Dr. Erin Talati Paquette - Applying a Racial Equity Lens to Research: The Importance of Engaging Diverse Populations

I'm Marylyn Stein, one of the attending and sociologist in the main OR. It's my pleasure to introduce Dr. Aaron Piquet, who will be speaking this morning. Dr. Piquet is an assistant professor of Pediatrics at Northwestern University Fineberg School of Medicine. Assistant Professor of Law by Piquet City at Northwestern University puts her school of law, as well as chair of the Ethics Advisory Board, Associate Director of Clinical and Organizational Ethics, an attending physician and pediatric clinical care at Lora's Children's Hospital of Chicago. Prior to showing in the faculty at Northwestern, she was a teacher undergraduate degree from Northwestern University before completing her MD, JD, and Master's in Biographics at the University of Pennsylvania. Dr. Piquet completed her residency of Pediatrics at the University of Chicago, and she will be familiar to many in the audience today from her time as a clinical fellow in pediatric clinical care medicine, here at UCH, as well as her research fellowship in the Center for Outcomes and Policy Research at Dana Farber Cancer Institute. Dr. Piquet is interested in research advocacy and policy tools to reduce health disparities, address bias, racism, and other structural determinants of health, and promote social justice. Her research includes evaluating disparities in research and enrollment and participation, the use of medical legal partnerships to address the social determinants of health, and evaluating healthcare access in relation to health outcomes. Dr. Piquet will be speaking on applying a racial equity and lens to research, the importance of engaging the verse population. Her talk this morning is part of an ongoing series highlighting issues of diversity, equity, and inclusion, created by the EDI task force in collaboration with the Grand Round Leadership. Please join me in welcoming Dr. Piquet. Thank you for that very lovely introduction. And it's really great to be at least virtually back on my old stomping grounds. I'm very happy to see a lot of familiar faces in the Zoom audiences morning. Okay, so before we get started, just wanted to give a few disclosures and disclaimers. My funding is listed here. The talk today is supported by my work through the NICHD. I also have funding to the American Bar Foundation and through HRSA, though those don't pertain to what we'll be discussing today. I also want to start with a brief disclaimer in that references to diversity by race are really a function of available data with recognition that raises a social construct and differences that are seen are the result of racism or other social conditions that are so seated with race, which acts only as a proxy for them. As we begin to think about issues of equity, diversity and inclusion in a particular of some of the work that I'm going to talk about today, I like to keep in mind this quote by Reverend Martin Luther King that I think really encapsulates what we're trying to break down in terms of barriers to good health. And decades ago he stated that of all the forms of inequality, injustice and health is the most shocking in any human. I think if we keep that in the back of our minds as we execute our work in equity, diversity and inclusion, we're bound to end up on the right side. I also want to frame our discussion before I begin to talk about diversity in research and how to maybe address that barriers to diversity. I want to take a moment to situate where we are historically, particularly in medical and research abuses against black Americans. And black, actually, black Germans as well. We often think of the Tuskegee Cypheus study as being the marker of what we have called mistrust or distrust among the African-American community and other minority communities. And we often attribute black of participation in research to this distrust. But the Tuskegee Cypheus study was really in the late to end of a series of injustices and medical abuses against black people. Beginning with in the 1850s a repair of pro-Lafs Udari and black slaves without anesthesia, moving into the 1930s and 40s where African-American children in Germany were underwent forced sterilization to improve the purity of the area in race to the 1950s when the healer's cell line was created without the consent to tendriata laks, which many of you will be familiar with. Under the 1960s where in Cincinnati, there were total body radiation exposure experiments conducted under the Department of Defense in order to learn more about the effects of total body radiation on both cognitive and physical outcomes on people. But these were conducted on 60% of the population on which they were conducted. Was African-American and the majority of black were of low education. We then moved into the 1970s when the Tuskegee Cypheus study ended. There we saw other research abuses in that decade. There was a child care program. Had Hopkins in which individuals were enrolled and thought that they were receiving child care. There were some blood draws done where parents of children were told that this was for routine evaluation of for anemia and other childhood issues. But in fact, they were looking for whether there were differences and predisposition to crime where there were extra-wide chromosomes, so individuals who were XY versus XY. And this was done again without the knowledge of poor consent of the parents. It was during the decade that the Belmont Report, one of the foundational documents for research ethics, was also developed in which we established and key principles for research ethics, from including respect for persons or autonomy. Beneficence are doing the best or benefiting individuals wherever possible. And justice rose as a key concept. Even with that document developing in the 1970s and movement to develop federal regulations to govern research in the 1980s and the 1990s, we still saw abuse against low-income black and Hispanic individuals where there was a measles vaccine that was evaluated without consent largely in black and Hispanic population in infants and where there was a study with breaking a sealed juvenile records to evaluate propensity towards crime at Columbia University where the majority were these sealed records excluded white individuals. So you can see from this timeline, hopefully, that we're not talking about a single point in time. The lack of trust or the broken trust that exists with minority communities spans more than a century, almost two centuries of time that we need to contend with when we're thinking about how we begin to repair some of these abuses. Okay, I spoke on a previous slide that slide that Justice Rose is a critical concept on the Belmont report. And I