From the desk of Gus Mueller
It's only been a month or so since the 2012 Summer Olympics, and you all probably still recall the interviews with popular gold medal winners such as Gabby Douglas, Missy Franklin, and Ryan Lochte. One of the things they all emphasized was the importance of training; never missing a day for weeks or months or even years. Of course, you don't have to be an Olympian to appreciate the effects of training—remember your early days of writing in cursive, tying shoe laces, or riding a bike? Even animals benefit from training (maybe not so much for cats).
But what about training of the auditory system, and its application to people fitted with hearing aids? With training, can your hearing aid patients improve their listening skills? Reduce the effort required for listening? Improve their understanding of speech? Improve their recognition of speech in difficult listening situations with background noise? If so, then it would be appropriate to make auditory training part of the overall fitting process.
Most experts do believe that auditory training should be part of the hearing aid fitting protocol. Modern audiologic rehabilitation textbooks usually devote an entire chapter to this topic and the evidenced-based Best Practice Hearing Aid Fitting Guidelines of the American Academy of Audiology gives auditory training high grades. Not to mention—auditory training was part of the hearing aid fitting program developed by Raymond Carhart back in the 1940s. While most audiologists do not really "embrace" auditory training, to suggest that it may not be very helpful would be like going up against the Easter Bunny or apple pie (or Carhart). If we dig into the auditory training literature, however, we do see that much of the supporting data does not directly relate to training with typical patients using hearing aids in the real world. We need more research in this area, which takes us to this month's 20Q topic.
Gabrielle Saunders, Ph.D. is the Associate Director of the National Center for Rehabilitative Auditory Research in Portland, Oregon. She also is associate professor in the Department of Otolaryngology at Oregon Health and Sciences University. You are probably used to seeing Dr. Saunders' name associated with the NCRAR, as she has been conducting research, planning the biennial conferences and directing the Education and Outreach program there for the last ten years.
As you know from her many presentations and publications, for some time Dr. Saunders' work has been focused on supplemental approaches to audiologic rehabilitation, such as counseling and various psychosocial factors. She presently is involved with research looking beyond traditional measures to better understand hearing aid outcomes and other hearing-related behaviors, with plans to examine audiologic rehabilitation in the context of the Health Belief Model, and through patient-centered care. What brings her to 20Q, however, is Gaby's recent research involving auditory training. The findings from the large study that she describes certainly are interesting, and maybe not quite as "apple pie" as we would like. It's been 65 years since Carhart's recommendations, and we evidently still have some things to learn about the role of auditory training in audiologic rehabilitation.
Gus Mueller, Ph.D.
Contributing Editor
September 2012
To browse the complete collection of 20Q with Gus Mueller articles, please visit www.audiologyonline.com/20Q
1. How exactly do you define "auditory training?"
There are probably a few different ways to define it, but I'm interested in the training that relates to neuroplasticity. Scientists have shown that neural pathways and synapses in the brain change as a result of behavior, sensory input, damage, dysfunction and training. This neural reorganization is referred to as neuroplasticity. Specific to audiologic rehabilitation (AR), studies have shown that neural responses to sound can be changed intentionally through intensive listening (Tremblay, Kraus, McGee, Ponton, & Otis, 2001; Tremblay, Shahin, Picton & Ross, 2009; Orduña, Liu, Church, Eddins, & Mercado, 2010). This has led us to believe that auditory training (AT) can be used as a form of AR. As a result, a number of AT programs have been developed that can be used as a supplement to hearing aid use, or maybe even to assist hearing impaired individuals who have yet to be fitted with hearing aids.
2. How do these auditory training programs work?
Well, they are developed based on three premises. First, that people with hearing impairment have been deprived of normal sound input for some time and thus need to re-learn to use incoming sensory input. Second, that hearing aids neither return the auditory system to normal nor provide a 'perfect' signal, and thus the hearing-impaired person needs to learn how to use the new signal. Finally, and most importantly, they assume that the skills a person learns within the training program can transfer or generalize to real-world situations. By this I mean that if you learn how to distinguish /ba/ from /da/ while carrying out training exercises, you'll also be better able to hear the difference between the words /bad/ and /dad/ during a conversation.
