Democratizing Research With Data Dashboards: Data Visualization and Support to Promote Community Partner Engagement

Historically marginalized populations bear a disparate burden of preventable diseases.^1,2 Health outcomes result from the interaction of multiple domains over an individual’s life course (i.e., sociocultural environment, health care and government systems, built environment) and at multiple levels of influence (e.g., individual, interpersonal, community). Where we are born, live, work, and play are directly related to our health, risk, safety, prosperity, and life expectancy. Differences in these drivers of health lead to unequal access to community-based programs, directly contributing to health disparities.

The Rapid Acceleration of Diagnostics-Underserved Populations (RADx-UP) program is a consortium of research projects throughout the United States funded by the National Institutes of Health. RADx-UP seeks to measure and understand factors that have led to the disproportionate burden of the COVID-19 pandemic on historically marginalized and vulnerable populations so that interventions can be designed and implemented to reduce health disparities. As part of this program, RADx-UP built a data dashboard to support research dissemination. Generally, public health dashboards are data display platforms used for health surveillance and data reporting. As noted by Dasgupta and Kapadia,³ data dashboard priorities and dissemination products are typically decided by individuals and organizations that are external to the communities that the data represent. However, the wisdom and expertise of community organizations, community members, and stakeholders are needed to fully address health inequalities and improve population health. Community partners bring deep knowledge of the lived experiences, values, and historical legacies of their communities.

Building alliances between community partners and academic researchers can enable local communities to use and adapt data, knowledge, tools, and expertise in the design, implementation, evaluation, and dissemination of public health interventions. Furthermore, this collaboration fosters healthy communities that can close health equity gaps caused by education level, immigration status, language, income, place, race/ethnicity, gender identity, or sexual orientation inequalities. Incorporating community partners’ lived experience through community–academic partnership is necessary to address health disparities.

Data dashboards serve as vital visualization tools supporting community partner engagement in research and dissemination. Dashboards enable community partners to explore results, postulate new questions, and disseminate findings based on data visualization capabilities. Providing community members access to intuitive, well-described, accurate, and informative data visualization contributes to true bidirectional communication. Community partners must be central participants in research dissemination, given longstanding barriers to information access; lack of access contributes to mistrust and the spread of misinformation and impedes cultural competence.⁴ We describe the components of the RADx-UP Data Dashboard, the infrastructures established to support community partner dashboard use, and challenges and recommended steps for enabling data dashboards to bridge the information gap between researchers and communities.

DATA DASHBOARD

The RADx-UP Data Dashboard presents a set of common data elements (CDEs): a standard set of study questions spanning multiple categories, such as sociodemographic, COVID-19 testing, and health status (Figure 1). CDEs are collected from RADx-UP participants by all RADx-UP project teams, shared with the data coordinating center, ingested, harmonized, and finally visualized on the RADx-UP Data Dashboard. Linked with the dashboard, area-level data visualizations incorporate zip code and county-level RADx-UP project data with 2019 American Community Survey public data sets, including the Child Opportunity Index, the Centers for Disease Control and Prevention Social Vulnerability Index, and the Area Deprivation Index. The visual output was designed for community partners to dynamically engage the data, form their own questions, and help answer those questions.

FIGURE 1—

RADx-UP Data Dashboard (a) Interactive Panels and (b) Data Filters

Note. RADx-UP = Rapid Acceleration of Diagnostics-Underserved Populations.

Facilitating community partner engagement with RADx-UP data was a key driver in the design of the RADx-UP Data Dashboard, and obtaining feedback early from community partners was crucial for incorporating the following qualities:

