Select Page

Algorithmic Bias in Healthcare

3

18%

of Black patients were identified as needing further care by a hospital algorithm.1

82%

of White patients were identified as needing further care by a hospital algorithm.1

46%

of Black patients should have been identified as needing further care.1

53%

of White patients should have been identified as needing further care.1

200

Million people are affected by  algorithm biases per year.2

How did this happen?

$1,800

Black patients spend $1,800 less on medical costs than White patients with the same medical conditions.

The algorithm used the medical expenditures of patients to determine who would benefit from high-risk care management. Black patients spend $1,800 less on their medical costs than White patients with the same medical conditions. Black patients also tend to pay for medical procedures when they’re at a higher level of need than their White counterparts. This difference in spending caused the algorithm to flag less Black patients  compared to White patients.³

Is this the only example?

Algorithms within healthcare systems display racial bias which have direct affects on unequal healthcare. These can be high-risk care lists, transplant lists,  surgery lists, diagnosis standards, and much more.4 Even algorithms outside of healthcare can create racial bias in healthcare systems. hiring algorithms have been shown to have racial bias, which leads to unequal pay. This can lead to racial gaps in the affordability of good healthcare in the United States.5

NO.