3. This notion of generalization seems really important. Does auditory training generalize?
That's the million dollar question! Unfortunately, the data about generalization are mixed. Going back to some of the AT studies that took place before computers were widely available, we see that the program of Kricos and Holmes (1996), which trained syllable recognition and communication strategies, generalized to a speech-in-noise outcome measure. While on the other hand, the study of Rubinstein and Boothroyd (1987) found that their top-down training (training on skills such as using linguistic context or practicing auditory working memory) generalized, but that their bottom-up training (training the listener to decode the speech signal by, for example, practicing how to hear the difference between /ba/ and /da/) did not. At the time of Sweetow and Palmer's systematic review (Sweetow & Palmer, 2005), there was evidence that both bottom-up and top-down training can be effective, but because so many different approaches to training had been used and the outcomes measured were so varied, the effectiveness of AT could not be firmly established. Since that review, research continues to show mixed findings. For example, Sweetow and Sabes (2006) showed that indeed, trained skills did seem to transfer or generalize to other stimuli. Specifically, they found that after training on a speech-in-babble task their participants did better on the QuickSIN test than prior to training. On the other hand, Burk and Humes (2008) showed otherwise among a group of older hearing-impaired listeners. They found benefits following training with words, but the skills learned did not generalize to untrained words or untrained keywords in running speech. Of course, it is not fair to directly compare these studies since they used different trainings. Sweetow and Sabes used the Listening and Communication Enhancement (LACE™) program that had people train on 6 tasks for 20 sessions over a 4-week period, while Burk and Humes used a word-based training protocol three times a week over a 12-week period. This is pretty typical of the variation across AT programs. Some, such as the Speech Perception Assessment and Training Program (SPATS; Miller et al., 2007) focus on training bottom-up skills, while others, such as LACE™, train both bottom-up and top-down skills.
4. Have you worked with auditory training programs such as LACE™?
Yes, as a matter of fact we have just completed a large multisite trial. I'll tell you a little about it. This was a 3-site randomized clinical trial that involved 263 Veterans. Some were new hearing aid users and some were experienced users of hearing aids. We compared hearing aid outcomes following training with the two commercially-available forms of LACE™ (computer-based and DVD-based) with outcomes following two control interventions - placebo training and counseling, which I'll describe in a minute. The computer-based and DVD-based forms of LACE™ are pretty similar but not identical. They differ as follows: First, for the computer-based form, users are expected to train for 20 sessions over a 4-week period (AT20), while for the DVD version, users are expected to train for just 10 sessions over a two-week period (AT10). Second, the computer-based form trains five different skills (understanding speech-in-babble, understanding speeded speech, listening in the presence of a competing speaker, auditory working memory, and use of context), while the DVD version trains only four (use of context is not trained). Finally, the computer-based version allows the user to upload their training data to a website, while the DVD version does not.
Getting back to our control groups, one group of participants did a form of placebo training (PT) in which they listened to digitized books via a computer for 20 sessions, while the other control group received routine informational hearing aid counseling (i.e. they had no structured auditory training). We measured behavioral outcomes for all participants using tests of speech-in-noise, time-compressed speech, divided attention, and working memory. We also measured subjective outcomes with the Hearing Handicap Inventory (HHI) and Abbreviated Profile of Hearing Aid Benefit (APHAB). We chose our behavioral measures because the skills required are all ostensibly trained with LACE™, and we could therefore look to see whether training generalized. We measured outcome at baseline, immediately following training completion and then again six months later.
5. Let me interrupt—could you tell me why you included the PT and counseling groups?
Sure. We used the informational counseling control group as an approximation of current standard of care to answer the question, does AT help relative to current clinical practice? As you probably know, most audiologists fitting hearing aids conduct some type of post-fitting counseling but few recommend auditory training. We used the PT group to test whether actively using hearing aids is as effective as adding an AT program.
