Big DataAhead of Print Roundtable DiscussionFree AccessImportance of Community Engagement in Data Decision MakingModerator: Michael Crawford, Participants: Francisca Flores, Marynia Kolak, Amy Hawn Nelson, and Malaika SimmonsModerator: Michael CrawfordAssociate Dean for Strategy, Outreach, and Innovation, Howard University College of Medicine, Washington, District of Columbia, USA.Search for more papers by this author, Participants: Francisca FloresProgram Officer on the Gulf Research Program, The National Academies of Sciences, Engineering, and Medicine, Washington, District of Columbia, USA.Search for more papers by this author, Marynia KolakHealth Geographer, GI Scientist, Principal Investigator, The U.S. COVID Atlas, University of Illinois at Urbana-Champaign, Washington, District of Columbia, USA.Search for more papers by this author, Amy Hawn NelsonResearch Faculty and the Director of Training and Technical Assistance, Actionable Intelligence for Social Policy (AISP), Philadelphia, Pennsylvania, USA.Search for more papers by this author, and Malaika SimmonsChief Operating Officer, National Alliance against Disparities in Patient Health (NADPH), Woodbridge, Virginia, USA.Search for more papers by this authorPublished Online:10 Apr 2023https://doi.org/10.1089/big.2023.29057.rtdAboutSectionsPDF/EPUB Permissions & CitationsDownload CitationsTrack CitationsAdd to favorites Back To Publication ShareShare onFacebookTwitterLinked InRedditEmail Mr. Crawford: I would like to welcome everyone to the Big Data roundtable. My name is Michael Crawford. I am the associate dean for Strategy, Outreach, and Innovation at Howard University College of Medicine, founder and executive director of Howard University's 1867 Health Innovations Project. I am also a commissioner on Robert Wood Johnson's Commission to Transform Public Health Data Systems. It is widely believed that community voice is vital to developing 21st century public health systems, where data are used to understand the unmet needs of residents, especially those who have been historically excluded from the data gathering process.This roundtable discussion will feature panelists from organizations and initiatives designed to expand local data collection capacity so that it is beyond community informed, and it is community led. Expert panelists will highlight their efforts to build data capacity in underserved communities and enhance community involvement in shaping equitable data systems.The discussion will ultimately reflect the importance of authentic community engagement in data decision making and showcase how data can help communities target community relevant interventions and measure progress, which is necessary to pave the way for a healthy and more equitable future for all.Today I am joined by some amazing thought leaders and panelists. I am joined by Francisca Flores at the Gulf Research Program at the National Academies of Science, Engineering, and Medicine, Marynia Kolak at U.S. COVID Atlas, Amy Hawn Nelson at the University of Pennsylvania's Actionable Intelligence for Social Policy (AISP), and Malaika Simmons at the National Alliance Against Disparities in Patient Health (NADPH). I would like each panelist to briefly introduce themselves, and then we will dive into the questions.Ms. Simmons: Thank you so much for having me. My name is Malaika Simmons. I am the chief operating officer of the NADPH. I have a background in clinical research, public health, and drug development before moving on to policy and regulation in the federal health space.I was a policy development officer for the Division of AIDS working internationally to create equitable policies and practices related to the research of HIV and its effect on under-resourced populations. I have worked in pharma, private, and public sectors, as well as academia and nonprofit spheres. The constant thread of my work has been the empowerment of people, to have agency and autonomy when doing the hard work of just being a human in this country or in the world, and especially for women.I am especially intrigued and committed to the journey of world-changing women internationally, but it started here at home, observing stark disparities in women's health care at the intersection of community health and public policy. I am also a certified design thinking practitioner, and I seek to include human-centered design in all aspects of what we do at NADPH.Dr. Flores: Hello. My name is Francisca Flores. I am a program officer in the health and resilience unit of the Gulf Research Program at the National Academies of Sciences, Engineering, and Medicine. My PhD is in public health, specifically community health sciences with training in community-based participatory research, which is all about harnessing community power to achieve better health outcomes.Dr. Kolak: I am Marynia Kolak. I direct the Healthy Regions and Policies Lab, where I lead the U.S. COVID Atlas. Both are based out of the Department of Geography and GIScience at the University of Illinois Urbana-Champaign. I am a health geographer. I use open science tools and geospatial data science to better understand neighborhood health and the social determinants of health. I also work with government groups, community groups, and other researchers to develop applications and decision support systems to access those same data and make it useful for different purposes.Dr. Hawn Nelson: I am Amy Hawn Nelson. I am research faculty at the University of Pennsylvania and Director of Training and Technical Assistance with AISP. We are a small initiative that supports government agencies to collaborate and use data more ethically and responsibly. What that looks like practically is we support a lot of data initiatives across the United States, mostly through our network (https://aisp.upenn.edu/about-aisp-network/) and our Equity in Practice Learning Community (https://aisp.upenn.edu/eiplc/). We describe our work as being singularly focused