Find Resources

The nonprofit Center for Open Data Enterprise and the U.S. Department of Health and Human Services (HHS) convened the Roundtable on Sharing and Utilizing Health Data for AI Applications in 2019. These discussions identified five high-level applications of AI in healthcare. First, AI automation tools can be utilized to improve clinician and administrative workflows and reduce the costs and administrative burden of manual data entry and billing processes. AI is also a valuable tool that can be used to connect patients with available health and social service resources, especially in rural areas. Researchers can use AI to inform and advance population health management by synthesizing data to tell clearer stories about the prevalence and incidence of different diseases. AI has the potential to be a revolutionary tool in the medical diagnosis space. Lastly, AI can aid in the development of new drugs throughout the entire cycle, from matching clinical trial participants to appropriate trials to contributing to precision medicine research. Importantly, the report emphasizes the dependence of all these promising AI applications on the widespread sharing of high-quality, clean, and accurate data.

The Robert Wood Johnson Foundation formed the National Commission to Transform Public Health Data Systems in response to the health inequalities and insufficient cross-sector data-sharing capabilities that characterize the American health care system. The 16-member Commission, comprised of experts from the healthcare, community advocacy, business, government, and public health spaces, was tasked with examining the data and data-sharing systems needed to ensure a public health system that is representative of the situations and needs of all Americans. In a report published in 2021, the Commissions put forth a series of recommendations focused on the need to remedy the current disconnect between data on health inequalities and the history and social conditions that shape poor health outcomes to begin with.

This study seeks to understand the unique experiences of gay and bisexual men who live in culturally conservative, rural communities and the implications of these experiences for the provision of health care. Researchers administered qualitative interviews to 44 individuals to determine the ways that their environments impacted their health and social behaviors. Results of this study revealed the powerful role that stigma plays in inhibiting access to health care and social support networks. The authors emphasize the need for greater discussion and awareness campaigns around stigma as an etiologic factor that impacts the health and well-being of rural gay and bisexual men. This is especially important in light of the fact that the incidence of HIV in America is particularly prevalent in rural communities where there are dispersed populations, fewer health care resources, and a lack of culturally responsive providers that understand the needs of stigmatized populations.

Researchers conducted a series of interviews with administrators from 40 different nursing homes across the US between July 2020 and December 2021 to better understand their experiences amidst the COVID-19 pandemic. Interviewees described the adverse effects of the pandemic on the physical and mental health of the residents in their facilities as restrictions around visitation, socializing, and communal dining created isolating, inactive conditions. Administrators also expressed concerns for the longevity of the nursing home industry in the wake of the pandemic and resulting reductions in revenue and pervasive negative perceptions of nursing homes in the media.

This study assessed differences in COVID-19 cases and mortality rates in traditional nursing homes and Green Houses. Green Houses are smaller, more intimate residential living environments that promote personalized care and more tight-knit communities of residents and care givers. Using CMS data, researchers found that, adjusting for age, gender, and disability status, small and large nursing homes had higher COVID-19 cases and mortality rates compared to Green Houses.

This article discusses the need for a disability data justice framework to support health equity goals. Even though 27% of American adults report having a disability, there exists limited COVID-19 data for the disabled community. Some public health surveys administered during the pandemic did not incorporate questions around disability until 2021. The article presents a series of recommendations on ways to achieve disability data justice. These include making disability a core element of all demographic data collection alongside race, ethnicity, and gender identity, increasing representation of disabled people across health professions, collaborating with members of the disabled community when developing and implementing disability data collection, and making health data accessible to all people with disabilities.

To make sure we’re accurately understanding people’s health, we need to use the same measuring tools in the same way for everyone, no matter who they are. In this study, Cintron and their team checked if a widely used depression test, the Patient Health Questionnaire (PHQ-8), works the same for different groups of people based on their age, gender, education, and race. They found that in about 24% of the cases, the test worked a little differently for these different groups. However, this is just slightly below the 25% limit we usually use to decide if the test is working the same for everyone. So, in general, the PHQ-8 seems to measure depression in a similar way for all the groups they studied. To make sure we’re treating everyone fairly when collecting health data, we should do similar research for other common tests and questionnaires to see if they work the same for different groups of people.

This episode from the podcast 99 Percent Invisible reflects on how the COVID-19 pandemic exposed the American public to a long-standing problem: the disjointed nature of the U.S. health system. Because state and local health departments largely operate independently and use their own data collection and analysis tools, health systems across the country lack standardized data definitions and systems. These inconsistencies made it nearly impossible to collect and analyze comprehensive, standardized data on COVID-19 cases, deaths, and vaccines administered amid the pandemic. Health experts featured on this episode believe that the pandemic made the need for an overhaul of America’s informatics system very apparent. When rebuilding this system, it’s important to focus on remedying existing inequalities in data collection and classification that in some cases render the health status of certain populations – think Native American communities and other communities of color – completely invisible in the data. By virtue of being small populations, it can be difficult for health departments to collect sufficient and/or statistically significant data on minority communities. Another issue discussed in this episode is the use of broad racial categories like “other,” “multiple races,” or even “Asian American,” which, if not disaggregated, obscures the health status of diverse populations who are grouped under the same category. Without comprehensive and inclusive health data, it’s difficult to identify disparities and implement policies and programming that promote social mobility and health equity.

This study from the American Journal of Managed Care identifies the main factors influencing the implementation and sustainability of cross-sector data sharing between a community-based organization and an academic medical center via a case study of the Mid-Ohio Farmacy (MOF) produce referral program. The MOF is a unique program developed by the Mid-Ohio Food Collective and Ohio State University Wexner Medical Center (OSUWMC) to provide food insecure patients who have a qualifying chronic condition with money to purchase produce from food pantries on a monthly basis. Researchers conducted a series of key informant interviews with OSUWMC and Mid-Ohio Food Collective representatives to characterize the challenges of data sharing between the academic medical center and the community-based organization. Challenges were grouped into three main buckets: data sharing regulations, data exchange capabilities, and cross sector data integration. Some keys findings from this paper include the need for clarification around HIPPA regulations for data exchange between covered and non-covered entities, the importance of incorporating social-risk data (such as a food security screening tool) into the electronic health record, and the need for additional research and guidance around how to create fully integrated shared databases between health care and social service agencies.

Data Across Sectors for Health (DASH) is a national initiative funded by the Robert Wood Johnson Foundation and co-led by the Michigan Public Health Institute and the Illinois Public Health Institute. Created in 2015, DASH promotes and supports multisector data-sharing ecosystems with the goal of fostering more equitable, informed decision-making practices and ultimately improving community health outcomes. They provide financial support, technical assistance, resources, and programming to help foster community collaborations around data-sharing. Their website features information on their two active funding programs, a link to the DASH data-sharing framework and accompanying webinars, as well as information about their knowledge-sharing platform, DASH Knowledge Base, that is still in development. DASH is a great resource for organizations and community leaders seeking guidance on how to foster relationships and enhance data sharing capabilities between governments, community-based organizations, and other local players.