6. Sounds like an interesting design. So what did you find?
Well, in a nutshell we found no differences in outcome between our four groups of participants on any of the outcome measures. We measured significant but small changes in performance between baseline and follow-up among all four groups, but didn't see any more (or less) change in the groups who did LACE™ training than we did in the control groups. This was the case for all of our behavioral and subjective measures.
7. That's a bit disappointing. How do you explain this?
It is a little disappointing, but we don't think it spells the end of AT because although we didn't find any differences in outcome between our four groups of subjects, we do know that there were individuals in the AT groups who did benefit from AT. However, because there were also subjects who benefitted from the PT and the informational counseling, there were no differences when data from all four groups were compared. What this suggests to us is that we need to find a way to figure out who will and will not benefit from AT.
8. This sounds as though we need to get back to thinking about individual differences. Is that right?
Well, yes and no. Examining individual differences is one approach - and a good one. However, from a research point of view it is difficult to conduct studies like that because it would mean we would have to measure any variable we think might influence outcome, and then look to see whether it predicts benefit obtained from AT. We'd likely end up with a very complicated model that may or may not explain much of the variability. So, we've come up with a different approach - one that assumes the patient knows best what they will benefit from, and what they are willing to do in the way of AR. For the study, we plan to compare outcomes from hearing aid users who select their own supplemental AR (for example AT or Group AR) with those obtained from participants who are randomly assigned to a form of supplemental AR. We hypothesize that the participants who get to choose their AR will have better outcomes than those who are assigned the AR by the clinician.
9. That makes a lot of sense, but do patients have enough information to make good choices for themselves?
I'm pleased you asked that, as it gives me the chance to tell you about patient decision aids and shared decision making. In shared decision making, both the patient and provider are involved in the selection of interventions. The clinician's role is to communicate facts about the condition, intervention options and the risks and benefits of each, while the patient communicates their own ideas, concerns and personal values so that a joint decision is made about the selected health management option with which both the provider and patient agree. In order that patients can effectively participate in these types of decisions they need tools to guide them through their choices. This is where patient decision aids come in. They provide supplemental information about specific interventions, for example, should I do AT? It is important that their content is unbiased and that they tell the patient all about the likely outcomes of an intervention - both positive and negative. They should also describe the various components of an intervention, the commitment required from the patient and the potential outcome if no action is taken. For our study, we have developed a patient decision aid to help participants decide whether they want to supplement their hearing aids with AT, Group AR or perhaps nothing else. This seems to us a reasonable approach since we know that some people do benefit from AT, we just can't yet figure out in advance who that will be.
10. Interesting, are decision aids something new?
Not at all, in fact there is an international organization called the International Patient Decision Aids Standards (IPDAS) Collaboration that works to enhance the quality and effectiveness of patient decision aids. It was originally established back in 2003 to develop criteria for decision aids used in healthcare (Elwyn et al., 2009). Over 100 researchers, practitioners, patients, and policy makers participate. Now that standards have been developed, members of the collaboration rate decision aids and then make them available online. Check out this link to the Decision Aid Library Inventory (DALI), which is maintained by the Ottawa Hospital Research Institute (decisionaid.ohri.ca/). It's really impressive. Next time you have a health-related decision to make, see whether there is an aid that can help you through the decision-making process.
11. Are there any decision aids for hearing-related matters on the website?
Oh yes, absolutely. There is one on the website entitled 'Hearing loss: Should I get a hearing aid?', and another called 'Ear infection: Should I give my child antibiotics?' There are others too but I won't list them here. In the future we plan to post ours on that website also.
12 . Getting back to your AT study, is it possible you didn't see benefits from AT because participants in the AT groups didn't actually do the training?
Good question, but actually, no, that isn't possible. We know for sure that the AT20 LACE™ group did do the training because each person's training data were uploaded to a server. We saw from the data that 84% of the participants did all 20 training sessions, and that as a group the participants improved on each training task within the program (i.e. they showed on-task improvement). In fact, our participants showed more on-task improvement than did Sweetow & Sabes' (2006) participants. We can't say for sure that the AT10 group trained because the LACE™ DVD training data is not stored however, 85% of the AT10 participants reported doing all 10 training sessions.
