Journal Logo

Speech Perception Gap

Causes, Prevalence of Speech Perception Gaps: Part Two

McCraney, Anna L. AuD

Author Information
doi: 10.1097/01.HJ.0000669888.64686.48
  • Free

Editor's note: This is the second installment of a four-part series. Read part one in our April 2020 issue (

In part one of this article series, we uncovered the walking wounded—patients with underwhelming hearing aid outcomes despite having aidable hearing loss and appropriately fitted hearing aids (HAs)—and discussed how these patients often return for an inordinate number of fine-tuning visits without achieving HA satisfaction. Some of these patients may complain little, if at all, and return every few years looking for something better. In many cases, the walking wounded are those who have given up and mostly wear their HAs in the drawer. While hearing care professionals can guess from these behaviors which patients may be among the walking wounded, research updates on the speech perception (SP) gap can help us address the problem sooner and not wait until a patient is exasperated or clinical time has been wasted.

istockphoto/SIphotography, Hearing loss, research, audiology
Figure 1
Figure 1:
PB max and aided single word scores as a function of four-frequency PTA (adapted from Hoppe, et a.,l 2016).4 Hearing loss, research, audiology.
Table 1
Table 1:
Example of SP Gap Calculation.

An SP gap is a discrepancy between a patient's maximum cochlear potential for word recognition (PB max) and actual aided word recognition (WRaided), which can be expressed as SP gap = PB max WRaided.1 To review, because PB max is the best possible word recognition (WR) score a patient can achieve given the extent of cochlear damage, it also serves as the upper limit of possible WRaided, making it a useful treatment validation metric. Instead of patients simply doing better with their HAs than without, we want them to achieve their full cochlear potential for WR as a result of treatment.

Why is this important? Recent studies have revealed that 25 to 30 percent of well-fitted HA patients have a significant SP gap (defined as ≥ 18% based on the Thornton-Raffin criterion2), indicating that despite having been appropriately fit with a hearing device, these patients are unable to reach their full cochlear potential.1,3 One study reported that the average SP gap among all patients was 20 percent, with some being as large as 70 percent.4

Figure 1 shows an example of a patient with an SP gap. Notice that despite having a PB max or cochlear potential of 68 percent in the right ear and 80 percent in the left, her corresponding WRaided scores were only 44 percent and 48 percent, giving her an SP gap of 24 percent in her right ear and 32 percent in her left ear. It's easy to see why this patient returned to her provider multiple times, complaining that she should be hearing better than she actually was. And she's right: 24 to 32 percent of her cochlear potential remains untapped despite wearing appropriately fitted hearing aids.


Before we discuss the particulars of measuring a patient's SP gap, it's helpful to understand the prevailing thought about why it occurs in patients with properly fitted HAs. It has nothing to do with the hearing aid. Patients who have an SP gap have it no matter what type of HA they are wearing and despite the device being well matched to gain and output targets. The issue is thought to lie in an inefficiency in the middle ear transfer function. This is not to be conflated with middle ear disorder. Just as one can have an inefficient metabolism without having a metabolic disorder, so too can one have an inefficient middle ear transfer function without having a middle ear disorder.

A transfer function refers to how energy makes it from point A to point B, in this case, from the tympanic membrane to the oval window. We know that acoustic energy transfer through the middle ear is least efficient for frequencies between 2 and 4 kHz.5 The HA is delivering appropriate acoustic energy (gain) to the eardrum, as evidenced by a good match to real ear targets, but too much of the sound is absorbed in the middle ear and therefore never makes it to the cochlea. Consequently, some patients can be ideally fitted using real ear measurement (REM) and still find treatment ineffective. And why shouldn't this be true? It's rare for a treatment to work equally well for everyone who tries it. Hearing aids are no exception.


We know why SP gaps occur and that they do so in roughly a quarter of patients with sensorineural hearing loss (SNHL). The question now is this: Since most hearing loss cases are sensorineural, can we narrow the field so we can identify which patients with SNHL are at the greatest risk of presenting with an SP gap? Hoppe's 2016 study on speech perception in HA users shows that although SP gaps exist at all hearing levels, the most significant ones seem to affect those with moderate to severe high-frequency loss or a four-frequency (0.5, 1, 2, and 4 kHz) PTA of 50 to 85 dB HL, as shown in Figure 1.4

It's worth remembering that 38 percent of all patients with hearing loss have moderately severe to severe loss. The same patients wear 70 percent of all hearing aids.6

The assumption that good real ear measures equate to good outcomes persists despite recent evidence to the contrary. Recall the 2016 McRackan and Dubno study referenced in part one of the article series that found that appropriately fitted, audiometrically similar patients with statistically equivalent PB max scores had WRaided that differed from one another by a range of 50 to 60 points.7 This major finding showed that the current fitting paradigm of assuming that similar patients have similar outcomes is inherently flawed.

Patients respond differently to treatment and vast outcome variability prevents the accurate prediction of treatment effectiveness even when word recognition and REMs are good. This begs the question: Is it necessary to test aided word recognition on every patient? Recent research shows that we should be testing those with moderate to severe hearing loss at the very least.

It's clear that aided speech testing can be a reliable means of validating treatment effectiveness. In part three of this article series, we'll discuss the particulars of how to use aided speech testing to identify SP gaps. In part four, we'll talk about what patients can do to reach their full cochlear potential when it can't be done with hearing aids. Hint: Because their HA isn't the problem, it also won't be the solution.


1. Dyer RK, Spearman M, Spearman B, McCraney A. Evaluating speech perception of the Maxum middle ear implant versus speech perception under inserts. Laryngoscope. 2018;128(2):456-60.
2. Thornton A, Raffin M. Speech-discrimination scores modeled as a binomial variable. J Speech Hear Res. 1978;21:507-18.
3. Hoppe U, Hast A, Hocke T. CI candidacy audiometry-based screening procedure. Otol Neurotol. 2015;36:1001-5.
4. Hoppe U, Hocke T, Műller A, Hast A. Speech Perception and information-carrying capacity for hearing aid users of different ages. Audiol Neurotol. 2016;21(suppl 1):16-20. doi:10.1159/000448349.
5. Goode RL. The history and development of the implantable hearing aid. Otolaryngol Clin North Am. 1995;28:1-16.
6. Kockhin, S. MarkeTrak VII: Obstacles to adult non-user adoption of hearing aids. The Hearing Journal. 2007; 60(4).
7. McRacken T, Ahlstrom J, Clinkscales W, Meyer, T, Dubno, J. Clinical implication of word recognition differences in earphone and aided conditions. Otol Neurotol. 2016; 37:1475-81.
Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.