The Second Patient: Racialized Medicine and Provider Care

 

Patient X

Patient X, a middle aged Black woman who monitors her sugar and blood pressure for a history of type II diabetic and high blood pressure called at 3AM experiencing chest pain and shortness of breath. Patient X lived in the apartment complex for decades, but her neighbors opened the door for EMS. Two older black women were waiting outside of the bedroom door, night robes pulled tightly, telling me her pain had become so frequent that they would listen through the walls at night to hear if she was crying out. I announced my crew and entered the bedroom, Pt X was reclined in bed, propped up with pillows to stay off of her bad back and, careful not to lie too far back and exacerbate her labored breathing. The bedroom was small, with piles of folded clothes leaning against the wooden bed frame, plastic bags filled with pill bottles and blood pressure cuffs hung from the posts. 

My crew and I took a set of vitals as routine assessment and assisted her to the stretcher. Before leaving her apartment, she was insistent on bringing her own “machines”, a personal pulse oximeter and automatic blood pressure cuff. At first I took her assertiveness as impatience with my crew for taking a “wrong” reading, but I understood that she was scared. She wasn’t doubting our care, but distrustful of doctors that would disregard her pain as untraceable and nonexistent. Expectant of the fight against medical providers that would minimize her concern, asking her to prove her complaints, rather than care for her sickness. “I need my machines, I can’t leave without my machines. They tell me if I’m okay”

Racism in medicine

According to the Center for Disease Control and Prevention (CDC), African American communities represent the highest rates for preterm and low birth weight births and childhood chronic health conditions, culminating in the highest rates of adulthood hypertension and cancer mortality (Ellis et al. 2020). Disproportionate predisposition for long term medical conditions throughout all ages and socioeconomic statuses in tandem with racial inequity in the healthcare system represents the physical toll of the prolonged stressor of systematic oppression. Social adaptation to racism and inequality with increasing health disparity within the Black community represents the nature of resilience and the nurture of oppression as the prolonged experience of racism serves as a social determinant of health. 

Doctor’s offices represent the focal point of localized public health to facilitate patient care. The landscape for healthcare consists of a majority of white providers as 56.2% of US physicians are white compared to 5.0% of Black or African American physicians, despite serving diverse patient populations (“Diversity in Medicine: Facts and Figures 2019” 2019). This provider disparity reflects a lack of understanding histories of cultural distrust and racialized medicine instilled within the physician role as provider unawareness creates a subset of patient care for Black communities and the reality of patient care in a Black America. Implicit bias, a form of bias that occurs unintentionally based on subconscious judgements and microaggressions, against Black patients often exists within predominantly white offices and forms racialized concepts of the public health community that exclude Black patients from the standard of care. 

Racial bias influences pain management within preventative and emergency care as Black patients are more likely to receive inadequate pain treatment compared to the racial majority in America. As physician pain perception differs from patient pain ratings, bias about socio demographic groups influence medical judgement (Mathur et al., 2014). Race continues to influence provider treatment actions within emergency care, highlighted through the administration of pain medication for long-bone fractures. A study examined hospital cases of a long-bone fracture over a 40-month period for all black and white patients at an urban ED in Atlanta, GA. Evaluating the proportion of 127 Black patients to 90 white  patients with similar medical cases and pain complaints that received ED pain treatment, results found that Black patients were 22% less likely to receive pain medication. The overall risk of not receiving pain medication while in the ED was 66% greater for Black patients than for their white counterparts, specific to the pain complaint of a long-bone fracture (Todd et al., 2000)

Cultural competency training 

To understand the levels of overt and subconscious application of race-based guidelines as formal patient care, I interviewed Dr. Kelly Price, an experienced family physician that serves a predominantly white community. As a family doctor, she highlights the holistic approach to patient care that separates her from a specialized doctor as she gains a patient-provider relationship from the longevity of repeated appointments. “You see family units much more so than just one person. And I think that plays a role in the dynamics. So, if you know a kid struggling in school and the mother’s stressed out, you know more about the family function and dysfunction than if you just saw the patient as a cardiologist or pulmonologist,” she says, addressing the psychological and social interactions that factor into overall health. Price was mentored under two white male physicians that founded the practice, in which she was the first female doctor to join. 

