Improving Registries and Representation in Alzheimer’s Disease Research

Transcript

Intro: I’m Dr. Nathaniel Chin, and you’re listening to Dementia Matters, a podcast about Alzheimer's disease. Dementia Matters is a production of the Wisconsin Alzheimer's Disease Research Center. Our goal is to educate listeners on the latest news in Alzheimer's disease research and caregiver strategies. Thanks for joining us.

Dr. Nathaniel Chin: Welcome back to Dementia Matters. I'm here with Dr Josh Grill. Josh is an associate professor of Psychiatry & Human behavior and Neurobiology & Behavior at the University of California, Irvine. He is also the director of the Institute for Memory Impairments and Neurological Disorders, also known as UCI MIND. Very great name and acronym for you. In August 2021, Dr. Grill published a paper in The Journal of Prevention of Alzheimer’s Disease which looked at a popular tool in research recruitment – registries. Looking at how registries represent different neighborhoods and demographics, Dr. Grill and his team found that people in more disadvantaged communities may be underrepresented within these registries leading to an increased risk that these communities will be underrepresented within the research itself. Here to talk more about his study is Dr. Josh Grill. Welcome to Dementia Matters, Josh.

Dr. Josh Grill: Thank you so much for having me Dr Chin

Chin: I would like to start with some basics because I'm not sure our listeners really understand registries and how important they are. So if you could talk about, what are recruitment registries and how can they be helpful in Alzheimer's disease research?

Grill: Sure. I think it's probably important to begin by saying that one of the most consistent challenges we face as a field to trying to learn and advance clinical practice and including diagnosis and treatment is that doing research studies involving people with Alzheimer's disease, or even people without it, is always a difficult challenge to recruit a full sample. Most studies are delayed by the challenges in recruiting a full sample. Then, when we do recruit a full sample, it's pretty rare that that sample truly represents all the people with disease, all the people we're trying to help with these studies. We've, for years, been trying to study recruitment and produce an evidence base to try to get better at it, both doing it faster and more efficiently and doing it better by including people who represent all the disease sufferers out there. There aren't a lot of great tools to make it go faster. I joke sometimes that my career has been about identifying the many challenges, but we do need tools. We do need to identify ways to get better. One tool that is quite promising is these recruitment registries. A registry is based on the idea that typically what we would do in a new study is we'd work really hard to get the funding for the study, design the study, get it IRB approved and then say, ‘Okay, let's go find some people to be in this study.’ Wouldn't it be better if we had a very long list of people who had already said, ‘If there's a study that's right for me, let me know.’ So instead of serially recruiting people one at a time or once a week or, in many clinical trials, one per month for years, what if we could just invite 100 people or 200 people who all might qualify for that study and try to quickly accrue the sample. This is how registries are intended to work – large lists, databases, repositories of individuals who are open-minded, at least, about participating in research studies and let us try to recruit very quickly, very efficiently, and in some cases very strategically to studies, so that instead of our 12 months study taking three years to recruit and then twelve months from that point forward to be finished, let's recruit over 12 months and take two years to answer our 12-month question.

Chin: I mean, it seems very clear that registries are very important. They can accelerate research to some degree, but also people can change their mind, right? You can enroll in a registry. I imagine sometimes people either don't respond when contacted or want to leave. What do people do once they're in the registry or what do scientists do to maintain that relationship with people who have said, ‘I would be interested at some point?’

