Secondary Logo

Segmenting dispensers: Factors in selecting open-canal fittings

Johnson, Earl E.

doi: 10.1097/01.HJ.0000286219.07211.99
Article
Free

Earl Johnson, MS, is a licensed audiologist and has been a research assistant in the Dan Maddox Hearing Aid Research Laboratory at Vanderbilt University for the past 4 years. He also attends Vanderbilt University as a PhD student in Hearing and Speech Sciences. Readers may contact the author at earl.e.johnson@vanderbilt.edu.

The trend is clear. More clinicians and patients are choosing open-canal (OC) fitting hearing aids as a treatment option for hearing impairments. Now, just a few years after GN ReSound's launch of an OC device called the ReSoundAIR, nearly every manufacturer has a similar product. In a very real sense, an OC is more or less a small behind-the-ear (BTE) device with a tube fitting, although some products classified as “open” have the receiver in the ear canal and use a wire rather than a tube to connect with the ear canal.

Last year, BTE sales in the USA increased significantly (no doubt largely because of the OC product), jumping to 33% of all hearing aid sales. In the second quarter of 2006, BTEs accounted for 44% of the domestic market and a large portion of these were OC devices.

Based on OC products' cosmetic appeal and their use in fitting mild to moderate high-frequency hearing losses, one might assume that the OC aid's closest competitor would be the completely-in-the canal (CIC) style. However, according to the Hearing Industries Association (HIA) data,1 the surge in BTE sales is coming at the expense of all three custom hearing aid styles. Sales for the in-the-ear (ITE) styles have fallen from 41% of the market in 2003 to 32% this year, while the in-the-canal (ITC) and CIC styles have fallen from 21% and 14% of the total market, respectively, in 2003 to 15% and 11% in 2006. The OC device is clearly shaking up the hearing aid market.

Back to Top | Article Outline

USER ACCEPTANCE

As Mueller points out, there are several potential advantages to the OC product.2 As has been generally stated about this style of device, hearing aid users have reported a greater satisfaction level with or preference for OC aids relative to their old or current hearing aids in several studies.3,4

According to Taylor, two groups of experienced hearing aid users reported significantly higher satisfaction ratings for OC than for non-OC aids in regards to the quality of their own voice, phone comfort, sound localization, and appearance.5 However, overall ratings in satisfaction with OC and non-OC instruments did not differ significantly.

The patient satisfaction ratings also indicated that the OC style never received a satisfaction rating significantly below that of the non-OC style for any assessed dimension. Thus, these results suggest that from the standpoint of patient satisfaction there is no down side to fitting OC aids (assuming, of course, that appropriate hearing aid selection techniques for the patient are used at the onset).

Back to Top | Article Outline

DISPENSER ACCEPTANCE

As discussed, preliminary data suggest that this new technology has received generally positive ratings from hearing aid users. It is difficult to determine, however, if their dispensers influenced these ratings. Surveys have often found a “halo effect” associated with new technology. Given the labeling effect that has been shown with digital versus analog (new versus old) technology with hearing aid users,6 it's possible dispensers are reflecting the same type of labeling effect, only this time it's open versus closed. Therefore, one goal of this study was to determine if some dispensers had a positive bias toward OC hearing aids.

Back to Top | Article Outline

DISPENSERS SURVEYED

To find out how hearing aid users derive their perceptions about hearing aids, it seemed reasonable to start by assessing the opinions of dispensers. The latest annual survey of hearing aid dispensers by The Hearing Journal (HJ) and Audiology Online (AO), conducted in January 2006, included a group of questions about OC hearing aids. Their primary purpose was to discover why some dispensers are fitting patients with OC devices rather than conventional, closed-canal instruments. Of the 504 dispensing professionals who responded to the survey, 365 (72%) identified themselves as audiologists, 123 (24%) as hearing instrument specialists, and the remainder as physicians or “other.”

The 13 questions in this section asked about a broad range of factors that might influence someone to dispense an OC. Many of the potential factors have been validated by research or at least by some indirect evidence of patient benefit (e.g., reduced occlusion effect, improved sound of own voice, improved cosmetics, and improved localization).

Other potential factors influencing the selection of OC fittings were based on anecdotal reports or manufacturers' claims (e.g., ease of delivery and fitting, reduced return rates, fewer return visits for service, ease of insertion/ removal). Three of the suggested factors listed on the survey were foils, reasons for selecting an open-canal instrument that had little or no support in evidence or logic. The foils were: improved speech understanding in quiet, improved speech understanding in noise, and greater ability to attain real-ear fitting targets.