• 1.Easy navigation: A home page button and navigation bar on all pages and an actionable table of contents offer many ways to navigate;
• 2.Easy to learn: A consistent and uniform layout affords quick familiarization, so community partners can focus on interacting with the data rather than how to use the dashboard. Also, the RADx-UP Data Dashboard presents a teaching page at the beginning;
• 3.Simple and consistent design: Familiar graph types, such as column, line, and donut charts, and consistent colors assist interpretation;
• 4.Data visual interactivity: Data visuals that can interact across multiple panels on a given page by clicking on the visuals, meaning more complex questions can be asked;
• 5.Data visual filtering: Data visuals are filterable via multiple filters, allowing deeper drilling down of the data;
• 6.Data visual footnotes: Footnotes for visuals provide more detailed context to help community partners better understand the data;
• 7.CDE category grouping by pages: CDE category grouping reduces the need for shifting attention back and forth among categories;
• 8.CDE question grouping using multiple panels per page: Multiple panels on a given page produce more interactivity;
• 9.Data value denominators: Either per page or per panel, visuals display value denominators to provide full context for each CDE; and
• 10.Data comparing with public data sets: Integrating RADx-UP data with public data sets provides greater data context and deeper meaning of the RADx-UP data.

Community partners’ use of the RADx-UP Data Dashboard for research and dissemination was also integrated into the RADx-UP community partner engagement infrastructure (Figure A, available as a supplement to the online version of this article at http://www.ajph.org). Specifically, RADx-UP not only explicitly encourages community partners to lead articles of their own but also promotes their engagement in research and dissemination in the following ways:

• 1.RADx-UP staff support online collaboration sessions with authors to discuss ideas for research study concepts, analysis proposals, and writing;
• 2.Data science team members can assist authors in completing an analysis proposal (including assessing data availability) and statistical analysis plan, and can provide data sets via a secure platform and analysis support as needed;
• 3.Writing assistance at all levels is provided through writing workshops, data workshops, office hours, editorial support, and writing tools;
• 4.Authors’ proposed research study concepts are shared via newsletters and online to promote collaboration across the consortium;
• 5.Transparent and early discussions of authorship are facilitated to maintain healthy and productive working relationships in research;
• 6.Connections are facilitated between authors and community members with relevant lived experience and expertise to serve as coauthors for research questions specific to a minoritized community; and
• 7.Collaboration platform options (e.g., Box, Inc.’s Box; Microsoft’s SharePoint) are offered to facilitate efficient group work.

UNANSWERED QUESTIONS AND NEXT STEPS

We recently developed the discussed initiatives and are still evaluating them for effectiveness. This iterative process is key to ensuring that the dashboard is useful and usable by researchers, operational staff, and community partners. Despite efforts to build a data dashboard and support infrastructure that will promote community partner engagement, barriers to authentic engagement must be overcome.

To begin, inviting community partners to this effort is insufficient to successfully guide, inspire, and empower them to use the tool to better serve their community. Democratizing the research and dissemination process requires that we redefine who can initiate and answer research questions and who determines the validity and value of research questions to address. That is, we need to provide data and information to community partners as well as let them discuss what those data mean, what their implications are, and what steps to take next. Moreover, a culture shift is required from the start of a partnership to ensure that the community has equitable decision-making authority throughout the project. This shift needs to address historical issues of distrust that remain in communities of color for research that may not have the benefit of the community as a guiding principle. Research partnerships must ensure that the community voice is heard, valued, and protected. Moreover, sustainability in terms of relationship and partnership, funding, service delivery, public policy action, and commitment to future engagement should be addressed continuously.

If community partners are to use the RADx-UP Data Dashboard to ask and answer questions of immediate relevance to the other members of their community, we need to consider training options for those community members beyond what we have developed in the current infrastructure. Coursework, mentorship, and other formal and informal training typically benefit new and aspiring researchers; such capacity building grows trust by the community as a tangible asset which ensures that the community leads research and that successful outcomes are achieved.

CONCLUSIONS

The RADx-UP Data Dashboard and community engagement support infrastructure aim to foster community partner leadership of research and dissemination and to allow community partners to ask and answer valuable questions for their communities. Success and sustainability of this effort require a shift in current power dynamics and honest community partner engagement. These measures have the potential to reduce health inequities and associated disparities in chronic conditions by fostering alliances between community partners and academic researchers aspiring to understand the differences between root causes and associated disparities.