Mapping Hearing Loss Across America: Insights from the Sound Check Study Revealed

All right, everybody, and welcome to another episode of This Week in Hearing. I’m thrilled to be joined today by David Rein. David, thank you so much for coming on the show today. David is a senior fellow at NORC, which is the National Opinion Research Center at the University of Chicago, and he is the principal investigator of the Sound Check study. so let’s kind of go through and unpack some of that there. David if you wouldn’t mind sharing a little background on yourself NORC, and how this study came to be. Sure, yeah, thanks for having me. I appreciate being here. I’m excited to share our research with your audience. So it’s always a thrill to be able to show things online like this. so my careers as a researcher I run a group at NORC called the Public Health analytics program. And what we do is use usually national publicly available data sets to answer questions of public health importance. and the key thing to that is that usually no one data set has all of the answers that you want. So we have to apply different types of methods to combine the data sources to get the answers that you’re interested in. I’ve been working along with a number of other topics in the area of vision health for a long time and was relatively unfamiliar with the hearing health world. And then CDC, through the division of population Health issued an announcement for funding for applications to do work in an area that CDC’s chronic health division wasn’t currently doing work in. and just looking at the national estimates of hearing loss, I was struck one by the high prevalence of hearing loss, and then also struck by the fact that CDC does not have designated adult programs for hearing loss although they do have some excellent work related to hearing loss in infants and children. so we pitched the idea to CDC to do this project. and the idea behind. Well, I’ll back up for 1 second. So the idea behind the grant announcement was to raise education and awareness about the condition that you picked. and our strategy to do that was to create new national, and really more importantly, state and county level estimates of hearing loss. And the thought process was by creating this map that would give people something engaging to interact with helping to drive interest in hearing loss, and building sort of the national conversation around hearing health and what we can do about it. Yeah, I mean, the map is absolutely fascinating. We’ll pull it up here in a second. so that the viewers can see what you all have created as you’re doing that, I’m curious, how did you all actually sort of populate the data set? it’s through surveys, correct? Right. So it’s a collection of different surveys, right. And surveys use all sorts of different types of methodologies. So the backbone of our estimates is a national survey called the National Health and Nutrition examination survey, or NHANES which has been running since, for decades. and the key thing about NHANES to be aware of is that they do direct measurements of people’s hearing. So they have mobile examination centers and recruit people to come for different types of health examinations, one of which is the audiology examination. so we were able to get four frequency pure tone threshold measurements for respondents in NHANES over a number of different years. And then by pooling those data from multiple data collections together, get enough sample to create national estimates of hearing loss. So that’s been done before with some excellent work. we applied a method, a statistical method to it, to bring a little more stability to the estimates. but I think the key thing for our model was that we weren’t interested in just the national level. We were interested in the state and local level. so for that we had to use other data sources particularly a survey called the American Community Survey, which is if not the largest, probably one of the largest surveys in the world. It’s conducted by the Census Bureau and it asks people at the local level are you deaf or do you have serious difficulty hearing? which is a self report question. Got you. Okay, cool. So that’s how the data set gets populated. do you want to go ahead and pull up the map and we can start taking a look at that and talking through some of the key findings that your group has come to. Those aren’t only the only data sets that we use, right? So when you think of survey measures those are what are called our direct estimates and those are measured with some error. So we need to bring the strength of up sources to be more confident in the estimates that we’re getting from the surveys. So we use data sources like information from Social Security on people with disability in each county, data from Medicare on people with diagnoses for hearing loss data on how urban or rural a community was, data on the median home value, which is a good indicator of the overall economic health of an area, all to strengthen the estimates and be as competent as we could about the estimates at the local level. There’s another key thing that I’m sure your viewers are aware of, which is that a self reported hearing loss is not the same as a measured value of hearing loss. And problematically, from a statistical perspective, the way people answer that question, yes or no really varies by demographic groups. So for example, a man in his 50s might answer that question very differently