The NAACP Calls for the Elimination of Race-Clinical Equations in Patient Care

WHEREAS, Clinical algorithms have been used to support clinical decision-making in patient care and are routinely based on the predictive power of population-level datasets, the use of race as a modifier in treatment decisions has been disputed. It is well understood that race is a political and social construct, not a proxy for genetic differences. When race and ethnicity are used in "diagnostic [criteria] and practice guidelines [to] adjust or 'correct' their outputs," this defines race-based medicine (Vyas et al., 2020, p. 874). The role of race-based clinical algorithms in assessing individual risk and guiding clinical treatment has led to disparate and harmful impacts in historically excluded populations; and

WHEREAS, Kidney function can be assessed using equations to diagnose and stage chronic kidney disease and guide medical decisions. Race correction assigns a different value for a person being Black in an equation to evaluate kidney health, giving rise to an estimated glomerular filtration rate (eGFR) that is higher based on race alone. Race modification in kidney health based on inherent biological differences is patently false, which leads to inequities in kidney care. "A 2021 national study estimated that removing race adjustment from the MDRD eGFR equation could result in an additional 3.3 million Black Americans receiving a stage 3 CKD diagnosis, 300,000 more qualifying for a nephrologist referral, and 31,000 becoming eligible for transplant evaluation and inclusion on a waitlist" (NYC CERCA, 2022, p. 10); and

WHEREAS, The National Kidney Foundation and the American Society of Nephrology have recommended alternatives to race-based eGFR estimates since 2021, multiple barriers remain to wider implementation; and

WHEREAS, Black women/pregnant people are approximately three times more likely to die from pregnancy-related causes than their White counterparts in the United States, the vaginal birth after cesarean (VBAC) risk calculator may worsen maternal health disparities. Pregnant people are offered a trial of labor after the cesarean section (TOLAC) to increase the chances of a subsequent safe vaginal delivery. Data has shown that Black and Latino pregnant people experience higher rates of primary cesarean deliveries and lower rates of VBAC. In the VBAC risk calculator, Black and Latino women are assigned a different numerical value than White women based solely on race and ethnicity criteria, which predicts a lower probability of successful VBAC. This race correction is due to proposed ethnic differences in pelvic anatomy, and the validation study that identified being White, among other social characteristics, as a protective factor associated with a greater chance of successful VBAC. Because of this flawed inflation of risk in certain racial and ethnic groups, Black and Latino pregnant people are not recommended for TOLAC, and maternal inequities are perpetuated; and

WHEREAS, Researchers have found that using a VBAC calculator that excludes race and includes the current treatment of chronic hypertension accurately predicts successful VBAC among pregnant people choosing to have a TOLAC. The American College of Obstetricians and Gynecologists (ACOG) issued a practice advisory in December 2021, endorsing the non-race-modified VBAC calculator; and

WHEREAS, there are other examples of race-based clinical algorithms in the literature that have caused disparate impact on Black populations and other historically excluded groups, including pulmonary function test race modifiers, race-based equations to predict the risk of urinary tract infections in children, and breast cancer surveillance calculator, among others.

THEREFORE, BE IT RESOLVED, the NAACP strongly opposes using race-based clinical equations and clinical decision tools where there is evidence of disparate outcomes and perpetuation of health inequities. The NAACP demands that the Department of Health and Human Services (HHS) implement and enforce rules to prevent discrimination based on race due to using clinical decision-support tools and algorithms. The NAACP demands that HHS republish the systematic review done by the Agency for Healthcare Research and Quality on algorithmic bias and enforce regulations of clinical algorithms through fines and sanctions for those who fail to address algorithmic bias.

BE IT FURTHER RESOLVED, the NAACP affirms that "a biological definition of race in medicine is harmful and inaccurate, [rather] using race as a political or social category to study the impact of structural racism and its physiological effects is extremely relevant," as stated in the House Ways and Means Committee Fact Versus Fiction: Clinical Decision Support Tools and the (Mis)Use of Race issued by Majority Staff on October 14, 2021.

BE IT FURTHER RESOLVED, the NAACP demands that Section 1557 of the Affordable Care Act be enforced that prohibits covered entities from discriminating on the grounds of race, color, national origin, sex, age, or disability in the administration of health programs and activities, including preventing covered entities from using algorithms that could lead to discriminatory practices.

BE IT FURTHER RESOLVED, the NAACP demands that HHS enforce Section 1557 require algorithm developers to demonstrate how they have addressed the risk of bias and make this information publicly available, allowing users to review how algorithms were tested for fairness. The NAACP demands that regulatory frameworks require the collection and analysis of data on algorithm use and their impact on different racial and ethnic groups to identify and address any disparities. The NAACP also demands that regulatory agencies conduct regular audits of algorithms used by payors to ensure compliance with fairness standards and a strong enforcement mechanism, such as fines or sanctions, for those who fail to address algorithmic bias.

BE IT FURTHER RESOLVED, the NAACP calls for regulatory agencies to collaborate with healthcare systems, technology developers, and advocacy groups to develop and implement effective strategies for addressing algorithmic bias and, where appropriate, to establish legally enforceable responsibilities.

BE IT FURTHER RESOLVED, the NAACP calls for comprehensive efforts to improve data quality and ensure algorithms are trained on diverse and representative datasets, which can help reduce bias. The NAACP also calls for regulations to encourage collecting and using data on social determinants of health to understand and address the factors contributing to disparities in care. The NAACP advocates for identifying, recruiting, training, hiring, and developing diverse and inclusive healthcare providers to ensure high-quality, safe, and culturally responsive care for all, and training all healthcare providers on how to identify and address algorithmic bias to ensure clinical support tools are used appropriately and ethically.

BE IT FINALLY RESOLVED, the NAACP calls on state legislatures and state attorneys general to prevent racial bias in algorithms through legislation and investigations (such as the actions undertaken by the California AG to request information from hospital CEOs about their use of clinical decision-making algorithms and the District of Columbia Stop Discrimination by Algorithms Act), and on municipal governments to intervene to end the use of racially biased algorithms, using the New York City Department of Health and Mental Hygiene's Coalition to End Racism in Clinical Algorithms (CERCA) as a model to convene health systems, academic centers, and clinicians to establish time-bound goals for eliminating race correction in clinical care.