Respondents rated how much each factor influenced them on a scale from 0 to 10 where 0 signified “no influence at all” and 10 “a very strong influence.” Demographic information collected from the respondents was also used to analyze differences and similarities in OC dispensing and fitting practices as well as to identify the type of dispenser most likely to fit or not fit OC aids.

Back to Top | Article Outline

SURVEY FINDINGS

Figure 1 summarizes how dispensers responded to the question, “What percentage of your hearing aid fittings were open-canal?” Responses ranged from 0 to 100%, with a mean of 17%.

Figure 1

Figure 1

The dispensers surveyed generally believe that their patients are more satisfied with OC products than with traditional instruments (Figure 2). Overall, 92% of respondents said that patients were at least as satisfied or more satisfied with OC than with non-OC fittings and 62% said that patients were either somewhat or significantly more satisfied with OC aids. This could be, in part, because OC aids are frequently fitted on patients who have been dissatisfied with other technology (e.g., people who stopped using hearing aids because of occlusion effect problems).

Figure 2

Figure 2

As mentioned earlier, the primary purpose of this survey was to establish which factors prompted dispensers to recommend OC instruments. Figure 3 shows respondents' ratings of each factor from most to least influential. The mean data shown are presented for all 504 respondents, as analysis by profession (i.e., hearing instrument specialist or audiologist) revealed no significant difference in ratings between the two groups. The mean ratings are accompanied by 95% confidence intervals and relate the likelihood of recurrence for the range of mean ratings. All the significant differences discussed are at a probability of .05.

Figure 3

Figure 3

Back to Top | Article Outline

Reduced occlusion effect is #1

As shown in Figure 3, “reduced occlusion effect” and “improved sound of own voice” were the two highest rated factors in fitting an OC product. These two factors received similar ratings, but were rated as being significantly more influential than any of the other factors listed. Given the close relationship and validity of these two factors, this finding was expected. While there are few data regarding the sound of the user's own voice, a study about eliminating the occlusion effect with OC aids has shown that this is a clear patient benefit.7 Factors ranked third and fourth for selecting an OC fitting were “more comfortable fit” and “improved speech quality,” respectively.

Improved cosmetics also received a relatively high rating—sixth overall. Again, this was not surprising, as Johnson et al. demonstrated the validity of this factor by comparing the visibility of six different hearing aid styles from three different viewing angles.8 The OC ranked second among the six styles examined, just behind the least visible style: a CIC hearing aid viewed from 45° and 90° angles.

Back to Top | Article Outline

Many pick foils

Interestingly, factors included on the survey as foils (shown in red in Figure 3) received high mean ratings. This was particularly true for “Improved speech understanding in noise,” which was rated the fifth most influential attribute of OC fittings. Another foil, “Improved speech understanding in quiet,” ranked 10th. It is puzzling that these two factors were given such high ratings, since there is no logical reason why an OC fitting would improve speech understanding either in quiet or in noise more than a closed fitting.

Regarding speech recognition in noise, at least two studies have shown that with an omnidirectional microphone, performance with an OC is actually worse or not statistically different from unaided performance.9,10 Furthermore, even when an OC instrument has a directional microphone, research has repeatedly shown that OC aids have a lower directivity index in the lower frequencies than do more closed fittings. This reduced overall directivity index translates into poorer performance in noise, not better.11

The third foil on the survey, “Better able to attain real-ear fitting targets,” was rated as having the least influence, which seems appropriate. It was included in the survey because there have been some reports of “free gain” with OC aids.12 However, survey respondents apparently recognize that even though there may be residual ear canal resonance with OC aids, this does not equate to a superior target fit over what could have been obtained with a closed fit. It is true that current phase cancellation-based, feedback-reduction systems, as demonstrated by measures of increased gain headroom,13,14 can enable OC aids to attain real-ear fitting targets that would have previously been impossible with more open fittings. However, a closed fitting with the same feedback-reduction system and its additional attenuation of the feedback loop response would enable the fitting to provide as much or more gain than an open fitting. Therefore, a closed device would always provide a comparable or better fit to target than an OC device.