Grill: Yeah, I think that's a really important point and one that we've been putting a lot of attention toward. I think everyone agrees that there's potential power in these tools. I've written in manuscripts, ‘bigger is better,’ right? The more people we have in a registry, the more people we have to try to recruit quickly, but then there are differences in approaches around registries. For example, our registry, we refer to it as a local registry. We have about 5,000 people from Orange County, California who are all adults and have all consented to be in our registry ,basically saying, ‘Yeah, if there's a study that might be right for me, let me know, and I'll think about whether I want to be in it.’ That's what we try to sell them on, but when we enroll them we do two – what I think are very important – things. First, we collect a lot of data about them, their demographics, their medical history, the medications they take with an eye toward contacting only people we think are going to be eligible for a specific study, about that study. We don't want to reach out to someone and say, ‘We have a new study. It's for people 55 to 80,’ then all the fifty-year olds are saying, ‘Well, I don't want to be – I can't be in your study, so why are you contacting me?’ So we try to be very careful about collecting data that helps us get efficient on who to invite. The other key thing that we do when people enroll is we ask them nine questions about their willingness to engage in studies that require specific research procedures. Are you willing to have an MRI? Are you willing to have cognitive tests, blood draws, PET scans, lumbar puncture, etc. So we only want to reach out to people who are eligible and, a step further, we only want to reach out to people who are eligible who we think are likely to say yes. Anytime we are recruiting to a new study or anytime an investigator comes to us wanting to recruit from our registry, we mandate a query be performed of our registry to match people. I sometimes refer to our registry as a dating service between the adults of Orange County and the investigators at UC Irvine trying to to make a perfect match to help science move faster. We think this works. Our registry has referred more than 5,000 people to studies here at UC Irvine. As you point out, some people don't answer the email or respond to the calls. Some people say, ‘No I'm not interested in being in a study right now,’ or, ‘I'm not interested in being in that study.’ But so far more than 30% of the people we've referred to be in studies have actually enrolled in those studies. The screen failure rate of people referred to studies who get past a single phone screen – the screen failure rate is 1%. We are confident that we've moved the needle locally and that this model seems to really be pretty efficient in making those key matches.

Chin: And so, you mentioned in your response a couple of words that I think are key. One, consent, getting their consent. Then also, getting a lot, well not a lot, but a good amount of personal information. Knowing that and knowing the stigma that can exist in Alzheimer's disease, Alzheimer's disease research in our current times, I imagine people listening are wondering, what about protection. How are you safeguarding this registry? Do I need to be worried that I'm on some sort of list? How do you address that?

Grill: Yeah. Our registry is completely self-reported information. We don't tap into the medical record and, as yet, we don't get biomarkers or even test people's memories in our registry. We do have a scale that assesses their subjective performance on memory and cognitive tasks that is useful in our research and has actually been useful in using our registry as a data source for additional research, which I could talk about as well. But the registry, of course, like all forms of research does have some risks. They're mainly loss of confidentiality, although we have firewall, carefully protected servers that house this data. People do give informed consent and we describe that risk. Really the risk of unwanted emails or phone calls from investigators desperate to fill their studies, but that's really the big, main risks. We, importantly, don't really limit our registry to use in Alzheimer's disease studies. We enroll any adult who wants to be in the registry and we have made our registry available to investigators across all areas of science here at UC Irvine. People have participated in hearing studies, skin cancer studies. We've got an investigator getting ready to try to study post-chemotherapy cognitive performance. There's some studies ongoing about drinking water, so targeted recruitment of people based on where they live. So, you know, it really doesn't have that risk of stigma because it's not exclusively an Alzheimer's registry, and only about 20 to 25% of people in our registry have a family history of Alzheimer's disease. Now it was built and powered by a group of Alzheimer's investigators, so the median age of people enrolled in our registry is 58, a little more a little older than the county as a whole, but we really try to partner with people from all areas of research happening here at UC Irvine and let them use this as a tool or, even if they're doing community outreach, let them use it as a way to capture people they're reaching so that they can recruit them to studies later on.

Chin: What are the limitations of registries, and then specific to our topic today, how can they perpetuate health disparities?

Grill: Yeah. I think we're still learning the limitations of registries. The evidence base around their effectiveness, the degree to which they actually accelerate science or their cost efficiency remains an open area of study. I think we and others have tried hard to get papers in the literature, let others be guided by our successes and failures. You know, I'm thinking of the Brain Health Registry, and the Alzheimer's Prevention Registry, and the Alzheimer's Prevention Trials Webstudy. There are several large national registries that have done an outstanding job of putting into the scientific peer-reviewed literature their experiences, their successes, and the like, but I think there's still a long way to go. There are different models that are being used, some that incorporate cognitive testing online, some that allow for remote, genetic testing etc. So what's going to be the number one way or or the most effective approach to this, I think, is something we need to keep trying to figure out as a field. And again, we know that some of them are really designed around prevention trials for Alzheimer's and ours is taking a broader approach but, as you point out, there are some other limitations that we have to think about. We know that research as a whole doesn't always do a great job of being inclusive of the diverse populations that make up our nation. I don't think we really know the extent to which registries can be a tool to help us overcome many barriers to being more inclusive in our prospective in-person longitudinal research. Can we use registries to not only accelerate recruitment – one of those two main things I said was a challenge – but also to better recruit populations that represent all the people we want to help. I think there's some hints that maybe they could be, but also some hints that if we're not careful or very purposeful in the way we use registries that we run the risk of perpetuating or perhaps even worsening some of the challenges with representativeness in research.

