Integrating Race and Ethnicity Data in Public Health: Local, State, and Territorial Insights

December 17, 2024 | ASTHO Staff

Integrating race and ethnicity data in public health is critical to understanding and addressing health disparities among diverse populations. While there have been successes in this area, public health professionals must address ongoing challenges to ensure the accuracy and reliability of the data.

Cohort 4 of Diverse Executives Leading in Public Health includes subject matter experts who use data in their daily roles at local, state, and territorial health departments. Christopher Whiteside from New Mexico and Monica Tavares from Rhode Island offer state-level perspectives, while Eman Addish from Philadelphia, PA, provides a local viewpoint, and Emman Parian from the Commonwealth of the Northern Mariana Islands (CNMI) brings a territorial perspective. Whiteside and Addish both have backgrounds in epidemiology, and all have experience utilizing data for programmatic activities. They share insights into the successes and challenges of integrating race/ethnicity data in public health and future directions in this field.

What success and challenges exist in integrating race/ethnicity data to improve collection, completeness, and interpretation?

Local Perspective

EMAN ADDISH (Philadelphia, PA): On one hand, these efforts can significantly enhance data analysis, offering a more comprehensive view of health disparities among diverse racial and ethnic groups. Completeness of race and ethnicity data for people with different health behaviors or conditions may depend on how the data was collected and for the purpose. Matching to other registries, like immunization records and other disease registries, can help improve the completion of race and ethnicity data. Before we started matching our chronic viral hepatitis data to other registries, fewer than 50% of cases had complete race/ethnicity data. However, we have increased the completeness to 85-95% for newly reported cases.

That said, ensuring the accuracy of race and ethnicity data poses a challenge. Factors such as self-reporting, provider reporting, and data entry errors can affect the reliability of the data. Additionally, the lack of standardized data collection methods and definitions across different datasets hampers integration efforts and limits comparability. Addressing these challenges is crucial when integrating race and ethnicity data in public health.

State Perspectives

CHRISTOPHER WHITESIDE (New Mexico): When integrating race and ethnicity data in public health datasets, one must consider several key aspects. One is that the data promote equity when complete and inclusive of all races and ethnicities. Integrating race and ethnicity data allows for comparisons and increased awareness of structural racism and inequities. Within the New Mexico Department of Health, successes with integrating race/ethnicity have come from involving and explaining the reason for collecting these demographics with community members and the general public.

Unfortunately, these successes have also come with some challenges. For example, historically, marginalized communities do not have complete trust in the government or public health due to historical racism. Comparing race and ethnicity categories also poses a challenge in New Mexico because there are some racial groups with small numbers, which creates a statistical challenge and makes it difficult to compare appropriately. Consequently, data are often aggregated by year or geography, which allows more descriptive data by race/ethnicity but also means the data lose some specificity.

MONICA TAVARES (Rhode Island): There are many challenges to integrating race/ethnicity in data collection. As it exists today, the categories are incomplete, and it isn’t very clear or does not provide relevant options for individuals to describe their racial identity. Because of my unique view of race (I identify as Cape Verdean and consider this to be both a part of my racial and ethnic background), I am considered multiracial according to the official U.S. racial categories/definitions. This makes it difficult for multiracial individuals (those with two or more races) to see themselves accurately represented in the data. I know this concept to be true for many Hispanics and other individuals who often identify as multiracial.

In my view, the gold standard should be providing individuals with options to self-report their own race/ethnicity, allowing them to reflect their identity accurately. However, in the context of overdoses (fatal or non-fatal), a significant barrier arises when data is collected from individuals who are unconscious or deceased, making accurate self-reporting challenging. Managing the challenge posed by numerous self-reported factorial combinations in data aggregation requires implementing strong strategies. Utilizing advanced analytics can streamline this process, enabling the extraction of valuable insights despite the data's complexity. Additionally, creating a more uniform system for data collection and varying methods of inquiry across different systems can make it easier to compare data.

For overdose specifically, Rhode Island has been doing an excellent job of transitioning from counts to looking at rates, which highlights the racial disparities in overdose burden. We have also started doing this for naloxone to ensure equitable distribution of supplies. Over time, this will be helpful in guiding our efforts and getting materials where needed most!