The practice serves a patient community predominantly South or Southeast Asian, then Hispanic, and a minority of Black patients. Price attributes her patient care to experiential learning and peer debriefs as during her mentorship under the practice owners, and denies exposure to formal training in patient care for diverse populations. Price emphasizes her approach to the ritual of patient assessment and engagement as something developed from her predecessors rather than guidance in medical school or residency. “We didn't get any of that, it just wasn't part of the focus of education. I think that fresher residents get a little more training, but you really do pick up along the way.” 

She highlights a level attitude towards diversity training in which she claimed to provide equal treatment by not factoring in color into how she develops a provider-relationship. Although Price didn’t recall training modules for patient care, she had taken a “Culture and Diversity” training module administered by Atlantic Health System in 2021 that provided a template of patient behavior and pain responses by ethnicity. Price recalled learning that the module identified Hispanic patients as being “highly symptomatic” and more likely to provide medical complaint information, and as a result have communication that is “dramatic” with “moaning and groaning”. Conversely, Asian patients were stated to “minimize their pain and symptoms” in a stoic demeanor, and the module suggested that providers follow up with Asian patients to mitigate pain suppression. Finally, Black patients were listed to “be on edge” when being seen by a white provider in which the module suggested that providers be aware of health predisposition and patient health behavior. 

Price summarized the training module as the only cultural training she received, finding the simplification of patient ethnicity insightful for understanding various “cultural communication styles” outlined in the module. While Price claimed that she utilized the module’s recommendation when treating asian patients, by expanding her exam and conversational minutes to become more attune to suppressed pain, she denied factoring race into her patient relationships.

It’s historical, not natural

Disease burden and racial predisposition is historical, not natural. Race medicine influences the perception of pain, falsely projecting the white provider impression that Black patients have unique medical and physical bodies to white patients. In a 2016 study examining beliefs associated with racial bias in pain management among white medical students and residents showed that 40% of first and second-year medical students endorsed the belief that “black people’s skin is thicker than white people’s” (Hoffman et al., 2016). From a history of race medicine that pervades modern medical curricula/provider bias, Black pain and patient history is reevaluated by provider bias to under-treat patient complaint compared to the white patient. 

Historian of medical science and Africana studies, Lundy Braun, represents a large shift in scientific methodology of racial medicine towards the standardization of medical instruments and statistics. Braun’s work “Spirometry, Measurement, and Race in the Nineteenth Century” published in 2005 reflects a contemporary change in the production and application of racialized medicine in the form of health markers and vital taking as an indicator of overall health and disease susceptibility. The invention of the spirometer, a medical tool used to measure lung capacity vitals central to clinical research and diagnosing pulmonary illness in the scene of tuberculosis was marketed as a “seemingly discrete and objective entity” (Braun, 2005). Spirometer data was segregated between white and Black patients, establishing white patient data as the standard practice of vital taking. Measurements taken from the spirometer were compiled to confirm biased medical statistics such as “mean difference” between racial groups concerning lung capacity that racially governed medical practice to alienate the Black body from standard care.

Spirometry advertisement from The Medical Gazette 1870

Race language

The social lens of intergenerational trauma represents how race is inherently social but overtime effects of racism are internalized, presenting as Black predisposition for chronic cardiac and metabolic diseases. Evaluating the social determinants of health for Black patients, it’s important to understand how provider implicit bias compounds the effects of structural racism and epigenetics to exacerbate the current health disparity between Black patient health and their white counterparts. Examining the US medical education and curriculum concerning patient care for culturally or ethnically diverse patients exposes how provider bias begins with misrepresentation of race within formal training, or lack thereof. Specially within categorization of race within medical curricula, the use of race-based clinical guidelines endorse the use of racial categories in diagnosis and treatment of diseases without framing superficial guidelines with structural, political, or socioeconomic context (Amutah et al. 2021). 