than that same man 30 years later who’s come to accept his hearing loss or something like that. so we needed to adjust for that in the estimates. And we used a methodology called calibration to basically translate the self report values at the county level into what we call the currency or the values of the evaluated measures at the national level. Awesome. No, this is fascinating how a data set like this comes together and all the different inputs that it takes to build these kinds of estimates. I think it’s just absolutely fascinating. So let’s do it. Let’s hop into the map and Start taking a look. the whole idea, I think, look and feel of the website was really important to us just because we wanted people to be able to access the data. and then we also wanted the website to be able to house prevention and treatment information. So when you land on the soundcheck website, we have our branding up here. We have a pretty direct statement of what the website is for. So it’s basically we want you to understand hearing loss prevalence and increase awareness. You can then go to two documents right away to either explore the map or to our overall PDF report, which has all the project information in it. If you don’t want to access everything on the Internet we have quick stats on each page which vary based on the type of user. And then we have some icons to guide people to the specific resources or information that might be most useful to them. So we have one for data seekers. We have a page for primary care practitioners. We have one directed to public health professionals, and then we have two directed to public audiences. And then just scrolling down. The website also contains resources including custom made resources by the project that audiologists, primary care providers, whoever can download and use for themselves. So I know everyone wants to just go straight to the map, but those other features if you go back up here and you just click explore the map it brings you to this page, which is a county level map of the United States. I’m going to zoom in a little bit, with darker colors showing higher levels of hearing loss and lighter colors showing lower levels of hearing loss. the number that it defaults to is just the overall estimate for the entire population. But we have the ability to select the level of hearing loss, any hearing loss, mild hearing loss, moderate or severe hearing loss. And the map changes. We can change our age groups. So let’s say we’re interested in people 35 to 64. We can change that and get a different view. let’s say we’re only interested in men. We could look at that. and then we can also look at either what’s called the crude prevalence, which is just the number of people we estimate with hearing loss divided by the population or the adjusted prevalence. And so the adjusted prevalence is a value that assumes that each county has the same population distribution as the United States as a whole. we know hearing loss varies by demographic factors, like so, for example, by age, by race and ethnicity, by gender, et cetera. But if you assume each county was exactly the same in those respects, then where has more or less hearing loss. After you do the adjusted prevalence, you get a different looking map. but one thing that comes out to you from these overall maps either the adjusted or the crude, is the much higher rate of hearing loss in rural areas than in urban areas. this is hard to see. With all the counties, you also have the ability to switch to a state level view that shows you the states. And then the reason why we made the map is, like I said before, I’m relatively new to the hearing loss world, right? So I don’t have all the answers or even close. so what I’m really hoping people will do would be for health professionals with experience in hearing loss, to examine the map and use it to generate hypotheses about what’s going on. and to help with that, we’ve created different overlays of different variables from different data sources that you can place on the map. So most audiologists immediately want to go to audiologists per capita. We can click that, and we can actually see some of the places with the highest rates of hearing loss. Also have relatively low numbers of audiologists. Zoom out just a little. But you can also look at other variables, like so, for example, I’m going to clear that and look at high risk occupations. we can look at socio demographic variables like 8. Like, if you want to see which states have the highest proportion of people age 65 plus. That’ll show you educational attainment and then also some economic variables. so if you click any state like, let’s take west Virginia, for example, it’ll bring you up the value of their prevalence estimate just as, like a quick check. you can also do this at the county level, which I’ll show you in a minute. You can view the details of that state. in this case, we’re looking at the crude prevalence among all ages, all races, all genders, which is 18.2%, which is actually the highest estimated prevalence for any state in the country. gives you our uncertainty around that estimate, which is basically, you can think of that as like a margin of error. we assume that all of our things are measured with a little bit of error, but we’re pretty confident that it’s between these two numbers. and then we have some comparison values to look at. So if 18.2% doesn’t really mean anything to you, like, off the top of your head, you