New York state officials launched an investigation into whether Optum’s algorithm used by hospitals to identify patients with chronic diseases has a racial bias.
New York's Financial Services and Health departments sent a letter to UnitedHealth Group’s CEO David Wichmann Friday regarding an algorithm developed by Optum, The Wall Street Journal reported. The investigation is in response to a study published in the journal Science that found an algorithm used to assign risk scores gave the same risk level to blacks as white patients that were healthier.
“These discriminatory results, whether intentional or not, are unacceptable and are unlawful in New York,” according to the letter.
The study said that the algorithm reduces the number of black patients that need extra care by half.
“Bias occurs because the algorithm uses health costs as a proxy for health needs,” the study said. “Less money is spent on black patients who have the same level of need, and the algorithm thus falsely concludes that black patients are healthier than equally sick white patients.”
The algorithm needs to be changed so it doesn’t use cost as proxy for healthcare need and would eliminate the racial bias in predicting who needs extra care, the authors added.
“By relying on historic spending to triage and diagnose current patients, your algorithm appears to inherently prioritize white patients who have had greater access to healthcare than black patients,” the letter said.
New York’s regulators called for Optum to stop using the algorithm or provide evidence to contradict the study’s claims.
UnitedHealth Group said in a statement on the study's release that it appreciates the “researchers’ work and validation that the cost model was highly predictive of cost.”
But the algorithm is “just one of many data elements intended to be used to select patients for clinical engagement programs, including most importantly, the doctor’s expertise and knowledge of his or her patient’s individual needs.”
Industry analyst Cynthia Burghard, who has conducted research on information technology and reviewed the Science study, said, "The big question is, is it the tool or how the tool was used?"
Burghard, research director for value-based healthcare IT transformation strategies at IDC Health Insights, said the study looked at one academic medical center’s use of the algorithm. Industry best practices for using this type of algorithm to dictate who gets care management often involves using multiple pieces of information, not just health costs.
"There is a lot of nuance to people’s health that is not reflected in cost," she said, noting that the Optum algorithm is not artificial intelligence but a rules-based system.
Using algorithms to look at a cohort of patients is a common practice in healthcare, with companies like IBM offering these technology tools, Burghard noted. If hospitals and health systems stopped using these tools it would be "throwing the baby out with the bathwater," she said.
She added the study highlights the importance of involving clinicians in technology purchasing decisions when the technology will be used to support clinical decisions.
Optum, the provider services subsidiary, has been a major revenue driver for UnitedHealth.
The group posted $5 billion in profits in the third quarter of 2019 driven largely by strong performance by Optum, which posted double-digit growth for both revenue and earnings in the quarter.
Heather Landi contributed additional reporting to this story.