on ethical use of data, and we promote strong collaborative governance as the foundation of all data sharing and integration. This means we support the development of data governance, legal frameworks, and centering racial equity across data practices.Mr. Crawford: OK. Can you give us a brief overview about how your organization and its work to build data capacity in under-resourced communities enhance community involvement in data collection?Ms. Simmons: NADPH's work has largely centered on the state of inclusion, importance, and tangible value of engaging communities as persons with lived experience in all aspects of the data lifecycle within community data ecosystems. So, we look at it as an eight-part process—plan, collect, access, analyze, interpret, recommend, report, and disseminate, all the parts and pieces of the ecosystem for which people with lived experience should be involved along the continuum. We are a data-driven organization led by people with diverse lived experiences speaking to people with diverse lived experiences.We do not talk at our communities. We facilitate cocreation of platforms for engagement. We provide information to and lift up the voices of community experts on factors impacting their quality of health, health care across all the social determinants of health, which essentially is localized decision support.Dr. Flores: The mission of the Gulf Research Program (GRP) is to “develop, translate, and apply science, engineering, and medical, and public health knowledge to enhance offshore energy safety, environmental protection, and stewardship, and human health and community resilience in the Gulf region in ways that empowers its citizens.” That last piece is critical, because how you do things matters. The process will determine the outcome. In carrying out our mission, the GRP uses data to execute its four strategic lines of action:Advancing health and understanding, building partnerships and engaging networks, bridging knowledge to action, and monitoring for progress and change. Specifically, the health and resilience unit is building data capacity in frontline communities of the U.S. Gulf Coast. These are the communities that have been the first to experience the effects of climate change, and they have been experiencing those effects at a disproportionate rate and severity compared with everyone else.We do this by transforming the predominant paradigm of what Malaika called the data lifecycle. By using community-based participatory research (CBPR), our work is flipping the script on who decides what data are collected for whom, when, where, how, who owns the data, and how the data will be used to inform decision making that will solve real-world problems for the benefit, of the communities themselves.Dr. Kolak: While we are technically a research laboratory, part of the work that we do as researchers is focused on understanding data capacity, but then also moving it toward the next generation of what is possible. Our research includes a number of projects that seek to redevelop and actually flip the technical systems of different data infrastructures, as well as data decision support structures. So, that involves dashboards and other things you have probably gotten really familiar and sick of from the pandemic.One of the things that we specialize in is harmonizing data from dozens of different formats, from satellite data to census records and way beyond. And we make that available to communities, as well as other research groups and various civic organizations at spatial levels that make sense to them (i.e., neighborhood level aggregations). We make those data not just available to look at through maps, but also enable users to interact with, filter, and then, ultimately, download the data.We also work with communities in different organizations to identify data needs and goals for what they are trying to get out of those data. We do a lot of work in building toolkits and tutorials to enable others to learn how to work with the analytic and data wrangling data science pieces needed to work with those data.There is one project funded by NASA where we are working with several community organizations with the harmonized data that I talked about. Their goal was to develop a new filtered index to guide greening in their area. Through the U.S. COVID Atlas, we have added an oral histories component where people can add stories. That becomes embedded within the wider dashboard system to humanize the pandemic and the data associated with it a bit more. Through work with community groups in the past, we have learned to ensure that folks are able to retain ownership to those data and ownership to their stories—from a legal perspective, as well.Dr. Hawn Nelson: AISP has supported the work of government agencies across the United States in centering racial equity throughout data integration. We started this work in 2020 with the publication of A Toolkit for Centering Racial Equity Throughout Data Integration (https://www.aisp.upenn.edu/centering-equity/). This work emerged in 2017 at the request of members of our network who were concerned about the disparate impact of data access and use on their communities, and sites were looking for resources. We could not find a lot of helpful information, so we convened a workgroup and began a participatory process of learning alongside each other to think through how we can best center racial equity in the everyday work of using government-held data.This was pre-COVID and in person, so we ate a lot of muffins, drank a lot of coffee, wrote on lots of chart paper, and figured it out through hours of discussion. Our cocreated Toolkit was published in 2020, and provides applied and concrete ways to center racial equity throughout the data life cycle. We include many examples of work in action as well as resources to support conversations to shift practice. We have been working hard to operationalize these ideas and to support