can say, well, how does that compare to the United States? And you get that value there. but you’re not limited to only looking at this overall number. So, for example, if you wanted to look at moderate and severe hearing loss among non hispanic white men and you were curious about it among the 35 to 64 year old age group, you can generate that estimate, too. And then once you have this information, you might say, like, well, why is West Virginia so high? What’s going on there? I’d like to see some more data about that. so we can look at, for example, some economic data, and we can get the poverty rate, the median household income, the unemployment rate, deep poverty, and we can do that for all of these variables that we have listed on the chart. and we created this report functionality thinking someone might want to be writing to their county or their state legislature about the need for hearing prevention and resources. And this is a way to just get all of that information quickly at your fingertips. I’m just going to note that we have this also available at the county level. So let’s take like a particularly high prevalence county in West Virginia. and you can see we can do the same thing with all of the different drop down menus that we just had before. Yeah, I mean, the thing that jumps out at me the first thing is that this will be the backbone of so much future research. I just feel like this is such an amazing resource for anybody that is trying to generate new public policy or build support for some kind of grant, something of that nature or even just initiate some motivation at the county level to say we are 100% times two times more likely in this county than the national average for moderate to severe hearing loss. That’s just very powerful. I can see how this can be a tremendous resource for all kinds of different types of professionals, from people that are trying to enact policy change to the positions that need to be more aware of this in their consults. Yeah. And we try to make the data accessible to different types of audiences. Right. So many people without a data science background will really want to use the reporting functionality that we have here to just quickly get their number. Like say, I want to look at that number and I want to compare it to this number over here. And I just have it at my fingertips. But also for the data science types, we give you the ability to download all these data directly bring them into whatever program you want. They’re all saved as CSV files. And do your own analysis. The hope for this is that people are going to see the data be interested in it, bring their own experience to it in a way that each person has. So, for example, I don’t really know that much about this area of Idaho, but people who live in Idaho do. And I want for folks out there to be able to download their data and compare it to what they know about their area and see if it matches up and see if they can figure out ways methods for prevention and treatment. Yeah, I mean, the map itself, just kind of looking at it a bunch of stuff kind of jumps out, obviously the rural urban divide. but I used to live in Chicago, and that appears to be one of the. Cook county appears to be one of the least. The lowest levels, which is really interesting because, yes, I’m sure one contributing factor to this is like mining, heavy machinery, firearms, but it’s still such a densely populated area that I would have thought that there were higher levels up there. And so I’m sure there’s just all kinds of different ways that you can slice the data and tell different stories. but some of this is just, from a glance, kind of counterintuitive. Yeah. The rural-urban divide is pretty striking and not something that I necessarily expected to see before doing the model. It’s not just Cook county. Washington, DC, has the lowest estimated rate of hearing loss in the country. And places like Atlanta, New York City, Los Angeles also have similarly low rates of hearing loss. in fact, what we found was when comparing the most urban places to the least urban places the least urban places or most rural places have almost double the rate of hearing loss as the most urban places. part of that is demographics, right? So we know from a lot of research that non hispanic whites have the highest rates of hearing loss among any demographic group. and also older people have higher rates of hearing loss than younger people that we’ve known that since time immemorial. so the fact that rural counties tend to be whiter and older is definitely explaining part of the differences. but when we look at the adjusted prevalence we see, like, that’s still not the whole story. There’s still a lot going on in rural areas where even accounting for differences in demographics, rural areas still have higher hearing loss. there’s probably many reasons for this. rural areas have lower access to all sorts of health services, including audiologists, right? And if you think about the self report question that the local variation is based on, if you have someone with untreated hearing loss and you ask them do you have serious difficulty hearing, they’re much more likely to answer yes than someone who’s well adapted and has received treatment and hearing aids or whatever other strategies to treat hearing loss. So that’s one driver. We know that early like preterm delivery for births is related to higher rates of hearing loss. So a baby from preterm birth