Chin: So, now in your study you looked at diversity of the UC Irvine Consent 2 Contact registry. How are you defining diversity in this, and then what did you measure in this registry?

Grill: Yes. We call our registry the UCI Consent 2 Contract registry, or C2C. I like the palindrome and we are, here in Southern California, close to the ocean so I like to remind myself and others of that fact. In the C2C, we have put effort into trying to recruit a diverse and representative population. While I am at once proud of the diversity represented in the C2C, I'm simultaneously aware it's not representative. I knew that before the study that we're going to talk about because it's about 25-30% composed of non-white races or Hispanic ethnicity, and actually Orange County, California already has no majority when it comes to race and ethnicity, so we've got a ways to go to truly represent our local communities. In this particular study, we wanted to characterize the people who were enrolled in the C2C even beyond their race and ethnicity as they reported it to us when they enrolled and looked instead at what's known as the Area Deprivation Index, which has been really championed and made available to researchers nation and worldwide in part through the leadership of Dr. Amy Kind there at University of Wisconsin. She has just been moving the field at leaps and bounds by using this tool showing others its value and making it easy for investigators like me, who might be two thousand miles away, to also use it. What we did with Dr. Kind was explore through her data democratization tool, we took all the folks who had home addresses in the C2C and we examined what ADI strata they lived in – all the folks who were enrolled in our C2C registry – and whether more or less advantaged neighborhoods in Orange County might be more or less represented in the C2C.

Chin: So, Josh, what did you find in your study?

Grill: Probably to the surprise of no one, we found that the C2C was overly representative of our least disadvantaged neighborhoods here in Orange County. So about 70% of people enrolled in the C2C came from the 20% of our most advantaged neighborhoods, and only about 13% of our most disadvantaged neighborhoods were represented. I should say, only about 13% of participants in the C2C came from the most disadvantaged neighborhoods. We think this is key. I think work by Amy, work by others in the field like Lisa Barnes and Crystal Glover and many, many other tremendous investigators have begun to show us that race and ethnicity is only part of the story. Race and ethnicity are important. They're, of course, social constructs and other social determinants of health like where we live and our access to healthy foods, healthcare, etc. These things may matter to our health and they may matter to our risk of getting Alzheimer's disease. We took this data as an imperative to try to do a better job one, at recruiting people from the more disadvantaged neighborhoods, but also studying how to do that. We really don't know if certain methods or approaches might be better at reaching people in these more disadvantaged neighborhoods compared to the more advantaged neighborhoods, which apparently we do an okay job of.

Chin: With this information, what are you now investigating as far as techniques to reach these underserved communities?

Grill: Now, we're kind of taking the next step, again in close partnership with Dr. Kind, to do an experiment – a multi-year experiment, gratefully with support from NIH – to test different interventions, if you will, for trying to recruit people from the various strata of ADI, the various levels of disadvantage, and understand whether certain interventions, certain approaches to recruitment may work better in certain communities. So we'll be doing that starting this summer actually and looking at community outreach, which we've always done and we can do geographically targeted, but also things like Facebook advertising. The ubiquity of smartphones makes it possible to try to reach people everywhere through social media and other platforms. We know that more than half of people who enroll in C2C do so on a handheld device, but we're also going to look at things like postcards and sending postcards to specific neighborhoods where, you know, we can therefore target the level of disadvantage. I think – actually, one good thing from the pandemic is that it might help this study because now everybody knows what a QR r code is and is used to using them to order food and whatnot, and so postcards can have QR codes on them and and people don't squint and wonder what it is. So we're very excited to get this underway for the dual purpose of trying to do a better job on our own efforts and a registry and to increase the representation of people from diverse communities, but also to try to help others with the evidence that we'll collect about what works best, and what may not work best and shouldn't shouldn't be how you spend your resources.

Chin: So Josh, we've had Dr. Amy Kind on this podcast – and so for listeners, I would direct you to an earlier episode where Dr. Kind explains the Area Deprivation Index – but I'm wondering, Josh, can you explain to us or provide an overview of how you use this tool or what you were specifically looking at with the metric?

