April 2020 The Hearing Journal This is the first instalment of a four-part series.
Every audiologist can think of at least a few patients who have proven difficult to fit. Sure, some cases of hearing loss are harder to treat than others, but there are still patients who, on paper, appear to have a very aidable hearing loss but continue to present with various complaints. They have been fit with state-of-the-art hearing aids (HAs); the fitting has been verified to meet target; they have been counselled regarding realistic expectations, and still they keep showing up at the clinic to have their already expertly programmed HAs fine-tuned…again. These are often the same patients who have several other pairs of HAs that they've mostly just “worn in the drawer.”
A reality in health care is that rarely is a treatment effective 100 percent of the time. HAs are no different. Some patients will simply do better than others, much like some patients can control their cholesterol levels with a diet change while others also require medication. We tend to look at patients with similar unaided word recognition under inserts or maximum phonetically balanced word recognition scores (PB max), and think they will do about the same with HAs as other patients with a similar PB max. This is akin to giving all patients the same diet or drug and assuming they will all have the same results.
Figure. 1: Aided word recognition scores as a function of unaided word recognition under insert earphones by percentage correct, adapted from McRackan, et al., 2016. The highlighted area illustrates the variability of aided outcomes among patients with equivalent PB max scores. Hearing loss, hearing aids, audiology.
However, a 2016 study by McRackan and Dubno has shown that audiometrically similar patients with statistically equivalent PB max scores could have widely variable aided word recognition (WRaided) and, therefore, widely variable aided outcomes. For example, as shown in Figure 1, for patients with PB max scores between 80 and 90 percent, the WRaided scores ranged from 52 to 96 percent. Similar ranges can be seen across every equivalent PB max band. Keep in mind that these data were pulled from a tightly controlled FDA clinical trial, so not only did the patients have similar levels of hearing loss (moderate to severe high-frequency sensorineural hearing loss), their HAs were verified to meet targets using the same algorithm (NAL-R). Despite being a homogenous group, their outcomes were still widely variable. This major finding showed that a patient with good unaided word recognition cannot be assumed to have good aided outcomes even with target-matched HAs.
When looking at the patients in Figure 1 who had a PB max between 80 and 90 percent, it's not hard to imagine that a patient whose WRaided score was 52 percent was having a very different HA experience from that of a patient whose WRaided score was greater than 90 percent. Yet the current paradigm for fitting assumes that all patients in the same PB max range will have similar outcomes.
What happens to these patients now? They are the “walking wounded.” We have been lured by good PB max scores and real ear measurement (REM) fittings into believing they're doing as well as they can, but as the McRackan-Dubno study shows, they are not. So they may be viewed as difficult-to-fit patients—those who return for an inordinate number of fine-tuning visits. They may have given up altogether and mostly wear their HAs in a drawer. Some may even be considered good HA users because they come in every two to three years asking, “What's new?” or “What's better?”
Because it isn't customary to do aided word recognition testing unless and until a patient is struggling enough to be considered for a cochlear implant (CI), we're failing to identify our walking wounded. When patients come back and complain that they're not hearing as well as they think they should, our first instinct is to remind them that they've been well fit with the latest technology and they're really doing as well as can be expected given their hearing loss. But according to the McRackan-Dubno data, we now know that aided outcomes among audiometrically equivalent patients may be dramatically different–by as much as 60 percent from patient to patient. In short, some patients really aren't doing as well as they should. Appropriately, McRackan and Dubno concluded that we cannot rely on PB max to predict individual HA performance even for patients whose HAs meet target. Therefore, it's an absolute necessity to validate HA treatment effectiveness by testing aided WR for each patient.
Validation typically takes the form of having patients complete a subjective survey. In some cases, in lieu of formal validation, patients are simply asked if they're hearing better. While it is important to seek patients’ input regarding their perceived benefit, subjective measures are only confirming that a patient is doing better with treatment than without. But is that really the best metric? Objectively demonstrating that a patient is not just doing better but doing as well as his or her anatomy will allow is the ultimate evidence of treatment effectiveness. In other words, is the patient reaching his or her full cochlear potential as a result of treatment?
The question that follows is: What is a patient's full cochlear potential? Halpin and Rauch concluded that hearing loss is best characterised by a patient's word recognition ability, not his or her audiogram. They pointed out that PB max is a patient's best possible word recognition score. Therefore, since PB max represents a patient's cochlear potential for word understanding, it is also the upper limit of possible aided WR. In other words, if a patient's PB max is 84 percent, his or her WRaided should be around 84 percent as well, not something significantly less. When a patient's WRaided is consistent with his or her PB max, he or she is reaching his or her full cochlear potential, and treatment effectiveness has been validated, not assumed. When there is a difference between PB max and WRaided, then there is a speech perception gap (SP Gap), which can be expressed as SP Gap = PB max-WRaided. The greater the SP Gap, the less benefit the patient is receiving from hearing aids.
We all want our patients to do as well as they possibly can, not just better than if they hadn't been treated. Knowing what we know now, it is obvious that we need to validate—not just verify—aided performance by measuring WRaided. Objectively quantifying patients’ aided benefit has many advantages, including:
- providing a pre-treatment anticipated objective outcome measure to inform patient counseling and the setting of realistic expectations,
- identifying whether a patient is reaching his or her full potential, prompting either expeditious intervention when potential hasn't been met or cessation of fine-tuning if it has,
- differentiating between patients with unrealistic expectations and those with untapped cochlear potential, and
- providing value that cannot be replicated by practices or retailers who simply dispense products.
In part two of this article series, we'll discuss in detail the causes and prevalence of SP Gaps, ways to identify patients who have them, and which patients are most likely to be among those affected. Part three will explore ways to measure SP Gaps, and finally in part four, we'll examine the treatment alternatives that may help patients with SP Gaps reach their full treatment potential when hearing aids cannot. In the end, and in keeping with our standard of practicing evidence-based hearing care, whether a patient procured hearing aids online, at our practice, or from a different one, we have the power to make sure they're not among audiology's walking wounded.