sites committed to this work. Our Equity in Practice Learning Community began in 2022, and this is a place for sites across the United States to learn alongside each other as we work to center racial equity across the data life cycle. We are a year into this work and it is messy and challenging, and worth the effort.Mr. Crawford: OK. As we just talked about data in health equity, what does data equity look like to you?Dr. Flores: In short, to me, I think it is that everyone has a fair opportunity to be counted by data. So, if our data systems are designed in a way that you do not get counted, that system is not going to work for you. It is not going to be able to intentionally benefit you. In all aspects of the data lifecycle, everyone should have a fair opportunity to be counted.Dr. Kolak: In addition to being included within the data, I also see it on the other side of different groups being able to access the data, especially if it is about them and their communities, but also making meaningful insights from those data and being able to use those data and being part of the process of using those data for action and policy and all the things that come from that. I see data equity as a fight. It is something that we must persistently work toward. It is also slow, where when you move from a top-down approach to a more inclusive approach, more people are involved, so we also need better infrastructures to make that process participatory and led by communities throughout. So, I may be the pessimist in the group, but that is how I see data equity.Ms. Simmons will say that the inclusion of diverse data is what data equity looks like. Those who are impacted should be those that are represented. I am going to cite Francisca and say these communities typically get first impact and worst outcome. We know that this is what happens. So those impacted by the treatments, policies, and systems that are being developed, their data should be there. Then, fast forward, those systems, policies, and health care needs to be designed with those diverse data in mind. It is not just throwing it in there. How are the systems and policies and treatment using those data to make better decisions?And lastly, this data ecosystem works in ways that are congruent with and beneficial to the multilayered approaches to people with lived experience until such time as there is no under-resourced community. We do not think that is ever going to happen. So, that means that we are just going to continue. As Marynia said, we are going to continue to fight. All of this is to inform ongoing ethical approaches to access quality individual treatment and policy. I think data equity boils down to ethical approaches to all of that.Dr. Hawn Nelson: We see data equity as similar to a common definition of racial equity. We see it as both a process and an outcome. We see data equity as a series of decisions and choices, both large and small, that when put together, they make something new. We like to think about positive and problematic practices across the data lifecycle. And the focus is doing a lot of things a little differently. All of that adds up to equity. It is not one thing. It is many, many things. We tend to think about the data lifecycle as planning, data collection, data access, analysis, use of statistical tools, like algorithms, and then reporting and dissemination.We focus a lot on that earlier part of the data lifecycle for thinking about equity. It is critical to go upstream, planning involves staffing, and making sure we are resourced adequately. We often focus on the beginning parts of that data lifecycle, rather than where our national attention tends to go, which is the use of algorithms and how things are reported.Mr. Crawford: I am going to ask a follow-up question. How important are standards to this process?Dr. Hawn Nelson: We see standards as being very important, but the challenge is that standards are different in different places. So, my answer on pretty much everything is going to be data governance. It is the people talking about the policies, practices, and procedures of how data are accessed and used. We think that standards are really important, but these standards need to be collectively created and agreed upon by people who are impacted. That includes people who are in the data, but it also includes people within agencies and people who are creating and using the data. It needs to involve all parties, rather than a small group who are creating these on high, and then delivering them down through a policy. We just do not see that as an effective practice for data equity.Dr. Kolak: So, I think for standards, it is tricky because it depends on which standards for which part of the infrastructure, and also depending on the problems that are being tackled. The more multidisciplinary, the more complex the problems, the more voices you are going to be bringing from many different perspectives. And each of those groups may have their own standards. There are challenges from that perspective, and I almost prefer the idea of standardized approaches and including communities both early on and throughout the process.A standardized approach may be more effective than enforcing standardized data throughout the process in different places, because, in my experience, it is not the data by itself that pose the biggest challenge. It is not the technical pieces. It is the people part that makes things more difficult and challenging. For example, satellite data have really complex standards that exist in the scientific community. But when you are trying to communicate that as “greenness” at a neighborhood level, when you are working with community groups worried about climate change impacting their neighborhoods and potential translations to health, the question of data standards gets a bit more complicated. Data standards are not as important as how you translate those data to meaningful insights that impact the