think as we think about addressing barriers in research participation, trying to figure out how to maximize concepts of justice are important. Under the Rawlsian concept of justice, there's the notion that the distribution of burdens and benefits within society or across any intervention in society ought to be fair and also equal. And however, within this principle, Rawls also articulated that if there are to be any inequalities, these inequalities should really favor the least well off rather than falling to those who are better off. So as we think about how we maximize our opportunities for participation of diverse groups and research, we ought to keep those justice principles also in mind. So why is it problematic if we have inadequate participation in research by diverse groups? And three reasons are that it really threatens study feasibility and can lead to premature study termination. Secondly, it threatens study of validity and generalize ability because the study findings may not and truth be applicable to the entire population if there are race or other factors for which race acts as a proxy that might impact how individuals would respond to study intervention. And finally, if we're not equally enrolling or proportionately enrolling, we may be unfairly distributing burdens or benefits, which either may exploit if we're over representing a given population in a particular study or create a lack of benefit for that population in the study which may be provides benefits and majority of participants. So we're really looking for proportionate representation. And I'd like to start by looking at some of the adult data regarding participation, which shows us that there are racial differences and under-representation of minorities in clinical trials. When 98 clinical trials were looked at, 18 found differences. And these differences were relevant, 50% were related to pharmacokinetic differences, 39% to efficacy, and 11% to safety. So when we think about how do we apply clinical trials findings, it's important to think about the fact that there may be differences in how drugs are processed, how effective they are and how safe they are based on race or other things for which race represents. We also know that in the adult population, there's a lack of diversity in genomic studies. So this reflects two studies. To the left, we're looking at the numbers of association studies by disease and study population demographic group. So the large bar here to the left represents the European population or largely white population. The Asian population is next most represented to the right. And finally, underrepresented minorities are in this bar to the far right, which you can see is at a much lower rate of studies being conducted than any of the other races. Pope joined all of the study that they repeated between 2009 and 2016 to see whether any of the attention to lack of racial diversity and research had led to change. They really showed very little change. In 2009, 96% of the studies were of European ancestry. Those studies did decrease to 81%, so 15% increase in other populations between 2009 and 2016. But when we look at the breakdown between those populations, this lighter blue bar represents the Asian population, that the slightly darker aquamarine bar shows other minority populations. And we can see that the majority of the increase between 2009 and 2016 is really owing to an increase in those of Asian ancestry. Those are also represented by this larger segment, lighter blue bar here where the remaining diverse populations are in these smaller bars at the bottom, which we can see have not increased much in representation over this period of time. So our first conclusion from this data is that diverse adults are underrepresented in clinical trials and genomic research, but it doesn't tell us much about how diverse children are represented in research. We conducted a study looking at a pediatric biobang. We collected opt-in consent forms that were given out at the time of PICU admission to over 3,400 encounters between late 2014 and 2017. And I'm showing you the results of both those who did not complete their form, which we showed varied by race ethnicity. We showed the odds of returning an incomplete form were higher for every non-white race. So Hispanic, Black, and Asian races all had higher odds of not completing their consent form. Similarly, we looked at those who discomplete the form, but refused consent to participate in the pediatric biobang on their form. And we also found that except with respect to those who were marked as other in our electronic health record, all other races also had a higher odds of refusing consent to participate in pediatric biobang when given the option. So our second conclusion is that diverse children are underrepresented, at least in a single center study in genetic research and probably more work needs to be done to understand that on a larger scale. So if we've taken a look at participation of children in biobank research or genetic related research, what does that say or do we know anything more about participation of children generally? In order to study this, we look directly, respectively, at nine studies in the critical care and emergency department through two large research networks, the critical care pediatric research network, and the pediatric emergency care applied research network. And through looking at those studies, we were able to evaluate more than 60,000 individual patients through both interventional observational studies with variable consent mechanisms to take a look at whether there is different participation by race and ethnicity, which was the only socio-demographic marker that we had available to us. Because we looked at this retrospectively, we had to be somewhat creative about how we characterize participation by race and evaluated whether there were differences by race ethnicity. We only had enrollment numbers for individuals that actually agreed to enroll in the study. At the time that most of these studies were conducted, studies were not collecting the most precise estimate at the denominator, which was the eligible number of study participants. Ideally, you'd like to compare those who enrolled against those who were eligible, but we did not have that data available to us. I've put a chart or a diagram on the right to show what were the different potential denominators we could have looked at, and the red highlights the data that was available to us. So we could have looked at those in order to make the denominator more precise if we didn't have eligible data. We could have looked at the rate of breakdown of race and ethnicity by study institution. Again, we didn't have that data. Nor did we have the general population at the study institutions. What