Two top Democrats are putting pressure on the Trump administration and big-name health insurers to address concerns about racial bias in algorithms used to assess patients’ healthcare needs.
Sens. Cory Booker, D-New Jersey, and Ron Wyden, D-Oregon, sent letters to the Centers for Medicare & Medicaid Services (CMS), CVS Health, the Blue Cross Blue Shield Association, Cigna, Humana and UnitedHealth Group Tuesday requesting additional information on if and how these groups may be tackling bias in these algorithms.
“Both the people who design these complex systems and the massive sets of data that are used have many historical and human biases built-in,” the senators wrote. “Without very careful consideration, the algorithms they subsequently create can further perpetuate those very biases.”
In late October, Optum, a subsidiary of UnitedHealth Group, came under fire after it was revealed that an algorithm it built for hospitals to identify patients with chronic disease may have a racial bias.
A study published in the journal Science found that the algorithm would assign the same rank to sicker black patients and healthier white patients, which could prioritize care for white patients. New York insurance regulators are also probing the algorithm.
Industry experts note, though, that an algorithm like Optum’s is just one element of the discussion, and its impact is also highly dependent on how providers are applying these tools to make care decisions.
In the letters, Booker and Wyden ask whether the private insurers deploy diverse teams to develop algorithms to assist in care decisions and if they’re routinely auditing these tools to prevent bias. They also ask CMS whether they require health plans and hospitals to submit data on algorithms they may use.
They want answers by Dec. 31.
The pair of senators also requested that the Federal Trade Commission (FTC) commit to delving deeper into how these data analytics tools may be discriminating against minority patients.
The FTC previously held hearings on consumer protections regarding the use of predictive analytics and artificial intelligence in 2018.
“Congress and the administration must make a concerted effort to find out the degree to which this issue is widespread and move quickly to make lasting changes,” Booker and Wyden wrote.