By Justice Allen
- An algorithm created by Optum was significantly underestimating the healthcare needs of Black patients
- AI in the global healthcare market is expected to reach $45.2B by 2026
It appears that technology and algorithms are not immune to the realities of racism, even as the healthcare industry increases its reliance on artificial intelligence (AI). An algorithm created by health services company Optum was intended to offer objective decision-making advice for millions of patients. Instead, it was significantly underestimating the healthcare needs of Black patients. With AI in the global healthcare market expected to grow from $4.9 billion in 2020 to $45.2 billion by 2026, how important will it be to fix the racial bias in healthcare algorithms?
Why This Matters: Optum’s algorithm missed a collective 48,722 chronic diseases in its analysis of Black patients. In an attempt to exclude race from the algorithm, it focused its decision making on potential future costs to the healthcare system. Basically, the algorithm was really predicting who would have the highest healthcare spends in the future. What the algorithm doesn’t take into consideration are the centuries of implicit healthcare bias experienced by the Black community. The Urban Institute found that disparities among African Americans, Hispanics, and non- Hispanic whites will cost the healthcare system $23.9 billion.
Disparities among African Americans, Hispanics, and non- Hispanic whites will cost the healthcare system $23.9 billion
Black patients spent about $1,800 less than white patients, and for good reason. The likes of medical malpractice, an overwhelming absence and lack of access to insurance, along with mistrust of the system has led to Black patients to spend significantly less money than their white counterparts when it comes to healthcare. Even when insurance status, income, age, co-morbid conditions, and symptom expression were equal, an analysis by the National Academy of Sciences found African Americans receive poorer quality care than white patients.
Bias occurs because the algorithm uses health costs as a proxy for health needs. In order to circumvent this from happening, the algorithm needs to be reformulated so that it no longer uses costs as a proxy for needs, and thereby eliminates the racial bias in predicting who needs extra care. Finding a solution for the problem at hand would mean that the number of “at-risk” Black patients could double in size.
What’s Next: If racial biases were to be eliminated across healthcare technology, there’s no telling what it could mean for Black generations to come. Luckily enough for Optum, the researchers who’ve discovered the problem are working to find a solution that focuses on a patient’s future health conditions rather than who will have the highest anticipated costs.
CBx Vibe: “Through The Wire” Kanye West