Closing the Open Fitting: An Effective Method to Optimize Speech Understanding

Choosing a closed hearing aid coupling system can positively effect speech understanding in background noise—although, for some patients, at the sacrifice of own-voice sound quality. However, new systems like Signia Own Voice Processing (OVP) are making it possible to “get the best of both worlds.” The mean SNR additional benefit for a closed fitting in this study was 3 dB, and as much as 5 dB for some participants. With as much as 17% improvement in speech understanding for each 1 dB of SNR improvement, this can represent a magnitude sufficient to be able to change a patient’s socialization activities.

Tech Topic | April 2019 Hearing Review

By Matthias Froehlich, PhD, and Veronika Littmann, PhD

There is little documentation of the first open-canal (OC) fittings, but we know they received considerable attention immediately after the CROS type hearing aid was introduced in the 1960s. It was quickly discovered that the open earpieces (or tubing only) used for the CROS hearing aid also could be used for ipsilateral high-frequency amplification (hence, the term IROS [ipsilateral routing of signals], referring to a large vent).¹ In the ensuing decades, there wasn’t much excitement about open fittings, and this general attitude, encouraged by the popularity of custom instruments, continued until the early 2000s—a time when ITEs/CICs had an 80% US market share.

Starting around 2003, four main driving factors emerged which combined to move dispensing professionals and patients toward OC fittings:

1) Smaller “cuter” mini-BTEs;

2) Nearly invisible, thin-tube fittings;

3) Pre-made fitting tips, and

4) Most importantly, improved feedback reduction systems that allowed for acceptable gain with open products.

As reviewed by Mueller in 2006,² these factors combined to offer a wide range of potential patient benefits, and OC products rapidly took over a sizable share of the total hearing aid market. While no precise data exists, we suspect that today many hearing care providers (HCPs) fit 50% or more of their patients with open fitting tips/earmolds.

But, like many things related to fitting hearing aids, the potential benefits of an OC fitting come with some compromises. First, depending on the degree of the patient’s hearing loss in the low frequencies, it might not be possible to achieve desired prescribed gain for this frequency region. And, it may be necessary to use maximum amplifier gain, which has the unwanted effects of higher circuit noise, greater battery drain, and possible distortions. Reduced audibility often means reduced intelligibility. This especially can be an issue for streamed signals, such as telephone conversations, which may already have reduced clarity.

The second important issue is related to the advanced signal processing of the instruments, in particular, digital noise reduction (DNR) and directional technology. Consider DNR; in most cases, the unwanted noise has a low frequency emphasis. But with an open fitting there is little gain in the lows; the speech-to-noise difference (DNR benefit) cannot be any larger than the gain present for speech in quiet (simply stated, you can’t take away what isn’t there).

Moreover, the open fitting allows low-frequency noise to pass directly to the eardrum with little or no attenuation. Imagine a patient using hearing aids that have the capability to employ a strong DNR of 10 dB in the low frequencies. The patient is using his hearing aids at a large party with background noise of 80 dB SPL (mostly low-frequency). Will the noise at his eardrum be 70 or 80 dB SPL? Unfortunately, with an open fitting, it will be at or near 80 dB. This is illustrated in Figure 1, adapted from Mueller and Ricketts,³ which shows the magnitude of DNR in the real ear for an open fitting, for a hearing aid that had essentially equal DNR effects across frequencies for a closed fitting. Observe that in the mid to high frequencies, there is 10-15 dB of DNR benefit, but for 1000 Hz and below the DNR effect is negligible—yet this is where it is most needed.

The same general principles apply to directional technology. For example, the findings of Ricketts et al,⁴ who tested 14 different products with open fittings, illustrated that the average directivity index (DI) was only around 2.0 dB. This is in contrast to average DIs of over 4 dB for instruments with closed fittings tested in the same laboratory.⁴ Consistent with these findings, Valente and Mispagel⁵ found that the SNR benefit for speech recognition for directional processing for open fittings was only about half of that which was normally observed in other research at that time for closed fittings: 2 dB vs 4 dB or more.

So, we have the potential advantage of OC fittings, which most notably is the improved sound of one’s own voice, but to obtain this, we give up much of the advantages of DNR and directional technology. Is it possible to have it both ways?

