Common healthcare algorithm biased, reduces care for black patients by more than half, study finds

A commonly used healthcare algorithm exhibits significant racial bias and reduces the number of black patients identified for extra care by more than half, according to a study published Oct. 24 in American Association for the Advancement of Science. 

Researchers analyzed 2013-15 data from primary care patients with risk-based contracts at a hospital that used an algorithm sold by Optum of United Health Group, according to The Wall Street Journal. The sample consisted of 6,079 black patients and 43,539 white patients, with almost three-fourths enrolled in commercial insurance and over a quarter in Medicare.  

Of the patients needing more care, as identified by the algorithm, 81.8 percent were white and about 18 percent were black. Black patients were significantly sicker than white patients, researchers say, and if the algorithm accurately reflected the proportion of sick patients, 46 percent of black patients should have been identified.

Researchers believe the algorithm's use of health costs as a proxy for health needs creates the bias. Less money is spent on black patients who have the same needs as white patients, so the algorithm inaccurately concludes that black patients are healthier.

The algorithm manufacturer replicated the analyses on its national dataset of nearly 3.7 million commercially insured patients and found black patients had 48,772 more chronic conditions than white patients, confirming the racial bias.  

The study authors recommend a new algorithm that no longer uses cost to predict who needs extra care.

Copyright © 2022 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.


Featured Learning Opportunities

Featured Whitepapers

Featured Webinars