we did have was the network data, both by study diagnosis, as well as the general network data at the total institution population for the network institutions. So we used those to do our primary analysis for this study. And we validated that, which I'm not going to show you the validation data, but we validated that against the FIS database, which collects general population data for pediatric institutions. We felt that being able to look at this data was slightly more precise than some preliminary data, or some prior data that had looked at participation of children and trials against the general census. So this is just another chart to show you what we did. We looked at a participation proportion. So this was the number enrolled in the study by race and ethnicity, divided by the number of those individuals by that race and ethnicity in the two different denominators. The first being the institution method, we used core data projects for each research network, the PCURN core data project, and the CAVQN core data project. And this was the total population at the network institutions. And then a second method that we used for fewer institutions for which, for fewer studies for which we had the data was a disease prevalence mother, which we used the core project data, but we restricted it by the ICD-9 code of the disease of study to get a slightly more precise estimate. Okay, this is showing you that data for one study, the biosegniture study, just to orient you to what we did on the left axis, you're seeing race, on the right axis, you're seeing percentage participation by race. And the hatched marks show you rate of participation or rate of enrollment in the data project versus the dark black bar, the study itself. So in this study, the dark black bars show that Hispanics were overrepresented, what relative to the network institutions as were the non-Hispanic white population while non-Hispanic blacks were underrepresented. When we restricted that by individuals with sub-ryle neonates, which was what the biosegniture study looked at, we found the same trend that the Hispanic population was overrepresented relative to the core, whether it relative to the network population while the non-Hispanic white population and black population, non-Hispanic black populations were overrepresented and underrepresented respectively. And all of these findings were significant to a p-value of less than 0.001. Because we looked at a lot of studies and not going to show you that data individually, we compiled a composite graph to look at over an under-representation, over-representation is when the rate of participation in the network institution was higher than that in the study representation. Under-representation is when that population in the study was higher than that in that network institution. So across the multiple studies, we can see that there's both over and over under-representation. I'm pointing out here the representation of non-Hispanic blacks. You can see arrows pointing to their representation across studies both above and below the zero point demonstrating that they're both over and underrepresented. This is similar in the Hispanic population. When we get to the non-Hispanic white population, they were largely overrepresented, excuse me, overrepresented, both by the institution method, as well as when we restricted them to the core data projects by disease. When we looked at non-Hispanic black population by disease, they were largely underrepresented, and Hispanics were both under or were overrepresented by the disease method. Again, for the disease prevalence method, we were only able to look at three studies. And so I think that all of this data retrospectively demonstrate that there's likely a signal in that, we believe that there is disproportionate representation by race and ethnicity across different types of studies, but really the limits of the retrospective data suggest that some prospective work needs to be done. But what we can draw from this is that diverse children were largely likely both under and overrepresented in clinical research. And as we mentioned, this can be problematic if they're overrepresented in the types of research that could lead to exploitation and underrepresented in the types of research that could lead to benefit. So we started trying to look retrospectively about whether we could say anything about study characteristics and whether race enrollment varied by these characteristics. So here, you're looking at mechanism of consent across race, and we can see that indeed the mechanism of consent is associated with differential race participation. If we walk down this chart, we looked at the odds of representation of non-Hispanic blacks compared to whites, non-Hispanics compared to whites, and other populations compared to whites in that population of 64,000 children across those network institution studies. And we found that in the non-Hispanic, I'm sorry, in the studies which had written consent or us sent, so less formal mechanisms of consent where say signatures or documentation were not required, there was a lower odds of participation of black versus white individuals. However, in the studies that had, I'm sorry, I said, I think I said that wrong, where there were more formal mechanisms of consent, we had a lower odds of participation, where there were less formal mechanisms of consent with verbal consent or waiver of consent, we had higher odds of participation of a black participant versus a white participant. When moving to the Hispanic and other population, those trends went in the complete opposite direction. So for more formal mechanisms of consent and where us sent was required, there were higher odds of participation of Hispanic and other populations, where there were lower odds in the less formal mechanisms of consent. And I think these can be, there can be many hypotheses that really need to be tested for why this may be the case, but one might imagine that those who are Hispanic or other by race ethnicity might also be of limited English proficiency, which we didn't have the data of. But those individuals might be less likely to be able to consent for verbal and study populations, maybe less likely to approach them, worried about language barriers for studies where there was a waiver. Whereas with the black population, they are maybe more hesitancy when there's more documentation or formality attributed to their participation when there's less formal involvement. We also saw that there was differential participation by type of study, where there was higher odds, I apologize for the shading here, the blue shading didn't come out properly. There was a higher odds of having a black or other participant in an interventional study with, then there was a Hispanic