Back to Top | Article Outline

Dispenser-friendly factors

Also of note are the three factors listed that may have especially positive effects for dispensers. These are ease of fitting and delivery, reduced returns for credit, and reduced return visits for service. These factors were rated 7th, 8th, and 9th, respectively, in how influential they are in the choice of an OC aid.

Since OC devices do not require an earmold impression, their fitting and delivery are easier and often faster, as they may be fitted on the same day as the hearing assessment. As for reducing returns for credit, one clinical study has shown a significantly lower return rate for OC aids than for closed canal aids.15 It reported a 1.8% rate of return for OC aids and 11.3% for closed canal instruments at a particular clinic site.

On the other hand, the 2006 HJ/AO dispenser survey revealed an unexpected finding regarding returns for credit. A t-test of two independent samples on return-for-credit rates from the survey data demonstrates that return rates increased significantly with the percentage of OC aids that respondents dispensed. Specifically, respondents who dispensed OC aids less than 10% of the time had an average return rate of 5.0%, while those who dispensed OC aids more than 10% of the time had an average return rate of 5.9%. And, the 5% of respondents who dispensed more than 50% of their aids as OC aids had average return rates of 7.5%.

While the reason for this is unclear, a possible explanation is that OC devices are especially likely to be fitted on borderline patients with relatively mild hearing losses. As a result, users may receive less benefit than the typical patient given a closed fitting and thus be more likely to return the hearing aids for credit. Alternately, some dispensers may be more cavalier in their recommendation of hearing aids for milder or marginal hearing losses and have greater return for credits due to their patients perceiving little benefit or value from the devices.

It is also interesting that dispensers who fit OC aids more frequently reported significantly different usage of verification procedures. For example, 42% of those who fit OC hearing aids more than half the time said they never conduct probe-microphone measures. However, among respondents who fit OC aids less than half the time, only 28% said they never conduct probe-microphone measures.

Lastly, and perhaps predictably, im-proved localization was rated as the least influential of the valid factors (i.e., all but the foils). This, in my opinion, does not mean that OC aids cannot improve localization over closed fittings, since research has demonstrated that they can.16 More likely it reflects the fact that localization is not generally a major factor in the selection of hearing aids, and also that a satisfied OC user is unlikely to report the often-subtle effects of localization to the dispenser, particularly without an outcome measure. Using an outcome measure, Taylor did show that patients rated satisfaction with sound localization significantly higher with OC versus non-OC aids.5

Back to Top | Article Outline

DISPENSERS SEGMENTED

Using a simple market segmentation approach, the approximate dividing point between the upper and lower 50% of respondents in terms of what percentage of all their fittings were open-canal was identified. This dividing point occurred at approximately 10% (see Figure 1).

Doing this made it possible to explore how the attitudes and experiences of respondents who dispense OC aids more often differ from those of practitioners who fit them less often. An independent samples t-test showed that respondents in the over-10% group assigned significantly higher mean ratings to almost all the factors that might lead to the selection of an OC aid. The only exceptions were two of the foils: improved speech understanding in noise and better able to attain real-ear targets.

In the over-10% group, 42% reported that their patients fitted with OC instruments were significantly more satisfied than their patients fitted with other products. In the under-10% group of dispensers, only 19% said patients with OC fittings were significantly more satisfied than those wearing other closed products (Figure 4). For the group that dispenses more than 10% of their aids as OC fittings, only 2% said that patients were somewhat or significantly less satisfied than those fitted with other product types, meaning that 98% of their patients were at least as or more satisfied with OC fittings.

Figure 4

Figure 4

Regarding the dispensing rate of OC aids, an independent samples t-test showed a significant difference between hearing instrument specialists and audiologists. Specifically, OC aids accounted for 19% of all BTE fittings in the offices of audiologists and 12% in the offices of hearing instrument specialists.

Back to Top | Article Outline

Clustering analysis

We also used a more statistically sophisticated approach to categorize dispensers based on which factors they said influenced their decision to fit an OC device. Using Statistical Package for the Social Sciences (SPSS), this segmentation approach utilized a hierarchical type cluster analysis known as Ward's method.17,18

This method, which is commonly used in cluster analysis, either builds up or breaks down the clusters, respondent by respondent, based on the change in the error sum of squares associated with joining any pair of clusters.18 As this analysis requires respondents to answer all of the influence questions in order to be in-cluded, sample size was reduced from the original 504 respondents to 466. As a result, the average influence ratings across all clusters are slightly different from the earlier reported averages for all 504 respondents. However, the rank order of the factors remained the same.

