The Racial Differences Hidden in the Data

by Gayle L. Capozzalo, FACHE, director of The Equity Collaborative – The Carol Emmott Foundation and
Douglas Riddle, PhD, DMin, curriculum director of The Carol Emmott Foundation
Published in 
partnership with the American College of Healthcare Executives

The importance of social determinants of health is no longer new to most healthcare systems. Many healthcare organizations have taken steps to positively affect the communities they serve, including the circumstances of their own employees. What’s missing is sufficiently detailed demographic data that captures the pervasive influence of structural racism and other marginalizing factors that affect health outcomes.

Reporting population averages on any measure may disguise the problems caused by biased systems. Our contention is that large improvements in quality, safety and outcome metrics will come for healthcare institutions only when we begin to track the impact of race, gender, sexual orientation, ethnicity, disability status and other crucial factors on those metrics.

In her recent book, The End of Bias: A Beginning, Jessica Nordell devotes a chapter to the inferior care women, especially women of color, receive in part because of provider bias, myths and the lack of participation in medical research studies. Many landmark studies on aging and heart disease never included women because medicine considered the male body as the default and women a subcategory that could be safely left out of studies. Women of color are at particular risk for poor treatment. A recent analysis of their childbirth experiences found they frequently encountered condescending, ineffective communication and disrespect from providers. Socioeconomic issues do not explain all the poor outcomes. The gap between heart disease rates for Black and white women is greatest at the highest levels of education.

The concept of treating man as the default human and not studying sex differences in medicine is the cause for much of the poor outcomes for women, particularly women of color. Recent studies found sex differences in every tissue and organ system the human body, as well as in the prevalence, course and severity of the fundamental mechanical working of the heart. There are still vast data gaps, but the past 20 years have proved that women are not just smaller than men: male and female bodies differ down to a cellular level.

Why don’t we practice medicine and conduct research using this knowledge? The challenge is the universal and subjectively invisible nature of bias, especially when it is baked into the structures of healthcare. We need metrics that lead to change, and a few organizations are beginning to drill down to more specific and actionable measures that can lead to significant improvements in safety, quality of care and health results for all.

The National Committee for Quality Assurance has started including stratification by race and ethnicity in its health plan quality measure set. In 2006, the National Institutes of Health declared the importance of measuring to uncover health disparities. However, in 2019, 76% of commercial health plans had incomplete race data for their members, and 94% had incomplete data on ethnicity. Information is not available on how many health systems and hospitals are currently stratifying their information on treatments and health outcomes by these factors, but some resources are available. For example, the American Hospital Association provides tools and guidance on data collection, stratification and use of the stratified data, emphasizing REaL data (race, ethnicity, language). The AHA cites Henry Ford Health, which collects REaL data on more than 90% of patients through its “We Ask Because We Care” campaign.

At a recent meeting of The Equity Collaborative (an initiative of the Carol Emmott Foundation), leaders from the University of Chicago Health shared their preliminary steps toward actionable metrics to achieve equitable health outcomes and a more equitable environment for healthcare workers. Among those steps is reorganizing their data warehouse to focus on intersectional data. That drove the creation of a section of the organization’s operating plan for the current fiscal year that identified people goals stratified by factors known to be marginalizing (gender, race, ethnicity, etc.), including employee engagement, clinician engagement, workforce turnover, workforce diversity and promotion rates. University of Chicago Health’s Clinical Excellence Scorecard also includes an equity and opportunity lens on all quality data that reports differences related to gender, race, ethnicity, age, financial source and ZIP code.

Yale New Haven Health’s new office of health equity is developing a data infrastructure that can generate actionable responses. The first stages are involved in aligning work with the Yale University Office of Health Equity Research. One lesson that has already emerged is the value of asking patients and employees to define their “racial identity” rather than “race.” This reflects the complex nature of social identity and its impact on care, something that has required changing terminology in its EHR system. The alignment has allowed a comparison of COVID mortalities by race among patients with population data.

These disparities in healthcare have persisted throughout our history, but real change is possible. Getting the metrics right and having the right metrics are critical to bringing real equity to healthcare. The ubiquity of electronic health records and the possibility of a more universal understanding of disparate impact can make the difference.