participant in the opposite when you looked at observational studies. And so we need to think a little bit about with some of those interventional studies where the interventions were actually potentially harmful to minority populations, historically about what is it about participation in intervention studies that may promote participation by minority individuals, and are they being overrepresented in an inappropriate manner in interventional studies while representation in observational studies is not as robust. So that data was data from the work that we've done in my lab looking at race and ethnicity and participation in clinical trials. But I want to point out this study that looked at recruitment strategies that really highlighted that race and ethnicity are only one factor for successful recruitment of groups by diverse characteristics that we may want to try to promote inclusion of in clinical studies and all observational studies. So in this study, the dark gray plus shaded recruitment strategies were positive while the negative or the minus sign represent strategies that were not successful. And you can see that there were positive and negative strategies that varied by age, gender, race, ethnicity, education, employment, and income. So many of the social influencers of health and differential types of recruitment and introduction to study whether you were using an electronic health record, flyer targeted or moving into the community up to direct recruitment within clinics. There was variable positive uptake of those different types of recruitment. And when looking at race and ethnicity in particular, these also varied that different recruitment strategies. It wasn't that for race and ethnicity minorities in this particular study that I looked at, the Latinx population. And it wasn't like for this population that there was one strategy in particular that worked that also these strategies also varied. So we can't take a singular approach. And this brings to the conclusion that the above data, the different recruitment and consent strategies might be effective for different favors populations. And as well as the fact that different study types may suit individuals from different race, ethnicity, or other social demographics marker backgrounds differently. Just briefly before we move on to looking at some of the ways to address barriers to enrollment of diverse groups. I want to talk about a study we did looking at factors associated with retention and clinical studies. And again, what I went up point out very briefly about this work is that we looked at a study that at a single center that enrolled individuals in an intervention study for support intervention for their time in the PICU. And these individuals were surveyed at the time of their approach at discharge, three to five weeks post discharge and three to five months post discharge. And what we looked at was the rates of study return by a number of sociodemographic factors at these different time points. What I'd like to highlight is less important than the fact that we can see that there are multiple sociodemographic factors that play a role that are statistically significant at the multiple time points. It's set as we move farther away from the time point at which individuals have definitive contact with the study team. And individuals move more into their built environment. We see that they have more factors that are related to their lack of return of surveyed materials. And I think this, like what we know in healthcare that 60% of an individual's health is not defined by their clinical interventions, but by their social factors and social risks in their home environments, this suggests the same thing that with respect to research when those home factors begin to take play the likelihood to engage in continued engagement and research study to wings. So it also points out to us that retention barriers maybe different. We saw that across sociodemographic markers, these might be different than recruitment barriers. But we weren't able to really study recruitment barriers and the work that I showed you beyond race and ethnicity. So how do we move forward from this work? We've built a larger model thinking about what are the different things that might play a role in research enrollment. And we've thought about different influencers of the process, drivers that impact those influencers and then different factors that play a role in those drivers. And for time's sake, we're not gonna have time to run through this entire logic model, but you can see that it can get relatively complex with the number of influencers, drivers and factors. And just to take one of those to break down to demonstrate that complexity, when we look at the parent alone, and we break down the different characteristics, practices, experiences and resources that that parent has, that they play a role in their decision making towards research, we can begin to understand really how complex this decision making is and how much there is to really investigate and trying to figure out which of these factors might we be able to modify or intervene upon in order to improve participation by diverse children and research. When we're thinking about improving participation, I think it's important that we don't think about enrollment in a narrow sense. Think we often think about the regulatory elements of enrollment and our need to meet those requirements with obtaining consent, whether that's verbally or through written consent in a scent from study participants, but really our ability to recruit patients and their willingness to participate begins well before that. Recruitment signs will tell us that it begins at the time of formulation of the research question, through to identifying the population in which we want that question to be answered, in how we design and approach to those individuals, subsequently in how we think about consent in a scent, actual enrollment and then in retention. And there may be steps or interventions along the way that can help us to improve both enrollment and retention if we think about this as a spectrum and not a single point in time. So I want to jump back a little bit to the beginning where we talked about roles and principles of justice. And as we begin to think about interventions to improve or break down barriers, it really brings up concepts of equity. Just to briefly go through these ideas, inequality is the idea of individuals having unequal access to opportunities. So we can see here that the tree is bending to the left. So the individual on the left is going to have higher access to getting apples in the individual on the right, regardless of what the individual on the right tries to do when they, you know, without additional tools to help them. Equality