Probably. In recent years, Signia has developed an algorithm referred to as Own Voice Processing (OVP). This feature is designed to automatically detect the user’s own voice, and then instantaneously alter the gain and output to make the voice sound more natural to the user (see Høydal⁶ for details). Clinical studies have been conducted to evaluate the efficacy of the Signia OVP algorithm.^7-9 Overall, for both ratings of naturalness and satisfaction, OVP has been shown to make a significant improvement. The positive impact of OVP was perhaps best illustrated by the findings comparing the satisfaction of own voice for Signia OVP-On with a closed fitting to two competitive products with open fittings—group satisfaction for their own voice was 86% for Signia compared to 58% and 37% for the two competitors.⁷

This research reveals that if OVP is applied, in many if not most cases, there is no additional patient benefit for an open fitting. But does the use of a closed fitting improve speech understanding as one might predict? The answer to this question was the purpose of the current research.

Clinical Study Methods

Participants. The participants were 20 individuals with bilateral mild-to-moderate hearing losses (mean age = 74 years, 10 males and 10 females). All had symmetrical downward sloping configurations, with a mean audiogram ranging from 26 dB at 500 Hz to 66 dB at 4000 Hz.

Procedures. All participants were fitted bilaterally with Signia Pure 7 Nx which were programmed to the Nx-fit algorithm for a bilateral fitting for experienced users. All special features except for frequency lowering were activated, and set at the default. Experimental testing was conducted with four different ear couplings:

1) Signia open fitting tips;

2) Signia vented click sleeves;

3) Signia closed click sleeves and

4) Custom made unvented canal earmolds.

The three Signia coupling systems were sized for each participant.

The target speech material was the sentences of the Oldenburger Satztest, commonly referred to as the OLSA. All sentences in this material (spoken in German) follow the structure: subject-verb-number-adjective-object (eg, Robert buys five red flowers). The background competing signal was the standard OLSA multitalker babble, presented uncorrelated from seven different loudspeakers surrounding the listener.

Testing was conducted in a standard audiometric test suite. The array for the presentation of the target and competing speech material consisted of eight loudspeakers surrounding the participant, equally spaced at 45° increments, starting at 0° (ie, 45°, 90°, 135°, etc). The participant was seated 1.5 meters from all loudspeakers in the center of the room, directly facing the speaker at 0°, which was used to present the target sentences. The competing signals were presented from the seven other loudspeakers. These competing signals were calibrated to equal 65 dB SPL at the position of the participant when all seven loudspeakers were activated. The signal remained constant at this level for all testing.

For the speech recognition testing, the competing noise was introduced, and the target sentences were presented adaptively, according the standard test procedures of the OLSA. The step size was automatically reduced as testing progressed; the SNR-50 threshold in noise was automatically calculated. The order of testing for the four different coupling systems was counterbalanced, and participants were randomly assigned to a specific ordering.

Before the speech recognition testing was completed, a real-ear occluded response (REOR) using a 65 dB SPL ISTS input was conducted for each ear with each of the four coupling systems. This was to ensure that the four systems differed in their tightness of fit. The results of this testing is shown in Figure 2.

The results of the REOR testing revealed a very predictable pattern. For example, if we observe the mean real-ear occluded gain (REOG) for the open fitting, we see that it closely mirrors the well-known average real-ear unaided gain (REUG), with a peak around 2700 Hz of 17 dB, suggesting that the fitting is very open. As we progress to the more closed systems, we see a systematic reduction in the residual REUG. Finally, with the most closed system, the custom earmold, we not only observe the disappearance of the REUG, but note significant attenuation affects occurring (REOGs below the input level). Overall, these findings show that we indeed were testing ear coupling systems with different occluding properties.

Results

The OLSA speech test was delivered to all participants for the four different ear coupling conditions, and SRT-50 speech recognition values were obtained. Statistical analysis measures (ANOVA with a Greenhouse-Geisser correction) showed a statistically significant difference between the four different ear coupling systems (F[2,07,39,36]=50,31, p<0.001). Bonferroni-adjusted, post-hoc analysis revealed a significant difference between the open condition and all other conditions (p<0.001), and a significant difference between both click sleeve vented and click sleeve closed and the custom molds (p<0.005). The difference between click sleeve closed and click sleeve vented was not significant.

