Racism isn’t always intentional and fuelled by malice. Especially in the healthcare sector, implicit bias thrives under long-established racial stereotyping that extends from doctor to patient and even vice versa. Health inequity stands as one of the most prevalent concerns in racial disparity. Often, people of colour are less likely to be given the same medical treatment or appraisal by medical professionals – but that implicit and often non-malicious bias has translated to artificial intelligence. This shocking anomaly came to head in a jarring case in October 2019 when a health risk algorithm was found to be perpetuating existent inequality.
The algorithm in question was calibrated to help hospitals and insurance companies identify which patients would benefit from “high-risk care management” programs, which provide chronically ill people with access to specially trained nursing staff and allocate extra primary-care visits for closer monitoring. By singling out sicker patients for more organized and specific attention, these programs were aimed at active preemption of debilitating illnesses, reducing costs and increasing patient satisfaction.
Researchers found that the algorithm assigned risk scores to patients based on total health-care costs accrued in one year. The assumption was initially deemed reasonable as higher health-care costs were generally associated with greater health needs. However, a point of contention arose as scientists set the average African American person in the data set as having similar overall healthcare costs to the average white person. Through this, the algorithm would have swiftly preserved the systemic racial inequality it was built to reduce.
It is indisputable that racial inequalities are layered on top of economic inequalities, as there is a population-attributable risk of being of low socioeconomic status and belonging to a minority. This reflection of flawed human perception translated into a dispassionate program ironically left the indiscriminate creation as inherently discriminatory.
Race and income have always been correlated; people of colour are more likely to have lower incomes. And poorer patients, even when insured, tend to use medical services less frequently or have reduced access to them because of time and location constraints. For example, among all patients classified as very high-risk, African American individuals turned out to have 26.3 per cent more chronic illnesses than white ones (despite sharing similar risk scores). Since their lower recorded health care costs matched those of healthier white people, the program was less likely to flag these eligible patients for high-risk care management, causing a large proportion of African American patients to lose access to critical care.
The routine implicit bias people of colour face mean they receive lower-quality care and have less trust in the doctors whom they feel are exhibiting bias which leads to a continuation of healthcare disparity.
Several studies show that African-American patients are often prescribed less pain medication than white patients with the same complaints. A 2016 Journal of Clinical Oncology study found that African-American women, regardless of their risk level, are less likely than white women to undergo genetic testing – largely because physicians are less likely to recommend it to them.
Marginalized populations are also more likely to been seen as violent in emergencies and are more likely to be presumed to be non-compliant with their medication, medication-seeking, or having an ulterior motive to receiving needed care.
People of colour on average have far less trust in medical professionals than their white counterparts, reflecting the strained relationship between minorities and healthcare organisations that has existed for centuries. The roots of minority mistreatment in medicine run deep – from segregated waiting rooms to experimentation on African American patients without their permission.
The Tuskegee study conducted between 1932 and 1972 by the United States Public Health Service stands as an especially shocking testament to this notion, where medical exploitation of African American men led to terrible health outcomes for them and their families. Disclosure of the Tuskegee Study disrupted a slow convergence of black health outcomes with white health outcomes in the mid-20th century, accelerated erosion of trust in doctors, and dampened health-seeking behaviour and health-care utilization for African American men.
More historical events have fuelled this mistrust – experiments done on female slaves by Dr. James Marion Sims, who was lauded as the founder of modern gynaecology, or even Dr. Orlando Andy’s use of lobotomies as a ‘therapeutic’ treatment for institutionalised African American children during the 1960s. Dr. Sims is described as a ‘pioneer’ and lionised for his development of surgical techniques to aid childbirth, but little is said of his mistreatment of slaves and adds to the long unethical history of medical apartheid. When leaps and bounds made in medical discoveries are proven to have transpired with the aid of systematic oppression, it is not surprising that minorities would become wholly disenfranchised from modernised healthcare.
J. Marion Sims experimented on black women without anesthesia & without their consent. After much activism, a statue commemorating him will be torn down in NYC. Good riddance: https://t.co/f9eGCHPEcX #NotOurHero @BYP_100 @EH_Preservation pic.twitter.com/ihRs5RuRbV
— NARAL (@NARAL) January 13, 2018
As people of colour feel less respected by the healthcare system, their trust in medical diagnoses and treatment plans unwittingly crumbles. This impacts the patient’s health and the doctor’s ability to provide genuine medical and palliative care. Without a trusting doctor-patient relationship, African American individuals are less likely to request extra care and end up paying for it. Minority women, especially African American women, have more distrust in the medical system, which leads to delays in screening for breast cancer, new research shows.
With fewer people of colour desiring to take part in research, a lack of diversity in genomic research can limit the understanding of population-specific differences in disease risk. Also, underserved racial and ethnic groups may have difficulty accessing precision medicines that can provide improved outcomes.
The impact of unconscious bias in educational institutions, the judiciary and the workplace are well-documented. The connection between racism and healthcare have recently been exposed to public purview with the problems uncovered in algorithms. However, we must understand that the discriminatory decisions of these programs reflect those of our own, and move towards rectifying the ideals that preserve them.