refers to the concept of giving equally distributed tools and assistance to everybody. Let's give everybody equal opportunity by giving them the same amount of something, some good. So here we can see that both individuals have the ladders that are of the same height, but still the individual to whom the tree bends has greater access and improved chance of being able to get the apples. Equity really starts to move towards, how do we customize its tools? If we want to reach a point of equity, customize our tools that identify and address inequality. So here, even though this tree may be unimovable and leaned to that left side, we see that the individual on the right gets a higher ladder. So that his ability to access the tree is made better and he has equity with respect to the individual that has the lower ladder but has a tree bending towards his direction. And we'll talk towards the end of the talk about how we start thinking about principles of justice. So keeping in mind that equity is about thinking about tools and customizing those tools for our participants. And then we also need to know, well, what are we customizing them towards? And here's the summary of the reason for lack of diversity mixing both adults and child data. I think some of the prospective work that we're planning will potentially demonstrate some of what's been shown in the adult population and child population. But we know across the adult population, there's poor awareness of opportunities for research. There is mistrust or what I like to call broken trust between individuals and the research system and healthcare system. There's confusion and concern about research, not easily accessible transportation. I think some of the work that we've had to move to in a virtual platform for research has shown us that we can do a lot in that manner and that might actually improve, we're actually looking at some studies to see whether it improves participation by diverse groups. But also study criteria can be a barrier, documents that don't use clear language, fear on the parts of individuals about research interventions or about receiving placebo, worries about health insurance coverage as something goes wrong. And the lack of diverse study staff and lack of concordance between study staff and potential participants and insufficient values. So what is the, you know, the Belmont report, you know, raised the notion of the social value of research and individuals themselves may think about what is that value and is it worth their participation. And so one more concept to think about before I present what I have been thinking about in terms of how we can maybe at least conceptualize addressing these barriers is to keep in mind that how we've kind of moved to this concept of adaptive trial designs. So historically, we've had a big sample trial design where we've formulated a design for a study, we conduct the study and then we analyze those results. But we've moved towards within the study conduct having periods of intermittent review and adapting as we need to then move into a final analysis after the study has been adapted based on what we found. Similarly, I think we can think about adaptive engagement in the research setting. And what do I mean by this? I mean that at each stage in that recruitment, science, spectrum, we can think about, we can review how we've done something that will promote equity, diversity, and inclusion at this stage. And if not, how do we adapt our approach at this stage in order to promote equity, diversity, and inclusion? And I can provide some examples for you at each of these steps to demonstrate what I mean. So at the stage of the research question, we can ask ourselves, is the research question responsive to community and research your needs? And if we don't know the answer to this question or the answer to this question, we do know and answers know, then we need to move towards asking those that are in the community that we wish to study. How can this research also answer questions that are responsive to their needs and their interests? It doesn't mean that we have to change our original study question, but the research questions along the way should be inclusive of questions that are important to the study population. Next, we want to make sure that the study population itself doesn't exclude disparities populations unless it's necessary for the design of the study. And so if they answer the question, does the study design a recruitment plan risk excluding disparities populations? Then we need to go back to re-identify the populations of interests that we'd like to study and how we intend to recruit them. And that recruitment starts again with identifying the population, but also moves into how we approach that population. And we can think about whether the hours, location, primary language of recruiters, and preference for information delivery, if that is varied enough and responsive enough to individuals from disparities populations to facilitate approaching them for research. If not, we again would have to adapt those characteristics of the study. Next in the consent and ascent process, we talked about confusion and lack of clarity of research studies being a problem. There are several measures for health literacy as well as several measures for assessing health literacy, which range in time for administration, but some are very briefly taking a few minutes to administer that can be used during the consent and ascent process in order to ensure that there's adequate understanding before enrolling individuals, so that we ensure that we're enrolling individuals that have a intention to enroll in the study population in the study at that time. And that we're getting robust ascent from children who might not understand as much as adults, but for whom there are also assessments of understanding that we should be applying at the time of consent. And finally, when we're enrolling individuals, we should think about well, what are the specific barriers to enrollment? If that barrier is transportation for some, but for others, its childcare is it okay to give differential compensation? Do we have to have the same incentive plan for everybody in the study where we say, we're going to give everybody a $25 gift card or a $50 gift card and you can use it as you wish. Or if we know that parking for seven hours is gonna cost $40, all we just provide transportation for those in whom we identify transportation as a barrier. And should we provide an adequate childcare rate for childcare for those in whom that is identified as a barrier. But in order to provide sort of these more nuanced incentives that directly address the concerns participants may have, we have to know something about those potential participants as we're