Improving Collection Of Self-Reported Race & Ethnicity Data

Addressing health inequities requires detailed and accurate race and ethnicity data. This information is critical to improving health equity and outcomes for historically marginalized populations. It's also increasingly becoming a part of accreditation standards, quality requirements, value-based contracts, and federal initiatives.

As provider and payer organizations formulate a variety of interventions to improve health equity, decrease disparities, enhance individual and population health, reliable and timely race and ethnicity data is of prime concern. Yet, the exact data that can support these achievements is the information that many individuals are afraid to provide. The information may be withheld due to past experiences of racism, or because the current categories and standards don't align with the diversity of race and ethnicity in the US.

Race and ethnicity data lack robust consistency, availability, and accuracy

Even though most US hospitals collect data on an individual's race or ethnic group, few are proactive in doing so, and the data quality varies greatly. For example, race and ethnicity data are available for nearly all Medicare beneficiaries. And all Medicaid and children's health insurance program (CHIP) agencies ask applicants to voluntarily self-report race and ethnicity, although it's not mandatory for coverage. Half of Medicaid health plans and more than 25% of Medicare plans have race and ethnicity data gaps.

In 2019, 70% of states (36 of 50 states plus D.C.) had serious concerns about their race and ethnicity data, ranked as a medium or high level of concern or simply unusable. The DQ Atlas lists only 15 states that had low concern about their race-ethnicity data.

The level of data available is worse for commercial health plan members. According to researchers from the National Committee for Quality Assurance (NCQA), “while 61.2% of working adults ages 19 to 64 received health insurance through an employer in 2019, less than 25% of commercial health plans had race data for even half of their members."

Race and ethnicity do not have a well-defined biological correlation

As published in the May 2022 edition of Nature Medicine, “Ethnicity is a complex concept, incorporating notions of race (the classification of people based on physical appearance), religion, and culture, especially for individuals born into families with multiple heritages."

When the State Health Access Data Assistance Center (SHADAC) conducted an audit of state Medicaid enrollment applications, they found substantial variation in the race-ethnicity categories used. They ranged from five to 37 race categories and two to eight ethnicity categories. Some states even offered a different set of categories between their online application and their paper application. The SHADAC audit identified 62 unique race categories across all 50 states.

When the DQ Atlas looked into this, they found that 55% of US states have one or more categories of race and ethnicity that do not align with the race-ethnicity estimates from the American Community Survey, adding to concerns about accuracy from existing sources.

Are individuals hesitant to self-report race and ethnicity?

A 2020 study, published in JAMA Network Open, found that 21% of patients reported experiencing discrimination as part of a healthcare encounter. About 75% of those experiences had to do with race-ethnicity. These experiences make patients reticent to share personal information, particularly race and ethnicity.

The good news is that research have revealed that patients may not be as reticent to share race-ethnicity information as previously thought. A survey of over 3,000 Americans revealed with whom participants were willing to share race and ethnicity information:

• 81% are comfortable sharing with their provider
• 79% with their health plan
• 75% with a government health agency
• 76% with their employer
• 6-8% were not comfortable sharing their race-ethnicity with these entities

Honoring any personal choice to share or not share race-ethnicity data is critical. As we'll explore later, giving patients clear information about the purpose of collecting their information is key: how it will be used and protected, as well as how sharing could benefit them. Compassionate exchanges could positively impact the relationship and yield higher-quality data.

What's possible? Race and ethnicity data success case

One of the largest health institutions in New York City, providing care for over 150K inpatient admissions and four million outpatient visits every year, wanted to improve race and ethnicity data capture in an outpatient setting. 

The institution designed a program that enabled them to achieve a 76% improvement in the completeness of race-ethnicity information.

Prior to this initiative, the health system did not systematically collect demographic data and passively requested patients to report race-ethnicity. Even though New York's Statewide Planning and Research Cooperative System (SPARCS) instituted mandatory requirements for race-ethnicity data collection, these data were found to have missing or inaccurate entries.

The organization launched a multifaceted approach:

• assessment and evaluation of system needs
• modification of data infrastructure to align with goals
• training and education of relevant stakeholders
• implementation and responsive action to results
• acknowledging limitations and lessons learned

The program team implemented a patient registration data collection improvement process (PRDCIP), focusing on methodological interventions like improving system workflows, communication strategies, and procedural training.

Of particular interest is their use of cross-departmental collaboration, stakeholder engagement, institutional support, and culture of anti-racism that they found essential to their success. Additionally, their diversity staff provided unconscious bias training, “an essential step towards establishing equity as a collective priority and aligning stakeholders' understanding of their roles in reducing disparities."

The institution's long-term goal is to achieve >90% patient self-identification of race-ethnicity.

Strategies to improve race-ethnicity data collection

Like the health system in NYC, many healthcare organizations know the vital importance of reliable and self-reported race-ethnicity data. And, just like new mindsets and cultural commitments to equity are foundational to positive health outcomes, so too is the data behind each individual we serve.

Moving from passive to intentional race and ethnicity data collection

As explored above, while demographic and race-ethnicity information is collected, it's passive and doesn't have the rigor it should. One of the first steps to improving self-reported data is to make it an organizational priority and set intentional policies, procedures, and training in place to ensure data collection is proactive, compassionate, and gathered accurately.

Multiple stakeholders play key roles in this undertaking, including diversity and inclusion, IT, medical records, patient/member services, cybersecurity, enrollment, and data analytics or informatics representatives. Together, they can assess current data quality, determine which race-ethnicity frameworks and categories they need to comply with or want to institute, engage the communities they serve about how sensitive data is collected and why, as well as create a roll-out plan that includes education and training for employees, clinicians, and volunteers.

Take a respectful approach to collecting race-ethnicity data

A respectful approach to race-ethnicity data ensures that the exchange, or encounter where information is exchanged, honors the potential risk the patient may feel and takes a culturally competent communication style when sharing the reasons for collection, as well as why the organization can be trusted with this information.

One example of a thoughtful approach is Florida's Medicaid application instructions, which explain that individuals are not required to answer race-ethnicity questions. However, sharing race-ethnicity information may help the person get the most help possible. “Providing your race and ethnicity can be helpful since it can speed up the application process. It may be used to automatically create your case."

The American Hospital Association (AHA) created a Toolkit for Eliminating Health Care Disparities to equip healthcare organizations with practical steps, scripts, and samples. Their response matrix provides examples of questions or responses when collecting race-ethnicity information and how to reply in a culturally-competent, respectful matter.

Another respectful example is from Henry Ford Health System. Their “Why We Ask" materials outline why collecting this information is important, what health disparities are, and how the organization uses the data to help them receive better care.

In 2021, Blue Cross Blue Shield created an equity report after review of 2019 commercial member data to uncover inequities. From their research, they published best practices health plans can implement to improve race and ethnicity data.

Related to cultural competency and taking a respectful approach, they recommended that “all staff who are collecting data should be trained on how to ask members for their personal data. The data collectors should be knowledgeable on relaying to patients how the data will and will not be used, the importance of collecting the information, and discuss the measures the organization is taking to protect their personal information." And they emphasized transparency. “It is vital to explain to members why these data are being collected and how they will be used to build trust."

How could you improve health equity and outcomes with better, more accurate race-ethnicity data?

RTI Health Advance supports data science rigor while, also, ensuring that health equity goals don't get in the way of building positive patient experiences and encounters. Let's discuss your health equity data strategy and how best to assess your organization's cultural competence and race-ethnicity data collection approach. Contact us.