Dr. Price’s minimal experience with diversity training, relying on patient care approaches passed down from mentors and the closed-loop of provider discussion, is representative of the mainstream medical knowledge of patient care that internalizes culture and ethnicity as health itself, rather than social determinants of health. Superficial training modules that transfer race and ethnicity into medical templates for provider ease undermine social factors of health such as implicit bias, yet rely and test on race-based medicine that reflects the negative health effects of racism and structural violence against ethnic patients that are exemplified. 

Bias in, bias out. Medical AI algorithms represent the new face of racialized medicine 

As digital medicine becomes a proxy for face-to-face provider impressions and personalized treatment plans, predictive algorithms are subject to multiple layers of data bias in which ML models are trained against Black male patients. Algorithmic bias within medical AI is vulnerable to data sets that represent historical bias pre-existing within society, and representation bias based on training a model on data sets that lack diverse samples of the population it will serve. Within this scope of historical and representation bias, provider bias that disproportionately triages Black adult patients to lower acuity areas is rooted in differences in provider assessment based on subjective patient intake information such as primary complaints and lists of patient symptoms (Peitzman et al., 2023). While the ML model works with “raw data”, the algorithm is subject to similar bias as provider impressions and finds discriminatory societal patterns based on the biased training data and replicates them through the output of biased predictions (Belenguer, 2022). The application of ML training using biased data sets outlines the “parameters” of minority group membership as it reproduces health disparity between racial groups. 

A landmark study published in the journal Science in 2019, assessed racial bias against Black male patients within an algorithm used by hospital networks to identify high-risk patients that qualify for additional resources to manage their health, such as competitive specialized treatment plans for chronic illnesses. The study audited Optum, the health services company that developed the algorithm, and found that the algorithm overpredicted Black male patients' health and falsely concluded that Black patients were healthier than equally sick white patients. The algorithm was found to use total healthcare costs as a proxy for predicting illness severity in which the data reflects racial barriers to healthcare that result in lower health cost expenditure. Researchers estimate that racial bias reduces the number of Black patients identified for further preventative care by more than half as the fraction of Black patients within the threshold to qualify for care would rise from 17.7 to 46.5% (Obermeyer et al., 2019). Machine learning (ML) models are trained on biased data that perpetuate pre existing socioeconomic and racial health disparity as the generalization of Big Data undercut the quality of preventative care to minority communities of Black patients compared to their white patient counterparts. 

Medical AI changes the landscape for preventative care as well as emergency care that continues to discriminate against the minority community of Black male patients. In the journal, The Lancet Digital Health, Dr. Rahuldeb Sakar, a consultant physician of respiratory medicine and critical care at NHS Trust and researcher at King’s College London, assessed severity scoring systems widely used within the intensive care unit (ICU) in which the illness severity scores are used to inform triage decision for allocation of scarce resources such as ventilators during the COVID-19 pandemic. The study investigated the performance of four severity scoring systems that assessed vitals, past medical history, sepsis, and organ failure to predict treatment outcomes across four ethnicities in two large ICU databases, consisting of 60,000 ICU admissions, to identify racial bias. The study found patterns of over-predicting mortality for Black male patients versus calculations for Asian and White patients in which the systematic differences across ethnicities suggested that illness severity scores reflected statistical bias in their mortality prediction (Sarkar et al., 2021). 

Racial health bias is internalized within predictive medical AI that are trained through race demographics and parameters, yet claim to implement race-neutral algorithms. An approach called ‘Fairness through unawareness’ develops ML models by omitting racial stratification, which emphasizes the social construction of race and its contributions to social inequalities, and instead inferrs minority community membership from correlated data such as Optum’s cost versus severity predictors (Huang et al., 2022). Direct applications of predictive algorithms are race-neutral, choosing to exclude race stratification, and argue that race is not a biological component, but remain insensitive to social implications of implicit bias and racism. 

Unlike AI, patient care is not a black box

The danger of false objectivity surrounding digital medicine as race medicine within predictive medical algorithms becomes a detrimental proxy for provider care. This has direct implications for the broader question of racial health equity as race language within provider-patient relationships and medical AI work to assess and treat by race identifiers, yet exclude the critical context of social determinants that disease burden stems from. Race medicine is replicated and reinjected into current medical curricula and medical technology by relying on the coarse use of race language to imply biological differences rather than social consequences. 