Grill: Sure. The ADI, of course, is a very powerful tool for assessing a composite, if you will, of socioeconomic status and neighborhood advantage. We use the full ADI to examine the strata of ADI in our registry. We found some things that make sense with what we know about ADI. There were associations between ADI and race and ethnicity, and so people in the more disadvantaged neighborhoods were more often from non-white races or Hispanic ethnicities. I think one thing that is really important that we found was that while the demographic associations were what we predicted, again going back to the way we design the C2C, we could also look at whether people from more disadvantaged neighborhoods were equally or more or less willing to participate in research. We found no evidence to suggest that people who were from more disadvantaged neighborhoods were any less willing to participate in research studies, be that whether it involved MRI, investigational drugs, autopsy, or lumbar puncture. There's a clear sample bias we have to think about here. We have people who were willing to be in the registry in the first place, but it does suggest that there are people in these neighborhoods who are recruitable, that they want to be in studies, that they want to help investigators make a difference. I've long held a hypothesis that registries are a relatively low-burden ask. You know, there's not a lot of risk. You don't necessarily have to give blood or saliva. It might just be giving some data. Even if we had found differences in willingness, I would have been a huge proponent for trying to increase representation from these neighborhoods and registries because then they become – forgive the term – a captive audience for us to try to change their attitudes, make them more willing, help them better understand the various levels of protection for people when participating in research, the importance of our mission, and our desire to help all people. Based on this preliminary evidence, we don't need that added level of attitudinal change interventions to change attitudes. They're already willing, but I do think that we need to keep targeting these underrepresented communities and use registries as a place to try to increase people's willingness to participate. I've even thought about, what we call around here, gateway studies. If someone participates in an interview or a survey study that is not a big ask, doesn't have a large commitment or risks or burden, will they have a positive experience in research and perhaps help gain more trust, make us more trustworthy, and make them more willing to participate in something that might have greater risk or burden down the line? That's a thing that could be empirically and examined over the years through C2C and other registries and I'm eager to try to test that hypothesis.

Chin: Yeah, I'm so glad you use the word trust because as you were saying that to me, I thought to myself – well in many communities, there isn't the trust or the energy that has been put into that community, so doing something that is low burden, that's sort of an intro into what is research, can build that trust. It can also show the relationship and the people that are involved in it. I'm really glad you brought up that willingness to participate piece because I really enjoyed reading that in your publication and I do think that's really significant. If I could summarize and then ask you a question. So people from more disadvantaged neighborhoods are less represented in research but, based on what you found, they report the same willingness to participate. Why this disconnection between interest and actual participation?

Grill: Well, the barriers to making research truly representative are many. We have to attack all of the barriers simultaneously. I don't think improved diversity and research is as simple as ask more. We need to ask more, but we need to gain trust. We need to ensure that our research teams are diverse and that people have the opportunity to see people like them in the research setting. We need to make sure that our designs are thinking about the many communities from which we wish to recruit, and that we minimize the burden of participation as much as possible, and that we remember people from disadvantaged neighborhoods may have different jobs or the need to take the bus to the research site instead of driving their own car. We need to remember that not everyone speaks English as a first language and consider the language barriers either to recruitment or to participation. We need to think about the fact that people from disadvantaged neighborhoods may not be able to afford to take off from work to be in our study. That may mean we need to offer visits at the evening, on the weekend, or that we should pay people to participate in our studies. I am a staunch proponent of compensated people for being in our studies and there's a very nice paper in the New England Journal of Medicine – Gelinas et al – a few years ago that sets up a structure for thinking about financial payments to people enrolled in research. We can reimburse people so that they're not incurring cost to participate in our studies. We could compensate them, meaning we could pay people a fair wage, like minimum wage, for the efforts that they put into our research advances, or we could incentivize them. We could pay them more than a fair wage, which really we need to do if we want to change people's minds. There are people who are as eager for us to make advances as we are and are waiting for us to do a new study so that they can participate, but if we want to get more people who rarely are in our studies to be in them, maybe we have to think about financial incentives and other incentives and making all of those other changes that I brought up so that we can do a better job of including those populations, those communities, those different groups.

Chin: Well, thank you, Josh, for sharing that last answer. I feel like that's going to be a teaser for a future podcast that we do with you and as well as the others experts in the field. In many Alzheimer's disease studies a study partner is required to participate, regardless if you have a thinking impairment or not. Is this a potential barrier to enrollment? Are there differences among people from different backgrounds or communities when it comes to study partner involvement?