attempting to enroll them. So some socioe demographic screens as we're conducting our research is important. I do wanna point out this study because I imagine before I go to the last element because I imagine for some the idea of differential payment maybe a challenging one, but several prominent emphasis have actually very nicely worked through rationales that justify differential reimbursement, which differential reimbursement reimbursement occurs when you're providing money for expenses that were incurred out of pocket expenses. And these could be different for different participants. So it can be justified to be differentially reimbursing them. Compensation refers to time and effort, maybe not direct out of pocket expenses, but we know that studying participants may need to participate or contribute more or less to studies. And so differential compensation maybe you maybe justified because individuals are uniquely contributing valuable ways, even if not spending out of pocket expenses in those ways. And finally, it can be justified even to provide incentives in a differential way. If you can identify that there are subgroups that are particularly important for the scientific validity of the study to be met and you need to reach a certain incentive level in order to get study participants to participate and those individuals, some of those subgroups may need an incentive that's higher than others. We actually have looked at this in a couple of studies, both in rolling in COVID vaccine studies, looking at individuals that might differentially contribute to the research and so proactively enrolling those participants. And in another study, differentially granting incentives to individuals who come from neighborhoods that are low. So not at a participant level, but at a structural level, neighborhoods that are low socioeconomic status. And in working with our IRB, we've been able to justify both this differential enrollment approach as well as a differential incentive approach by outlining these equity and justice issues. And the last thing with respect to retention is that we may need to think about different factors again at the point of retention. So reimbursement for child care if that's an issue for ongoing study. The use of graduated payments and the use of credible messengers. If individuals are back in their community, potentially using messengers within the community in order to conduct ongoing study visits. So those are a number of interventions or factors that we can think about in terms of addressing barriers that are barriers to equity. I wanna talk a little bit about public health emergencies and within that we'll talk about structural barriers and then about how we address justice. So we know that principles of underlying public health emergencies require data-driven responses. So often require that research has already been done or that it's quickly done and that information is accurate, complete and that it's reflective of the population that's impacted by the public health emergency. But we also know the implementation of measures not only concoction of data, but implementation of measures at a population level can disproportionately impact segments of the population. And this occurs with respect to surveillance for particular diseases, prevention, or evaluation of interventions. And then often underlying structural barriers to equity are the reasons why we see differences in data collection as well as implementation. And what we really need is cooperation from those who are disproportionately impacted. So this played out in the COVID pandemic, these maps, we've seen everywhere, but these are maps of the metropolitan Chicago area, these show COVID cases by ZIP code. These are COVID deaths by ZIP code. You can see that the areas of darker and highlighted in green are more concentrating. To the periphery, these are areas in Chicago that are of lower socioeconomic status than in the city center, which is shown in the lower colors. The social vulnerability index are overall, which is an overall marker of neighborhood vulnerability collected by CDC data and maps to census tract. We see also that that collects in the periphery as those being areas of areas of higher social vulnerability. And we know that this impacted vaccine impact, individuals of lower socioeconomic status and of minority race ethnicity live in those areas. They're also lower for vaccine impact uptake. So the majority of those who have taken the vaccine are white followed by this manic Latino population, then the black population, and then the Asian population and then less by populations that either identify as other or of American Indian Native Hawaiian or Middle East Eastern descent. Those of black and Hispanic descent are likely to refuse the vaccine because of concerns about potential, I'm sorry, not less like the interviews, raised concerns about not getting the vaccine because of potential lack of access, though a number of other reasons relate to concerns about the vaccine, even if they did have access, including side effects from the vaccines, safety of the vaccines, being the most common barriers. And we can see, I know this is a busy graph, really what I want you to look at is that as you move down, we move into individuals that are of minority race ethnicity in these lower bars. Also those who are uninsured rural residents, those of Republican political affiliation but younger individuals also, and those without a college degree, that we seem less of the dark blue and more of all of these other colors, which are either those who are waiting and seeing to take the vaccine will only take it of required or definitely not. And this is important because this also reflects participation in trials across both vaccines and therapeutics for COVID-19. So here I'm showing you the COVID case rate and the COVID death rate by race and ethnicity. And then I'm showing you what is the participation rate if we haven't available to us by these race and ethnicities in the vaccine trials as well as in the clinical trials. So we can see that for Native American Pacific Island, there's American Indian and Alaska natives. There's virtually no participation in either vaccine or trials for therapeutics. There was more participation across Hispanics and black individuals, but when we look at the rates of, sorry about that, the rate of the incidence of COVID as well as the death rate, their rates of participation don't really match across all of all of these studies, particularly for the Black, African American population, whereas the white population is overrepresented, relative to the case rate and the death rate. Again, the Asian population is underrepresented. Some interventions to directly address participation by minority populations include