Black History: Race Based Algorithms In Healthcare
African Americans have a history of being disproportionately impacted during pandemics and epidemics. There are stories in American history about experiments being conducted on African Americans without their knowledge. It is not necessary to look back too far in American History history to learn about how African Americans were ignored during natural disasters and public health crisis. In society today, healthcare based Algorithms are used to to treat or not treat African Americans. The COVID-19 pandemic and the African American experience.
Pandemics and epidemics have a history of disproportionality impacting African Americans. The COVID-19 death rate for African Americans is double that of White Americans (APM Research, 2021). The HIV/AIDS epidemic has disproportionality impacted the Black community. African Americans represent 12% of the U.S. population, but account for 43% of HIV diagnoses (KFF, 2020). The crack cocaine epidemic destroyed the African American community (mass incarceration, addiction, homicide, crack babies). African Americans are ten-times more likely to be impacted by crack cocaine than White Americans (Fryer, Heaton, Levitt, Murphy, 2006). Natural disaster and the Black experience.
Natural disasters and health crises are a part of American history, but when they happen to the Black community the disparities are obvious. The natural disaster and levee breech of Hurricane Katrina killed and displaced thousands of African Americans. They were forced to live in horrifying conditions at the Super Dome until help arrived. The former first lady Barbara Bush made the statement that the conditions at the Super Dome were okay since the people were underprivileged anyway. African Americans have experienced environmental disparities. The Flint Michigan water crisis poisoned the drinking water of their 54% African American residents (US Census, 2019). The residents complained about the smell, taste and look of the water. They were ignored for almost one year (Kennedy, 2016). The United States has a history of experimenting on African Americans.
Medical experiments are believed to help advance healthcare, but that was not the case in three events in American history. Between 1845 and 1849, Dr. Sims experimented on countless enslaved women to create the sims retractor. His experiments were all done without anesthesia, and many slaves lost their lives to infection (Usslave, 2011). In 1932, the Public Health Service, along with the Tuskegee Institute, studied African American men with syphilis without their knowledge, and without treatment (CDC, n.d.). The experiment lasted for 40 years. In 1951, an African-American woman named Henrietta Lacks was diagnosed and treated for cervical cancer at Johns Hopkins University Hospital. During Henrietta’s treatment, Dr. George Gey snipped cells from her cervix without her knowledge. Henrietta’s cells have been cultured and used in experiments, they have been commercialized, and patented (NPR, 2010). Henrietta was not compensated for her historical contribution. Today, and everyday, African American are disproportionally impacted by race based algorithms in healthcare.
Algorithms are a set of rules for machines (Merriam-Webster, n.d.). Healthcare workers use these rules to make medical decisions. When race is used in healthcare it can result in diagnostic algorithms and practices which adjust the machine’s output of the patient, based on race or ethnicity (Vyas, Eisenstein and Jones, 2020). Doctors use algorithms to generate individual risk assessments and to guide their clinical decisions (Vyas, et al, 2020). These decisions can direct more attention and resources to patient A, while ignoring and withholding resources for patient B (Vyas, et al, 2020). Here are a few health related algorithms which disproportionately impact the lives of African Americans.
The Vaginal Birth after C-Section (VBAC) is an algorithm used to determine the success of a vaginal birth after Caesarean (C-Section). How it works: when the patient identifies as being African American, the chances of having a successful vaginal birth decrease by at least 14%. (Note: I tried it myself, I kept all of the answers the same and I only changed the race). African American women are less likely to be considered for a vaginal birth after the VBAC algorithm than a White woman with the same education and socio-economical status (Francom, n.d.). This is important for two reasons. One, C-Section procedures inherently riskier than a vaginal birth. Two, African American women in the United States are three times more likely to lose their life or their baby’s life while giving birth (Melillo, 2020). Check out the VBAC calculator for yourself here.
The American Heart Association, “Get with the Guidelines- Heart Failure Risk Score” is used to predict the risk of death (Vyas, et al 2020). How it works: Patients are given three points on their scorecard if they do not identify as African American. These extra points can increase the patient’s chance of being referred to a specialist (Cardiologist). The patient that identify as African American has lower chances of being referred to a specialist because they are not considered at a risk of death (Vyas, et al, 2020).