Grill: So, it's absolutely a barrier to recruitment. At minimum, the logistical burdens of recruiting two people are greater than the logistical burdens of recruiting one person. It is a burden. It is a barrier to recruitment, but it also is an important aspect of how we conduct our science. We call these people study partners now. We used to call them caregivers. In our protocols, in our recruitment materials, we call them caregivers. As the field has evolved to recruit people earlier and earlier in disease, and now even people who we think may be at increased biological risk for disease but have no cognitive problems, that nomenclature change was important to refer to these people as study partners. Even in the space of prevention trials, we believe this role is key on numerous levels. The data that they provide still is important to estimating functional changes over time and estimating cognitive changes over time. We've done a bit of work to look at who's the better source of information about cognitive performance. Participants themselves are better at predicting their own cognitive performance at the beginning of a prevention trial, but by the end of a study the informants or study partners will typically have greater correlation or accuracy for cognitive performance. The role is greater than that. Ultimately when people have symptoms, or let alone dementia, we rely on these individuals for information about adverse events. We may rely on them to help ensure compliance with the study, be it taking a medication or getting to visits. Again going back to even this prevention space, we recently did a study asking people who could enroll in a prevention trial. Their attitudes towards the requirement of having a study partner – and while certainly there were people who said, ‘Oh, this is a big problem for me, and we know it to be true. There are some people who say I simply don't have someone who could be my study partner. We actually found that much more frequently people said, ‘Oh yes, I wouldn't want to do this without my study partner.’ For some people, they recognized the value of the information that they would provide. For others it was, I want support there when I do this. In particular when I learn my biomarker results for Alzheimer's disease I want my loved one to be there with me. Now you brought up another really important point, and we don't yet know enough about whether this is the same in all communities, all cultures, all groups. In the A4 study, which was really among the first and certainly the first very large preclinical Alzheimer's disease study, the racial and ethnic groups differed in the frequency of the study partner types they enrolled with. So non-Hispanic whites overwhelmingly enrolled with spouses as their partner, but as you move to Asians, Hispanics, and African Americans, the frequency with which people enrolled with a spouse sort of steadily declined, all the way down to 27% of African Americans. So we're starting a new study now to try to enroll each of those four groups – not in a trial, just in an interview study – to understand racial and ethnic cultural predictors of availability and attitudes towards the study partner requirement. I'm very eager to explore this. Maybe it'll have to do with racial and ethnic groups, and maybe it'll have nothing to do with that whatsoever but instead have to do with the neighborhood they live in or other social determinants of health, which we are going to try very hard to measure somewhat comprehensively. And so I think, stay tuned. This is a double-edged sword. It's key to much of our research, but it certainly does produce barriers and perhaps those barriers are disproportionate in their impact depending on the communities we're trying to recruit. As we learn more about this, it may be that we need to further adjust the way we design our studies, the requirements we place on our study partners, the manner in which we incentivize them to participate – because we should incentivize them as well – or the manner in which we allow them to fill their role and provide the valuable data that we ask of them.

Chin: With that, I'd like to thank you for your time and being on our show today. I anticipate we're gonna be having you on again.

Grill: Thanks for having me Nate. I'm always happy to be on your podcast. I really appreciate the opportunity. Let me say, thanks for the good work you're doing and all the people that you're reaching. I think we as a field are working hard to move away from living in a world that stigmatizes the disease we study and your efforts are key in achieving that.

Outro: Thanks for listening to Dementia Matters. Be sure to follow us on Apple Podcasts, Spotify, Google Podcasts, or wherever you get your podcasts to be notified about upcoming episodes. You can also listen to our show by asking your smart speaker to play the Dementia Matters podcast. And please rate us on your favorite podcast app -- it helps other people find our show and lets us know how we are doing. Dementia Matters is brought to you by the Wisconsin Alzheimer's Disease Research Center. The Wisconsin Alzheimer's Disease Research Center combines academic, clinical, and research expertise from the University of Wisconsin School of Medicine and Public Health and the Geriatric Research Education and Clinical Center of the William S. Middleton Memorial Veterans Hospital in Madison, Wisconsin. It receives funding from private university, state, and national sources, including a grant from the National Institutes of Health for Alzheimer's Disease Centers. This episode of Dementia Matters was produced by Rebecca Wasieleski and edited by Caoilfhinn Rauwerdink. Our musical jingle is "Cases to Rest" by Blue Dot Sessions. To learn more about the Wisconsin Alzheimer's Disease Research Center and Dementia Matters, check out our website at adrc.wisc.edu. You can also follow our Facebook page at Wisconsin Alzheimer’s Disease Research Center and our Twitter @wisconsinadrc. If you have any questions or comments, email us at dementiamatters@medicine.wisc.edu. Thanks for listening.