The analysis identified three “unique” clusters of respondents. In market segmentation, it's usual to label each identified cluster according to its pattern and magnitude of responses. Average ratings for each of the influences for each cluster relative to the overall average are shown in descending order in Figure 5. Comparing the mean ratings for one cluster with those of other clusters and with the overall mean ratings helps one to assign an appropriate name to a cluster. Thus, clusters 1, 2, and 3 were named OC Enthusiasts, Middle of the Road, and Not Fooled by Foils, respectively. The OC Enthusiasts cluster accounted for 30% of the respondents, while the Middle of the Road and the Not Fooled by Foils clusters accounted for 41% and 29% of the respondents.

Figure 5

Figure 5

The OC Enthusiasts cluster assigned consistently higher mean ratings for all of the factors in fitting OCs, suggesting that this group is strongly influenced by this style of device. As its name suggests, the Middle of the Road cluster generally gave mean ratings between the other two clusters, except that they gave the lowest mean rating of any cluster to the improved cosmetics factor.

Respondents in the Not Fooled by Foils cluster followed roughly the same pattern of descending responses as the other two clusters while rating the three foils in the range of 1.0 to 1.5. The Not Fooled by Foils cluster gave the foils significantly lower mean ratings than did the other two clusters. While its ratings for the foils were appropriately low, this cluster still rated the reduced occlusion effect and improved sound of own voice near 8 on the 1 to 10 scale.

Had there been significant differences in the pattern of responses to the surveyed factors, the clustering analysis would have identified the unique group of dispensers (e.g., a strong influence of reduced return rate for credit, localization, etc.). However, it appears that the dispenser clusters mainly varied in the overall magnitude of influence they assigned to the various factors and not the order of influence, as the three clusters followed essentially the same pattern of descending influence.

Based on the responses from the three clusters, one might hypothesize that the clusters would differ on the percentage of OC hearing aids they dispensed last year. This hypothesis was confirmed as the OC Enthusiasts, Middle of the Road, and Not Fooled by Foils clusters dispensed on average 21%, 18%, and 14% of their aids as OC last year.

Additionally, respondents in all three clusters reported patient satisfaction levels with OC aids relative to closed-canal aids in a way consistent with each cluster's rate of fitting OC aids (Figure 6). Specifically, the largest portion of the OC Enthusiasts cluster, 42%, reported satisfaction levels with OC aids significantly higher than with other products. This cluster was followed by the Middle of the Road and Not Fooled by Foils clusters in which 25% and 21% of the respondents, respectively, reported significantly higher satisfaction levels.

Figure 6

Figure 6

One might also think that the dispensers with the best understanding of a product and who are most aware of its true patient benefits would enjoy the greatest success in dispensing the products. However, these data suggest otherwise.

The Not Fooled by Foils group, who seemed to recognize that these products probably don't improve speech understanding in quiet or in noise (compared with closed fittings), reported less patient satisfaction with the OC product than the other clusters did.

Thus, the satisfaction rating results most likely relate to the mean ratings of the factors that influenced dispensers to fit OC aids. That is, OC Enthusiasts, who assigned the highest mean ratings on the factors of influence, also reported the highest satisfaction ratings. In contrast, the Not Fooled by Foils group, who gave the lowest mean ratings on the influence factors, reported the lowest satisfaction ratings.

Back to Top | Article Outline

Relating demographic data to the clusters

Market segmentation techniques also attempt to utilize demographic data of the respondents to help characterize and allow for prediction of the type of respondents in identified market clusters. That is, if we knew that either audiologists or the hearing instrument specialists were primarily in the Not Fooled by Foils or the OC Enthusiasts cluster, this would be very useful marketing information.

However, analysis of our survey data revealed no such significant relationship for any cluster for any demographic variable (e.g., profession, degree, work experience, gender). For example, in the two cluster extremes, OC Enthusiasts and Not Fooled by Foils, the incidence of audiologists and hearing instruments specialists was nearly identical, (OC Enthusiasts: audiologists 28% and hearing instrument specialists 30%; Not Fooled by Foils: audiologists 30%, hearing instrument specialists 32%). There was similarly little difference among audiologists by their degrees. Of those with master's degrees, 28% were in the OC Enthusiasts cluster and 27% were Not Fooled by Foils. Among the doctors of audiology, 27% were OC Enthusiasts and 30% were Not Fooled by Foils.