Figure 3 shows the mean SNR values for the different ear coupling conditions. Interestingly, but not surprisingly, the findings are aligned much in the same way as the REOGs shown in Figure 2. Note that as the coupling system became more occluding, the speech recognition SRT-50 improved. A mean 3 dB SNR advantage occurs when we go from the most open to most closed.

In addition to the mean values, it’s useful to examine individual distribution. Figure 4 displays the differences (benefit) obtained for each individual for their SRT-50 score for open versus the closed mold fitting. The advantage for the closed mold condition was observed for all participants. For many, the advantage was in the 3-4 dB range, but for some, the benefit of a closed coupling was over 5 dB. It is important to point out that, even with open fittings, there is a significant SNR advantage compared to omnidirectional because of Signia Narrow Directionality, so the values shown here represent the increased benefit for a closed fitting (eg, a typical patient might have a 3 dB SNR Narrow Directionality advantage with open, and if they now have a 5 dB improvement with closed, the total benefit of Narrow Directionality would be 8 dB SNR for that patient).

Summary

The findings of this research clearly show that the coupling system chosen by the HCP can have a substantial effect on subsequent speech understanding in background noise for the patient. As shown in Figure 3, the mean SNR additional benefit for a closed fitting was 3 dB. It’s important to point out the clinical relevance of this. For the speech test that was used in this research (the OLSA), each dB of SNR improvement corresponds to roughly 17% improvement in speech understanding.¹⁰ This means that the average patient who is fitted with closed rather than open will experience a 50% improvement in speech understanding for some listening situations in background noise. A magnitude this large can change a patient’s socialization activities.

Over the past 15 years, there has been fast moving trend to fit individuals with OC couplings. It is true that, for some patients, this improves the sound quality of their own voice. But if the special processing of the hearing aid already does this—as it does with Signia OVP—then the open fitting isn’t really needed, and it is possible to use a more closed coupling system. We know that DNR and directional technology will operate more effectively with a closed fitting, and, as shown in this research, substantial benefit in speech understanding will be provided for the patient.

Citation for this article: Froehlich M, Littmann V. Closing the open fitting: an effective method to optimize speech understanding. Hearing Review. 2019;26(4)[Apr]:16-20.

Correspondence can be addressed to Dr Froehlich at: [email protected]

References 

1. Mueller HG, Ricketts TA, Bentler, R. Modern Hearing Aids: Pre-Fitting Testing and Selection Considerations. 1st ed. San Diego, CA: Plural Publishing; 2013.

2. Mueller HG. Opens is in. Hear Jour. 2006; 59(11):11-14.

3. Mueller HG, Ricketts TA. Open-canal fittings: Ten take-home tips. Hear Jour. 2006; 59(11):24-39.

4. Ricketts TA, Bentler R, Mueller HG. Essentials of Modern Hearing Aids: Selection, Fitting, and Verification. San Diego, CA: Plural Publishing; 2019

5. Valente M, Mispagel KM. Unaided and aided performance with a directional open-fit hearing aid. Int J Audiol. 2008;47(6):329-336.

6. Høydal EH. A new own voice processing system for optimizing communication. Hearing Review.2017;24(11):20-22.

7. Powers T, Froehlich M, Branda E, Weber J. Clinical study shows significant benefit of own voice processing. Hearing Review. 2018;25(2):30-34.

8. Froehlich M, Powers TA, Branda E, Weber J. Perception of own voice wearing hearing aids: Why “natural” is the new normal. Audiology Online. April 30, 2018. https://www.audiologyonline.com/articles/perception-own-voice-wearing-hearing-22822

9. Powers TA, Davis B, Apel D, Amlani AM. Own voice processing has people talking more. Hearing Review. 2018;25(7):42-45.

10. Wagener K, Brand T, Kollmeier B. Entwicklung und Evaluation eines Satztests fur die deutsche Sprache Teil III: Evaluation des Oldenburger Satztests. Z Audiol. 1999; 38(3):86-95.