communities most affected by climate change, in that case.Ms. Simmons: So, I am going to take a little bit of a different approach completely out of the realm of health, and data. I used to be a Subway owner; Subway has a big book on how you are supposed to run your Subway store. But we have Subway stores all over the world, with many diverse food items. What is an appropriate menu for one store may not be for another store. So, if I am in California and I want a Baja Fresh, I can have that. But if I am from New Jersey, and I want something different, I can have that, too, right? And so that is really the point of standardization. It is really for simplification. And it is not so that you must do the same thing everywhere every time. And so, I think of standards as more of a consideration of three things that you really need.You need access, use, and implementation of whatever the thing is. Ways in which one community might access, use, and implement, might be different than the way another community might access, use, and implement. But what could be the same is the mechanism for how you get to it, and the wherewithal that the community has to understand that they have the agency to go and grab the thing that they need the most. It is so that it is easy for you to access, to use, and to implement whatever it is that you need to govern. I think of it a little bit more as sandwich making than it is to applying the data. But I think it can be transferred.Dr. Flores: From the perspective of CBPR, we have guiding principles—akin to standards—that come from decades of best practices and lessons learned from addressing health inequities with communities. What I like to do in the beginning is review them with the community and have them reinterpret them in a way that makes sense and is meaningful to them. This gives the community a sense of ownership over the process that will guide our work.Mr. Crawford: What counsel would you provide other organizations in incorporating community voice and key decisions about data, specifically when it comes to health equity and racial justice?Ms. Simmons: So, it should be no surprise to you that I am going to say honesty—speaking truth. Tell communities the truth—no condescension, no sugarcoating, and no gaslighting. I think that is prominently number one. Many communities, for various reasons, do not trust government. They do not trust academics. They do not trust anyone outside of their living room. When you are trying to provide services and data to people, you need to also understand the lens in which they are receiving those data. So that is the crux of communication. It is going out, and it is the feedback, and it is the loop, and how is it coming back to you.So being very cognizant that your messaging lands with different people differently, and there is no pathology there. There is nothing wrong with people if they do not get the message the way you wanted them to get it. There may be something that you need to change in your messaging. Or you may be offering something that just does not align with that community. And so having that honesty and that honest look at what it is that you are doing as an organization, whether or not it is landing with your communities and why if it is not, and just being really honest with yourselves as an organization about what services, and offerings, and data you are collecting.Another one would be stop policing community approaches. Stop telling communities that they do not know what they are doing. I have talked to community members in their 80s. And they were nurses back when they wore the crazy nurse hats. Who am I to come and tell this person that that is not how you talk to someone that is just coming into a hospital?So, listen to them. Understand that they have a whole set of skills and expertise that you may not have. And it is important to the process. And lastly, you may have heard that this work is about moving at the speed of trust. What does that mean? Does that mean that you say that, but your policies and your expectations for communities do not actually line up. You cannot say, “Oh, we're moving at the speed of trust, but we need that report by noon tomorrow.” Oh, OK. I only need to go knock on people's doors. I cannot just get on a Zoom. I need to physically go get this information from communities, and you are telling me that I need it tomorrow. So, just being cognizant that you cannot say one thing and do another thing.And then listen. Listen some more. And then when you think you are tired, keep listening.Dr. Flores: I would encourage organizations to incorporate the community's vote, not just the community's voice. Organizations need to move beyond that consulting level of community engagement: “We are the experts. We hear you. We gave you a voice. And we'll take your feedback into consideration as we, ourselves, make these decisions in a vacuum somewhere else.” You must move beyond that into shared decision-making power by giving the community a vote and not just a voice. To do this genuinely, I think the organization needs to be honest with itself. It must value the community as an expert who can contribute local and traditional knowledge from their lived experiences.Once we have that value, once we hold that to be true, then our attitudes become more favorable toward involving and collaborating with the community as equal partners with equal decision-making power: “I, the researcher, will contribute scientific knowledge, you, the community leader, will contribute local traditional knowledge, and our combined knowledge together will produce a more comprehensive, equitable picture.” In this way we will have a collective and thus better understanding of the choices that we have to make. And, by extension, we can make more informed decisions.Dr. Hawn Nelson: The work is slow work, and that has to be built into the planning of it. To get to Malaika's point, we have to be really careful about pacing