has about ten times the risk of having congenital hearing loss as a birth that’s carried to term and rates of preterm birth are much higher in rural areas. there’s also likely more people with veteran or military status living in rural areas. And then there’s also potentially greater noise exposures at higher decibel levels in rural counties than in urban areas. so I was talking to a reporter from Montana after the study came out and he was saying it doesn’t surprise him at all, just from his knowledge of folks working with farm equipment and in the mining industry and sort of the habitual long term noise exposures associated with work in those industries, we can see that one of our states with very high hearing loss is Oregon. And I know that there’s a lot of forestry, Oregon but potentially there’s other recreational exposures such as the use of firearms snowmobiles chainsaws, just to clear your own land and things like that. That’s going on in rural areas, that’s not going on in urban areas. another thing about thinking about noise exposure in urban areas would be even if the city is noisy, you’re not necessarily out in the city all day long, right? You’ll go, you’ll get in your car or you’ll get in the subway, you’ll go from your home to your office. You didn’t sit in your office in front of your computer all day. that’s a much different type of noise exposure than working in an industry like mining where you’re basically exposed to high decibels of noise all day long. Yeah, I mean, I almost wonder if from kind of the environment that you’re in as it becomes more urban, is there almost like less social tolerance for loud noise? More or less. I mean, yes, you can’t control some of it, like the l in Chicago or like public transport, but that’s not going to be loud enough for. You’re going to have to sustain that for long levels of time for there to be prolonged damage that’s done. So it’s very striking to me that when you look at the map that’s one of the coolest things about it, is it really tells such a story. and the thing that as you do kind of get to these heavily populated areas in the country, it appears for one reason or another there’s significantly less levels of hearing loss versus the rest of the country. And like you said, there’s so many different contributing factors of why that might be that is so county even specific. Yeah, I mean, I think another thing to take into account would be health status and general weathering of people as they age. many rural areas are quite poor and lack access to a lot of health services. so people suffer for years with undiagnosed or lightly treated diabetes. There’s higher rates of smoking in rural areas than in urban areas. and all of these I think probably from an audiologist perspective, when you think of noise exposure, there’s an interaction between the noise exposure and your general health status. And these two things can interact together, making even the same noise exposure more damaging in one context than in another. One other thing that I want to show people on the website I know everyone goes to the map, but if you go up here to our resources page we have a national indicator report that summarizes the major findings from our study and then also provides hearing prevention, diagnosis and treatment information. for many audiologists, it might be review, but it’s a great document for the public or for primary care professionals to both get access to our information and also have access to some prevention information. And then we also have some project created infographics and materials. Like, we have an infographic on noise levels something that providers could put up in their waiting room in terms of further patients to look at and sort of do a self assessment of their need to talk to their doctor about hearing loss some information about how to prevent hearing loss, how to treat hearing loss. And then one thing that might be interesting is we did some communications research around this project about the messages that are most effective in reaching people and getting them to accept the need to take care of their hearing elf. And we’ve summarized that in an easy to read, two page document that you can download from their website. Yeah, couldn’t agree more with that. Well, this has been a great conversation a lot to think through. but I think that, most importantly, you’ve built a tremendous resource here that I’m very excited to see. kind of the follow on research and studies that happen on the back of this. So, one more time, where can people go to check this all out and access this information? Oh, absolutely. So if you want to go to our website, you go to soundcheckmap.org. It should bring you right here. so just think of, like, sound check. I want to check out sound, and then we’ve built maps. So. Maps. So, soundcheckmap.org. And it should bring you right there. That’s perfect. Well, David, thank you so much for coming on today, sharing all this. I’m sure that there was a lot of time and energy. I know there was a lot of different people involved with this, so I think it’s a really awesome thing that you’ve all built, and I’m looking forward to seeing what comes next when people now have access to this kind of information at the level of granularity that you’ve provided. Fantastic. Thank you so much for having me on your show. I really appreciate it. All right, everybody. Well, thanks for tuning in today. We will chat with you next time. Cheers.