Territorial Perspective

EMMAN PARIAN (CNMI): Challenges to integrating race/ethnicity in data collection, specifically in public health systems, present various gaps in collecting, monitoring, and analyzing data. For example, the population sample being collected is sometimes inaccurate, which is a collection challenge. There needs to be a standard in the collection of health equity data—one approach is to monitor the inconsistencies that are part of these systems, develop the standards, and work towards solutions. Then, it would be updating the information or the frequency of the data being fed and, lastly, analyzing the collected/monitored data. If the first two present various challenges, then it would be complex (i.e., inaccurate or incomplete) to analyze the data to produce the result from the analysis. We can combat this with improvement of data collection, focused on race/ethnicity.

CNMI continues to work on initiatives that produce better outcomes and healthier communities such as data modernization, building capacity amongst public health programs, increasing collaboration with stakeholders/partners, and working towards public Health accreditation.

What are some future directions in integrating health equity data to address disparities and improve health outcomes comprehensively?

Local Perspective

ADDISH (Philadelphia, PA): Health equity data integration efforts will aim to address disparities and improve outcomes comprehensively. Enhanced data collection methods will include intersectional data like the environment, neighborhood conditions, housing, socioeconomic status, and education, providing a nuanced understanding of disparities. Improved data sharing and interoperability among health systems will integrate diverse data sources (e.g., electronic health records and social determinants) for a more holistic view. Some indexes can be used with public health data, such as CDC’s Social Vulnerability Index, the Healthy Places Index, and the Index of Concentration at the Extremes. Additionally, there are indices that measure resilience, such as the Baseline Resilience Indicators for Communities. Community engagement will also ensure culturally responsive interventions by involving communities in defining questions and interpreting data. These directions collectively strive to create a more comprehensive understanding of health disparities and develop effective strategies to improve health outcomes for our communities.

State Perspective

TAVARES (Rhode Island): The future direction (at least for overdose) will be aggregating multiple data systems at the state level with self-reported race/ethnicity information and combining it into a database that we can use to inform our work. While this would miss individuals who don’t touch any systems, it could vastly improve some datasets. For example, for non-fatal overdose data in EMS, 85% of non-fatal overdoses have missing information. If Rhode Island could link this data with COVID-19 vaccine or COVID-19 testing data (while limited to people who got tested or vaccinated), we could dramatically improve that percentage, allowing us to better aggregate the data by race/ethnicity.

Ideally, systems statewide should use more rates than counts to highlight better racial inequities and control for each population's baseline size. The Biden Administration has made some proposals and commitments to do this, but the census must change how it asks questions on race/ethnicity. Some research from the Pew Research Center on American Trends Panel tackles this issue, but more needs to be done. As we get better at asking/collecting race/ethnicity information, we need data at the population level to calculate rates. If that data isn’t collected better, it is impossible to calculate rates.

Territorial Perspective

PARIAN (CNMI): Integrating health equity data in public health will involve modernizing data and advancing data quality. From an immunization program perspective, data is the driving factor in identifying what populations or subgroups need attention to improve vaccination coverage rates. What villages, gender, race/ethnicity, etc., have low vaccination rates? We utilize these data elements to pinpoint and guide our strategies. Our current immunization work heavily relies on data and guides strategies and activities. The COVID-19 pandemic brought visibility to health equity data. Utilizing cases of infection/vaccination data throughout the pandemic helped to address disparities and improve health outcomes—identifying which population or subgroup got high infection rates and low vaccination coverage to target vaccination strategies. Continuing to make advances in health equity data, such as modernization and improvement, will continue to contribute to creating healthier communities.

Conclusion

While there have been successes in integrating race and ethnicity data, challenges persist (e.g., ensuring data accuracy and overcoming historical distrust). Moving forward, it is crucial to continue refining data collection methods and incorporating social determinants of health data. Additionally, data must be collected and interpreted for a public health system that meets the needs of all individuals, regardless of race or ethnicity. Next steps for public health agencies and officials include:

• Enhancing data collection techniques by investing in advanced technologies and methodologies to improve the precision and comprehensiveness of data collection.
• Engaging with communities to build trust through transparent communication and community-driven research initiatives.
• Promoting data literacy through training and resources for public health professionals to effectively analyze and utilize diverse data sets.
• Advocating for policies that support equitable data practices and the allocation of resources to underserved communities.
• Encouraging cross-sector collaboration to leverage different expertise and perspectives in addressing public health challenges.
• Regularly assessing and updating data systems and practices to ensure they remain relevant and effective in meeting public health goals.