13. It's impressive that your participants adhered so well to the recommended training schedule. Is this typical?
Alas, no. We have another study going on at the moment in which we are looking at interventions for blast-injured veterans who complain of auditory difficulties and yet have normal or almost normal hearing sensitivity. In this study, ¼ of the participants are randomly assigned to do AT, ¼ are assigned to use an FM system, ¼ are assigned to do AT and receive an FM system, and ¼ receive informational counseling. I don't have time here to talk more about that study other than to say adherence to the recommended training schedule has been extremely poor. So far, about 60% of the participants have completed less than 25% of the recommended training. You should note, however, that the AT program being used is 'Brain Fitness' from Posit Science. It is very different to LACE™. It primarily trains temporal processing and auditory working memory. It requires a whole hour of training each day for 40 days. So, even those who did only 25% of the total training still trained for 10 hours - which is as long as the LACE™ DVD requires for full adherence!
14. It is interesting that you've seen such different degrees of adherence in your two studies. Why do you think that is?
We think that the difference in adherence in the two studies is probably to do with the study populations. In our LACE™ study all of the participants were older retired veterans, while in the blast study the participants are much younger veterans, many are students, most work full time, and some have young children. Other studies also show adherence to be fairly varied. For example, in a study of home-based computerized AT for cochlear implant users, Stacey et al. (2010) reported that about 1/3 of their users completed less than 1/3 of the recommended training. On the other hand, Stecker et al. (2006) reported that "most subjects completed 35-40 days of training" (p. 546) in their study of new and experienced hearing aid users. The report that most likely best reveals real-world adherence to AT is that of Sweetow and Sabes (2010). They examined clinical data from over 3,000 individuals who had signed up to use the LACE™ training curriculum. They reported that adherence (defined as completion of at least half of the recommended number of training sessions) was less than 30%. In other words, real-world adherence to AT may not be so great.
15. Hmm, that's a pretty poor adherence rate. Do you have any suggestions for improving it?
Funny you should ask, because indeed we do! We have just submitted a grant proposal to test the hypothesis that compliance will be improved if users are permitted to select the training tasks they do and the duration they train for each day, as opposed to having a computer program tell them what to do and how long to do it for. We think that if patients are given a description of each training task, what it trains, what it might be helpful for, and if they are provided with tangible training goals (like reaching a certain level), patients will be more motivated to train. We haven't collected any data yet, but we are optimistic that this will improve training adherence.
16. Another question about your study findings—do you have an explanation for the difference between your data and that of Sweetow and Sabes?
I think there are a couple of likely explanations for the difference. First, Sweetow and Sabes used a study design that was different to the one we used. We did group comparisons, while Sweetow and Sabes examined only pre- to post-training data for each subject. In other words, each participant was his or her own control. Because many of the participants in that study benefitted, the average pre- to post-training change was positive, even though benefit was pretty variable. We on the other hand, compared outcomes across different groups of participants. We found there were people who benefitted from the AT, but there were also people who benefitted from the counseling and the PT, so although we saw changes in score between baseline and follow-up, that change was the same for each of our participant groups. This doesn't entirely explain the difference in findings between these studies since the average improvement on the HHI among Sweetow & Sabes' trained participants was 7.5 points, while among our trained participants it was just 2.5 points. It would be nice to make some other direct comparisons between the two studies but it isn't possible as the HHI is the only outcome measure common to the two.
17. Since the study design doesn't fully explain the differences, what other explanations do you have?
Another possible explanation has to do with the study populations. All of our participants were veterans, while Sweetow & Sabes' participants were not. Our participants had likely suffered cochlear damage during their military service up to 50 years earlier. This would mean that the auditory signal reaching their brain would have been abnormal since that time. It seems extremely unlikely to me that 10 or even 20 sessions of AT could fix this. Perhaps training for longer would have yielded better outcomes. For individuals who have had hearing loss for many years I think we need longer and more intensive AT programs.