that idea of the credible messenger that I was raising. So bringing into settings where we're either distributing vaccine or trying to reach individuals for trial and enrollment that we include individuals that look like the population that we're trying to recruit, let's speak the language of that population and understand things about their neighborhood and build environment and how they can participate. And that begins to get at this principle of justice and justice is really, how do we begin to fix the system to offer equal access to both tools and opportunities? So how do we remove the leaning of the tree or the structural barrier rather than trying to create tools? Well, we need to know what are those key structural barriers and we know that trust is a big one of them. We need to have approaches that engender trust and embody trustworthiness. So we need a robust informed consent. We need to maximize transparency, demonstrating for public health in particular, where is the social value and what are the principles of solidarity underlying this research and the public health interventions that may follow. We do ensure post trial access to trial interventions, reassure about coverage from medical care of injured as a result of participation. And then we need to think about the idea that trust is many things. It's trust with the do share relationship that may form between a doctor who might also participate in referring to research studies and has a few serious relationships with their patients. Trust as confidence in the confidence of research professionals not to exploit participants and then perceptions of trustworthiness of institutions. How do we reflect that our institutions in general or that the research enterprise in general is a trustworthy one. And I think in order to do this, we need to understand individual social determinants and influencers for which race we've collected historically. But this is often acted as a proxy. And we need to have a more nuanced understanding of who our participants are so that we can begin to think about what are going to be particular barriers for this participation and how can we address those in a systematic way to really promote a diverse and equal and equitable research environment that maximizes justice. And going to stop there at a recent time for questions, I want to say all of the individuals that make up our labs and have worked on these projects as well as numerous other phenomenal projects that they've designed addressing social justice issues. And then leave you again with a quote. I think bringing us from those who have really inspired the country really from Martin Luther King 1960s to Amanda Gorman in 2021 when she stated we're striving to form a union with purpose to compose a country committed to all cultures, characters, colors, and conditions of man. And so we lift our gaze as not to what stands between us but to what stands before us. And I think as research professionals and not of what professionals would stand before us is really opportunity to create more diversity, equity, and inclusion in research. And I hope that this has given you some tools to begin that work. I took knowledge. I'll leave this up there. Everyone, again, that's participated in my funders. And I'm happy to take questions at this point. Aaron is Jeff Burns. First, as a graduate of our fellowship program, it's been amazing but not surprising to see how you have taken your background in law and ethics and your formal training in both and fused into such an important research portfolio. And I think that's a great way to make sure that you're able to make sure that you're infused into such an important research portfolio. As you know, I've sit on more than a few committees where I've heard your work presented and it's easy to say that you are really one of the leading researchers in this area in our field in the United States. I have one question for you. Is there any data on whether all of these issues are exacerbated in parents who are facing a new diagnosis as opposed to parents who are facing a known diagnosis? And if not, what is your gut tell you would a new diagnosis exacerbate likely many of these issues are not? I think that's a real, first of all, thank you, Jeff. That your comments are really particularly meaningful coming from you and all the mentorship over the years that the program and then you have provided in different formats for my work. So I want to thank you for that. To answer your question, I'm not aware of comprehensive studies that demonstrate with respect to research enrollment that the Q versus chronic diagnosis. Though I may not be aware of all of the cancer literature, I'm not particularly aware. But I can tell you that we are actually studying that in one of our prospective trials where we're looking across socioe, we're attempting to enroll individuals from lower socioeconomic status, but we're collecting data for all participants who enroll in a prospective study for a PICS intervention or post-intensive care syndrome intervention study. And we're particularly looking at whether children with acute diagnoses versus parents of children with chronic diagnoses differentially enroll in that trial. My suspicion is that they do. And again, there may be some cancer literature. If anybody in the group is aware of, please do speak up. That demonstrates that, but it is a question. I think that's important and that we're actively looking at. I'm not sure if I'm supposed to call on people, but the next hand AC is Bob Trug. Hi, Erin, this is Bob Trug. Again, congratulations on such wonderful work and a great presentation. Question I had was, my understanding is that as the Moderna trial was nearing completion, they did not have adequate representation of minorities and the FDA told them that they would not get the EUA unless that was improved and they went out and they specifically targeted certain communities and ended up actually with a fairly representative study population. Do you think that if journals refused to publish papers where the study population did not adequately reflect the treatment population or where the FDA made this a requirement, do you think that that would be an effective way of sort of really forcing better diversity and recruitment? Yeah, I actually do and I won't flip back to the slide just for time. I didn't go through this, but you could see in the Moderna trial, if you've looked at the Hispanic and Black population participation, they were higher than in the Pfizer trial and the particularly in the Hispanic population, they actually reached if not exceeded the incidence rate of COVID for that population. And I do think that unfortunately we need both a carrot or incentive approach as well as a stick approach to having sort of compliance