and timing. And if you have firm fast deadlines, then you probably need to re-evaluate your scope of work. This work is glacial if you are doing it right. And that is hard. That is not the way most of our government agencies work. And it is certainly not the way the private sector works. And so making sure that that is incorporated into any project planning and any kind of outcomes is really important. So, to be cliché, go slow to build that trust, and then you can go faster. But you must focus on building trust first.Dr. Kolak: So first, I just wanted to echo everything that has been said by the panelists. I could not agree more on bringing in community voices and perspectives first and not last at the inception stage versus the review stage. There are often well-intentioned goals to do things for communities or groups. We want to help. We want to do things for communities, and shifting that mindset to doing things with communities takes work. And then, ultimately, really democratizing the tools and the learning that comes from working with the tools to being done by communities is another important goal.So, I think that progressing toward that is important. A while back I did some work with Engineers Without Borders. They would have a 5-year commitment to any community that they were working with on a project. A big component was also training so that the project would be taken over by the community by the end of it. So that experience and long-term commitment stick in my mind.But then, also, many of the projects our laboratory does with communities recently were related to the pandemic or environmental justice issues. And a big part is recognizing the tremendous underinvestment and trauma that has occurred in many communities. From that perspective, you better bring money and resources when engaging with communities… Community groups may have different goals and policy action items that they are working on. And especially in the past few years, every community organization we connected with is overstressed and limited. There has been a lot of trauma happening across the world, but especially in some of these communities in the United States. So, recognizing that, and especially if the goals are not completely aligned, trying to be respectful and understanding.Mr. Crawford: What are the challenges of having a community-led process when it comes to data collection and gathering?Dr. Flores: So, I can imagine that some researchers might be concerned about the quality of the data collected—that there is a high fidelity to a systematic collection of data, that every variable that is collected in the exact same way with the most rigor. That said, this is the same concern that researchers have with anyone who is new to data collection, such as students. The remedy to having anyone—a student or a community member—collect data with that high fidelity is to train them. Set them up for success.Give them the knowledge and the skills that they need. Support them in the field. Pay them for their work. Offer to include them in the following steps of the process. I cannot tell you how rewarding it is when you are able to show someone what they helped the partnership accomplish: “This is what you helped us do, this is how you helped your community and the science community, and we could not have done this without you.Dr. Hawn Nelson: I am going to come at this from a slightly different vantage point. Our work is supporting government agencies. So, our work is focused on the use of administrative data, data collected by government to administer programs. Our data collection challenges have more to do with building understanding and documenting how these data are collected and helping analysts and researchers use these data appropriately to support residents. One obvious challenge is that there is a massive amount of data that exist in our world, especially in government. The challenge for access and use is often data documentation, knowing who collected the data, when, and for what purpose. And improved standards could really support this work.Dr. Kolak: So, this theme had been highlighted before, but just the need for patience. This is a slow process. And it is also very dynamic and can change a lot. With the Atlas Stories Project, we spent a lot of time working on building an infrastructure to collect stories by folks across the United States in as fair a way as possible. Anyone who submits a story about his or her pandemic experience retains rights to it. But, it is taking some time to gather those stories, in part because we worked with community groups to ensure the process was in line with their goals. It has taken more time than expected. So, things are dynamic. They might change.Getting buy-in from community leads is crucial. And many of the questions that we have already talked about start to come up, like, why should I participate in this project? Why is it important to our community versus a governmental or research group? The trust issues will really come back up as well, because even if there are good intentions, if there is not that history and buy-in that precedes the project, it just takes much more time to move forward.Ms. Simmons: Managing expectations, not just in timing, but also in process. And we talked a little bit about that before. And that goes to Marynia's point about trust—understanding and having that expectation of what is going to happen, and when, and what is the process, and what is best. Francisca talked about it a little bit a few minutes ago as well. For example, what if you are working with communities who do not do email and you have to do surveys? Then you are going to have to text them. You are going to have to go back and forth. We have had this come up in the work that we do, where we have a whole set of interviewees that we can talk to through Zoom, or through Teams, or what have you.And then we have a bunch of other folks that the only way to communicate with