18. So, you've covered a lot of material here, what is your take home message about AT?
Well, I'd say, let's not give up on it yet. We are optimistic that patients who select AT following a shared decision making process will benefit from it. However, would we say AT is the miracle AR? No, we would not, but combining AT with hearing aid use and other AR certainly can't be harmful, and likely is beneficial for some.
19. Aside from your shared decision making study, what else do we need to look at regarding AT?
I think there are still a lot of unknowns. As we discussed earlier, first and foremost, it would be nice if we could predict who will and will not benefit from AT. Also, we need to find out what types of training materials and tasks are going to be most effective. We need to work out what the optimal training intensity and duration might be. In other words, is it better to train for a short time each day or for a long time on a few days? How many weeks/months do people need to train for? Finally, if people do benefit from AT, do they need boosters every couple of months to maintain that benefit?
20. Well, I guess there are plenty more research questions to go around. Before you go, do you have anything else to add?
Yes, I have referred to 'we' throughout this interview so let me tell you who the 'we' is. 'The 'we' is the fabulous team of people I've been working with - Terry Chisolm, Melissa Teahen, Rachel McArdle, Sherri Smith, Richard Wilson, Michelle Arnold and Paula Myers. We've been working together on the studies I've mentioned here, so they deserve credit for this article, too.
References
Burk, M. & Humes, L. (2008). Effects of long-term training on aided speech-recognition performance in noise in older adults. Journal of Speech and Hearing Research, 51, 759-771.
Elwyn, G., O'Connor, A., Bennett, C., Newcombe, R., Politi, M., Durand, M.,...Edwards, A. (2009). Assessing the quality of decision support technologies using the International Patient Decision Aid Standards instrument (IPDASi). PLoS One, 4, e4705.
Kricos, P. & Holmes, A. (1996). Efficacy of audiologic rehabilitation for older adults. Journal of the American Academy of Audiology, 7, 219-229.
Miller, J., Watson, C., Kewley-Port, D., Sillings, R., Mills, W. & Burleson, D. (2007). SPATS: Speech Perception Assessment and Training System. Journal of the Acoustical Society of America, 122, 3063.
Orduña, I., Liu, E., Church, B., Eddins, A. & Mercado, E. R. (2010). Evoked-potential changes following discrimination learning involving complex sounds. Clinal Neurophysiology, 123, 711-719.
Rubinstein, A. & Boothroyd, A. (1987). Effect of two approaches to auditory training on speech recognition by hearing-impaired adults. Journal of Speech and Hearing Research, 30, 153-160.
Stacey, P., Raine, C., O'Donoghue, G., Tapper, L., Twomey, T. & Summerfield, A. (2010). Effectiveness of computer-based auditory training for adult users of cochlear implants. International Journal of Audiology, 49, 56.
Stecker, G., Bowman, G., Yund, E., Herron, T., Roup, C. & Woods, D. (2006). Perceptual training improves syllable identification in new and experienced hearing aid users. Journal of Rehabilitation Research and Development, 43, 513-518.
Sweetow, R. & Palmer, C. (2005). Efficacy of individual auditory training in adults: a systematic review of the evidence. Journal of the American Academy of Audiology, 16, 494-504.
Sweetow, R. & Sabes, J. (2006). The need for and development of an adaptive listening and communication enhancement (LACE) program. Journal of the American Academy of Audiology, 17, 538-558.
Sweetow, R. & Sabes, J. (2010). Auditory training and challenges associated with participation and compliance. Journal of the American Academy of Audiology, 21, 586-593.
Tremblay, K., Kraus, N., McGee, T., Ponton, C. & Otis, B. (2001). Central auditory plasticity: changes in the N1-P2 complex after speech-sound training. Ear and Hearing, 22, 79-90.
Tremblay, K., Shahin, A., Picton, T. & Ross, B. (2009). Auditory training alters the physiological detection of stimulus-specific cues in humans. Clinal Neurophysiology, 120, 128-135.