with, I think both publication standards, and we know that there are publication standards now. Again, for times sake, I didn't include these, but there are publication standards for the use of race and the word race in a study, which I think has dramatically improved how we think about race and racism as being factors that are relevant to minority populations. But I think you're exactly right that if we begin to have regulations at the federal level that really say, hey, your study can't continue if you're not meeting these targets, and then if they're not able to be published if they don't meet the targets, I think that that would help. I don't think it's gonna remove all barriers, and I'm not sure that it will get us to truly proportionate representation, but I think it'll bring us to steps in the right direction. I don't see a name on the screen of the next hand that's up, so I don't know if you just wanna speak up or if anybody recognizes, I apologize. I don't know the next individual. It might be me, this is Steve. This is Steve, I'm sorry if I'm not showing. I wanna thank you for a great talk. Also, as a fellow Wildcat, I noticed in looking at your background that you and I have trained and educated all the same three institutions. I don't have as many degrees as you, but I'm very proud. I have great affinity for both Boston Children's and Laurie. And so I thank you for persisting in advancing a field that has now become kind of a vote for all of us to focus on. And my question is kind of a challenge for us. You may be aware that Boston Children's has recently made a major commitment to create an institute for pediatric health equity. Inclusion. And it's specifically, and there was a lot of discussion about this, we're committed $25 to $50 million to this initiative. And as we were discussing how to do this, I was quite vocal in stating that we need to make sure this is not just an expensive bumper sticker. And then we actually find a way to make a difference. And suggested that we're not gonna be able to boil the ocean and solve all of the inequities in healthcare. And I said if there's anything Boston Children should focus on, it's specifically on pediatric components of equity and inclusion. And so we're doing that. My question for you is, how should we, as a pediatric institution, focus those resources to actually make a difference for children? I think there are a number of ways that you could go about answering that question. And I've been actively participating in conversations within our own institution about how to answer that question. Largely our institution in developing its equity be plan has been driven by two forces. One is that community health needs assessment and the factors that were actually identified in our community or outward facing efforts outside of the Mecca Children's Hospital in order to develop a plan that we're implementing over, we're in lockstep with the CHNA to implement that plan over five years. So we've used that community health needs assessment which all hospitals conduct in order to help guide our thoughts about particularly around equity. Like I said in outward facing, from an inward facing perspective, we have formulated a President's Council for Equity, Diversity and Inclusion that is led by our Chief Equity, Diversity and Inclusion Officer but is comprised of individuals working across the spectrum of EDI and we have chosen to outline a five year plan based on some data that we've identified in terms of. We started looking at our outcomes and we have started to create a dashboard of outcomes by race and ethnicity and focusing our early efforts really in a mechanism that's data driven. So what are we learning from our patient family surveys? Where is there a signal that race and ethnicity is coming out in a disproportionate way that's problematic? For us, we've learned that that's among our limited English proficiency group and so we know that part of our initial plan needs to be directed at improving efforts in our limited English proficiency populations and then we're looking at outcomes to see are there particular clinical areas that we need to have make areas of focus? And so it may not be the same type of approach that is implementable of Boston children's, but I would say, I think what's been the most successful is that it's data driven. We've started with where does the data take us and then we've developed a plan around that. Thanks so much, it might be valuable. But if they're not already for our leadership to lock arms and investigate in that. Absolutely. I like the success probably. I would be happy to facilitate that. I knew we're closer maybe just that time but I did see Adrian Randolph's hand earlier. I don't know if you still have a question or we addressed it. I guess I just had a quick question about that the IRBs have come back at me. When does a incentive become coercive? We are giving out these iPads so that the missed fee patients in neurologic follow-up can actually do the interviews and assessments because they don't have access to a computer. And so we gave them iPads and they came back to us say that's coercive, that it actually did those through finally, but I think that that's one of the things that comes back when you try to change the incentives to make it so that the patient has the ability to participate. Yeah, and I thank you for raising that because I think that's an important point. And in an upcoming study that we are doing, we will also be distributing tablets for individuals to participate in the intervention. And we have similarly gone back and forth with IRB. What has mostly been successful is exactly what you just said, which is that if you can demonstrate to the IRB that you're going to really lose the ability of an entire demographic population to participate because they don't have access to something that's necessary for the study or necessary to facilitate their participation in the study, then I think if you can raise that as a concern for being able to proportionately represent the population of interest that you have, that has been where we've had the most success with our IRB is that we've tried to upfront present that data and if we haven't upfront done a good job coming back to them with exactly what you said, which is it's about ability to participate, it's not about trying to make them participate or correspond to participate. And if you can frame it in that way, hopefully you'll have more success, but there is variability. I actually think this is another area to study, how much IRBs are willing to accept this type of incentive to facilitate equitable participation. Aaron, thank you for an absolutely wonderful talk. Thank you, Jeff. And everybody, I appreciate your attendance and your engaging questions.