The National Football League (NFL)uses a “Race Norming Formula” to determine the severity of a player’s concussions (Start Here Episode #16, 2021). African American football players are more likely to be denied monetary relief or medical care for the same cognitive function as a White football player (Start Here, 2021). It is presumed that African American football players have a lower level of cognitive ability. How it works: According to the Psychology Research (n.d.),“Race norming is the practice of converting individual test scores to percentile or standard scores within one’s racial group” (para1)… “For example, suppose that a White candidate and a Black candidate each earn a raw score of 74 points on a test. If the White candidate’s test score is converted to a percentile only in reference to other White candidates and the Black candidate’s test score is converted to a percentile only in reference to other Black candidates, then the percentile scores earned by the two candidates may not be equal even though they attained the same raw test score (para 2)”.
Nephrology. There are algorithms and formulas used in Nephrology to determine kidney function (glomerular filtration rate) and the need for a kidney transplant (Kidney Donor Risk Index). How it works: The Glomerular filtration rate score is higher for African Americans because it is believed that African Americans release more creatinine into their blood (Vyas, et al, 2020). The higher score could result in the delay of referring African Americans to a specialist. How it works: The Kidney Donor Risk Index helps determine if a kidney graft will fail. A donor with a high risk index is an indication of failure. African American donors automatically receive a higher risk index, compared to White donors. The concern is that African Americans will have to wait longer for a kidney transplant because African American donors are matched with African American patients (Vyas, et al, 2020). Try out the calculator here.
Artificial Intelligence. A study showed that an algorithm was less likely to send African Americans patients to beneficial health programs than White patients (Ledford, 2019). How it works: Patients that identified as African American were given lower risk scores than equally sick White patients.
Obstetrics and urology also use algorithms based on race to determine care (Vyas, et al, 2020). Race is used to focus on differences, even though Race is not based on science (Smedley, 1997).
References
APM Research. (2021). The color of Coronavirus: COVID-19 Death by race and ethnicity in the U.S. Retrieved from https://www.apmresearchlab.org/covid/deaths-by-race
CDC. (n.d.).The Tuskegee Timeline Retrieved from https://www.cdc.gov/tuskegee/timeline.htm
US Census. (2019). Flint City. Retrieved from https://www.census.gov/quickfacts/flintcitymichigan
Francom, J. (n.d.). 3 Unique VBAC Challenges Women of Color Face. Retrieved from https://www.thevbaclink.com/womenofcolor/
Fryer, Heaton, Levitt, Murphy. (2006). Measuring Crack Cocaine and Its Impact. Retrieved from https://scholar.harvard.edu/files/fryer/files/fhlm_crack_cocaine_0.pdf
Kennedy, M. (2016). Lead-Laced Water In Flint: A Step-By-Step Look At The Makings Of A Crisis. Retrieved from https://www.npr.org/sections/thetwo-way/2016/04/20/465545378/lead-laced-water-in-flint-a-step-by-step-look-at-the-makings-of-a-crisis
KFF. (2020). Black American and HIV/Aids: The Basics. Retrieved from https://www.kff.org/hivaids/fact-sheet/black-americans-and-hivaids-the-basics/
Ledford, H. (2019). Millions of Black people affected by racial biases in healthcare algorithms. Retrieved from https://www.nature.com/articles/d41586-019-03228-6
Melillo, G. (2020). Racial Disparities Persist in Maternal Morbidity, Mortality and Infant Health. Retrieved from https://www.ajmc.com/view/racial-disparities-persist-in-maternal-morbidity-mortality-and-infant-health
Merriam-Webster. (n.d.) Algorithm. Retrieved from https://www.merriam-webster.com/dictionary/algorithm#note-1
NPR. (2010). ‘Henrietta Lacks’: A Donor’s Immortal Legacy. Retrieved from https://www.npr.org/2010/02/02/123232331/henrietta-lacks-a-donors-immortal-legacy
Psychology Research. (n.d.). Race Norming. Retrieved from http://psychology.iresearchnet.com/industrial-organizational-psychology/corporate-ethics/race-norming/
Smedley, A. (1997). Origin of the idea of race. retrieved from https://www.pbs.org/race/000_About/002_04-background-02-09.htm
Start Here. (2021). Episode #16 Personnel Foul: “Racial Norming” in the NFL. https://cms.megaphone.fm/channel/ESP8447327256?selected=ESP9855757396
Usslave. (2011). Retrieved from http://usslave.blogspot.com/2011/05/dr-j-marion-sims-medical-experiments-on.html
Vyas, Eisenstein and Jones. (2020). Hidden in Plain Sight-Reconsidering the Use of Race Correction in Clinical Algorithms. Retrieved from https://www.nejm.org/doi/full/10.1056/NEJMms2004740