map.