Medicine derived from an isolated history of racial equality and health may strive to be color-blind, to treat all patients as equal, but the persisting health impact of white racism demands health equity. The mechanism through which generational racism has been biologically internalized through health concerns of a social body under duress cannot afford to be diagnosed through the proxy of race and call medicine “color-blind”. As racialized medicine shapes preventative care, emergency medicine, and patient diagnostics to value the white body over all else, medicine needs to systematically talk about race through the lens of racism and provider bias. As racial provider bias persists in the exam room and within predictive medical treatments, understanding Black disease burden and predisposition as historical trauma rather than inherent difference lies beyond cultural “parameters” and must be the provider’s responsibility to internalize in order to truly “do no harm”. 

References

Amutah, C., Greenidge, K., Mante, A., Munyikwa, M., Surya, S. L., Higginbotham, E., Jones, D. S., Lavizzo-Mourey, R., Roberts, D., Tsai, J., & Aysola, J. (2021). Misrepresenting Race — The Role of Medical Schools in Propagating Physician Bias. New England Journal of Medicine, 384(9). https://doi.org/10.1056/nejmms2025768

Belenguer, L. (2022). AI bias: exploring discriminatory algorithmic decision-making models and the application of possible machine-centric solutions adapted from the pharmaceutical industry. AI and Ethics, 2(2). https://doi.org/10.1007/s43681-022-00138-8

Braun, L. (2005). Spirometry, Measurement, and Race in the Nineteenth Century. Journal of the History of Medicine and Allied Sciences, 60(2), 135–169. https://doi.org/10.1093/jhmas/jri021

Diversity in Medicine: Facts and Figures 2019. (2019). AAMC. https://www.aamc.org/data-reports/workforce/report/diversity-medicine-facts-and-figures-2019#:~:text=About%2030%25%20to%2040%25%20of,and%2030.6%25%20of%20White%20physicians.

Hoffman, K. M., Trawalter, S., Axt, J. R., & Oliver, M. N. (2016). Racial bias in pain assessment and treatment recommendations, and false beliefs about biological differences between blacks and whites. Proceedings of the National Academy of Sciences, 113(16), 4296–4301. https://doi.org/10.1073/pnas.1516047113

Huang, J., Galal, G., Etemadi, M., & Vaidyanathan, M. (2022). Evaluation and Mitigation of Racial Bias in Clinical Machine Learning Models: A Scoping Review (Preprint). JMIR Medical Informatics, 10(5). https://doi.org/10.2196/36388

Mathur, V. A., Richeson, J. A., Paice, J. A., Muzyka, M., & Chiao, J. Y. (2014). Racial Bias in Pain Perception and Response: Experimental Examination of Automatic and Deliberate Processes. Journal of Pain, 15(5), 476–484. https://doi.org/10.1016/j.jpain.2014.01.488

Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019). Dissecting Racial Bias in an Algorithm Used to Manage the Health of Populations. Science, 366(6464), 447–453. https://doi.org/10.1126/science.aax2342

Peitzman, C., Carreras, A., Samuels-Kalow, M., Raja, A., & Macias-Konstantopoulos, W. L. (2023). Racial Differences in Triage for Emergency Department Patients with Subjective Chief Complaints. Western Journal of Emergency Medicine, 24(5). https://doi.org/10.5811/westjem.59044

Sarkar, R., Martin, C., Mattie, H., Gichoya, J. W., Stone, D. J., & Celi, L. A. (2021). Performance of intensive care unit severity scoring systems across different ethnicities in the USA: a retrospective observational study. The Lancet Digital Health, 3(4), e241–e249. https://doi.org/10.1016/s2589-7500(21)00022-4

Todd, K. H., Deaton, C., D’Adamo, A. P., & Goe, L. (2000). Ethnicity and analgesic practice. Annals of Emergency Medicine, 35(1), 11–16. https://doi.org/10.1016/s0196-0644(00)70099-0