Back to Top | Article Outline

SUMMARY

The data from this dispenser survey identify several of the influences that most affect a dispenser's decision to recommend open-canal over non-OC hearing aids for patients (i.e., reduced occlusion effect, improved sound of own voice, and more comfortable fit).

In addition, it was found that dispensers are not equally influenced by the factors listed on the survey for recommending OC fittings. Thus, three groups of dispensers were defined. Interestingly, the demographic data of the dispensers did not assist in classifying the demographic characteristics of the dispensers in each of these groups. In other words, it appears that a particular dispenser with certain demographic characteristics may be found in any of these three groups.

Among the OC Enthusiasts cluster (30% of survey respondents), it appears that a halo effect does indeed exist for these dispensers regarding the potential of OC aids relative to non-OC aids. On most factors that might influence a dispenser to recommend an OC fitting, even including the foils, this group gave high rankings to such devices. The high rankings the OC Enthusiasts cluster gave to foils on the survey (i.e. improved speech understanding in noise − 8.5, improved speech understanding in quiet − 7.0, better able to attain real-ear fitting targets − 5.1) accompanied by increased dispensing rates of OC aids indicate that an OC label may influence dispensers to recommend these devices.

In addition, this study showed that the amount of perceived influence for selecting the OC style corresponded to dispensing behavior. Specifically, increases in the amount of perceived influence resulted in respondents dispensing more OC aids. And the more OC aids respondents dispensed, the more patient satisfaction they perceived with this product.

Back to Top | Article Outline

REFERENCES

1. Hearing Industries Association: Supplemental Survey on Open-Fittings, Directional Microphones, and Telecoils. Alexandria, VA: 2006.
2. Mueller HG: Open is in. Hear J 2006;59(11):11–14.
3. Otto WC: Evaluation of an open-canal hearing aid by experienced users. Hear J 2005;58(8):26–32.
4. Flynn MC, Obeling L, Johansen J: Thin-tube open fittings: Preferred patient populations and study results. Hear Rev 2006;13(3):62–95.
5. Taylor B: Real-world satisfaction and benefit with open-canal fittings. Hear J 2006;59(11):74–82.
6. Bentler RA, Niebuhr DP, Johnson TA, Flamme GA: Impact of digital labeling on outcome measures. Ear Hear 2003;24(3):215–224.
7. Kiessling J, Brenner B, Jespersen CT, et al.: Occlusion effects of earmolds with different venting systems. JAAA 2005;16(4):237–249.
8. Johnson CE, Danhauer JL, Gavin RB, et al.: The “hearing aid effect” 2005: A rigorous test of the visibility of new hearing aid styles. JAAA 2005;14(2):169–175.
9. Fabry D: Facts vs. myths related to open-fitting hearing aids. Hear Rev 2006;13(5):20–25.
10. Kuk F, Keenan D, Ludvigsen C: Efficacy of an open-fitting hearing aid. Hear Rev 2005;12(2):26–32.
11. Ricketts TA: Directional hearing aids. Trends Amplif 2001;5(4):139–176.
12. Mueller HG, Ricketts T: Open-canal fittings: Ten take-home tips. Hear J 2006; 59(11):24–39.
13. Johnson EE, Ricketts TA, Hornsby BWY: The effect of digital phase cancellation feedback-reduction systems on amplified sound quality. JAAA In press.
14. Kiessling J, Margolf-Hackl S, Geller S: Technical Report #13 Field Test of an Occlusion-Free Hearing Instrument. Tastrup, Denmark: GN ReSound, 2003.
15. Gnewikow D, Moss M: Hearing aid outcomes with open- and closed-canal fittings. Hear J 2006;59(11):66–72.
16. Byrne D, Noble W: Optimizing sound localization with hearing aids. Trends Amplif 1998;3(2):51–73.
16. Byrne D, Noble W: Optimizing sound localization with hearing aids. Trends Amplif 1998;3(2):51–73.
17. Ward J: Hierarchical grouping to optimize an objective function. J Am Stat Assn 1963;58:236–244.
18. Lilien GL, Rangaswamy A: Marketing Engineering, 2nd edition. St. Victoria, BC: Trafford Publishing, 2004.
    © 2006 Lippincott Williams & Wilkins, Inc.