NICHOLAS REED: Hi. I'm Nicholas Reed, assistant professor in the Department of Epidemiology at the Johns Hopkins University Bloomberg School of Public Health. In this lecture, we'll be reviewing hearing intervention in the health care system. Hearing in health care. As a primer for hearing loss, what you're looking at on the screen right now is the audiogram. It's a clinical representation of hearing loss. As we go down the y-axis, sounds are going from soft to loud. And as we go across the x-axis, sounds are going from low pitch to high pitch. The curved line in the center of the screen represents a typical age-related hearing loss. You'll see that it's not steady across all frequencies. It's such that there's more hearing loss in higher frequencies, and this is typical of most age-related hearing loss. The dotted lines above that black line represent areas where someone with this hearing loss theoretically could not hear sound anymore. In the center of the graph, there's sort of a boomerang shape or a banana shape with little letters in it. These letters represent the parts of speech. And what you notice is that our hearing loss, our black line, cuts this speech sort of in half. And what this results in is hearing loss not being so much of a volume issue for all sound but rather a clarity issue, specifically age-related hearing loss. Now, if that line was much further down in this graph, in the profound region, this would be a very different conversation. But the vast majority of age-related hearing loss is such that there's a clarity issue. And what this looks like for someone, if you gave them the phrase, you should go to the pharmacy before you get to your house, that might come out more like, "you ould go to arma y be ore you get to your ou." You'll notice that the soft sounds, the soft consonants have been dropped. But the volume of my voice didn't change. And this is how clarity would be perceived for someone with age-related hearing loss in this mild-to-moderate range. It's not that they couldn't hear the sentence. They did not know I was speaking, for example, it's just that little bits and pieces are missing. Some people overcome this by using contextual cues of conversation and sort of putting things together in the bigger picture of speech being presented, such that they're able to follow along even though they're missing little parts. But generally, if someone has this type of hearing loss, the way that a hearing aid is intervening is by giving back those parts of speech specifically, those parts of sound that they might be missing. So we're very much intervening to get back the SH in should, for example, without over-amplifying the second part, the "ould" part of that word. In the context of patient-provider communication, patient-provider communication is a cornerstone of patient-centered care. Patient-centered care represents care that is respectful of and responsive to individual patient preferences, needs, and values. Needs is the key word here for someone with hearing loss. They may have different needs than someone without hearing loss in the health care system. We know that patient-provider communication is key to patient-centered care. And we know that patient-provider communication impacts many different aspects of health care and the health care outcomes. One aspect might be satisfaction with care. And so we ask ourselves, is hearing loss associated with satisfaction with care? And the idea here would be that if someone has hearing loss, it affects their patient-provider communication. Does not affect their satisfaction with care? To explore this, we looked at the Medicare Current Beneficiary Survey. We looked at the 2015 data, which gave us a weighted national sample of 12,311 US Medicare beneficiaries. This is an interview-conducted survey. The outcome variable is satisfaction with care. And it was recoded as either being satisfied or dissatisfied, but in the original response, somebody could respond very satisfied, satisfied, dissatisfied, or very dissatisfied. This is common to recode this scale because this satisfaction with care is actually a relatively rare outcome. Most people say they are either very satisfied or satisfied with their health care. In the final sample, we had 11,441 Medicare beneficiaries. 5,912 reported no trouble hearing, 4,665 said they had a little trouble hearing, and 864 said they had a lot of trouble hearing. What's important to note is that their reports of trouble hearing include if they used a hearing aid, which is why we call this a functional hearing status. It's very representative of their everyday status, even in their best aided condition. The weighted sample overall represents 48.6 million Medicare beneficiaries. Across the total sample, 3.94% were dissatisfied with care. Among those with no trouble hearing, this is 3.11% were dissatisfied with care, whereas those with a lot of trouble hearing, 6.53% were dissatisfied with care. When we built a logistic regression model to further explore this, and we're adjusting for having a usual place of care, age, sex, race, education, income, marital status, self-report general health, functional limitations by activities of daily living, functional limitations by activities of daily living counts, a comorbidity count, and self-report trouble with memory or concentration, we end up with results such that relative to those with no trouble hearing, those with a little trouble hearing have about 50% higher odds of saying they are dissatisfied with care. And those with a lot of trouble hearing have about 77% higher odds of reporting being dissatisfied with care. It's important to note that this data is based off of self-report hearing trouble, but it isn't a large nationally representative data set of Medicare beneficiaries. We also were able in our research to look at the same question in a smaller data set but using pure-tone audiometry, or the clinically accepted measurement of hearing loss. In the Atherosclerosis Risk in Community Study at visit five, there was a hearing loss study pilot. Hearing data using clinical pure-tone measures was collected on 256 participants aged 67 to 89 years. They were all white. It's a very homogeneous population. Our exposure is pure-tone audiometry as measured by the pure-tone average. And our outcome is that same self-report satisfaction with quality of care over the last year. And we report this as a binary variable, either being optimally satisfied or less than optimally satisfied. And this could be interpreted as similar to our last one, as either dissatisfaction or satisfaction. Our outcome variable is self-report satisfaction with quality of care over the last year. This is reported as a binary variable of being either optimally satisfied with care, which was self-report very satisfied, or less than optimally satisfied with care, which was self-report being satisfied, dissatisfied, or very dissatisfied. We had to recode it this way, simply because it was such a rare outcome, again, to be anything less than very satisfied. In our data, we found an interaction between age and hearing loss, such that for a 75-year-old participant, every 10 decibel increase in hearing loss-- this is 10 dB on the pure-tone average-- the odds of being less satisfied with care, less than optimally satisfied, really didn't change. However, for an 85-year-old participant, for every 10 dB increase in hearing loss, the odds of being less than optimally satisfied increased by about 33%. You'll note that the confidence interval does cross 0, but we think we have a signal of interest for what we're looking for, in that hearing loss, specifically among older adults in this study, the older old group, is associated with less than optimal satisfaction with care, even when using clinical pure-tone measures. So why is this even important? Satisfaction with care represents a strong patient-centered outcome that gives us an idea of how somebody perceives their health care. And it has been associated with in previous studies things like treatment adherence, maintaining care, and general health care utilization outcomes. In other words, satisfaction with health care may be a marker that researchers can use and health care systems can use to gauge the health care quality of their patients or participants in a research study and perhaps even use it to predict future risk of poorer health care outcomes. If hearing loss impacts this, it's notable because hearing loss may be modifiable risk factor. We may be able to do something about the hearing loss to improve satisfaction, which may then be a marker that we're improving overall care for people.
