Medical science continues to use long-held racial stereotypes about Black people to diagnose illnesses. But using a race-neutral approach provides more accurate health assessments, according to a new study published Thursday (June 1) in JAMA Network Open.

Researchers compared results from lung function diagnostic software that uses a race-based algorithm to results from software that used a race-neutral algorithm.

They found that using the race-neutral algorithm could have helped to diagnose up to 40 percent more breathing problems in Black male patients than the current diagnostic software. The research looked at test results from more than 2,700 Black men and 5,700 white men.

Dr. Darshali Vyas, a pulmonary specialist, told The Associated Press that medical researchers long suspected that integrating race-based assumptions into diagnostic software produced inaccurate results. Now there’s evidence.

One of the long-held racist assumptions is that Black people’s lungs were innately worse than those of white people. Modern guidelines adopted that unscientific bias, which was incorporated into computer algorithms that adjust for the patient’s race.

Typically, results of pulmonary function tests are an important factor in diagnosing lung problems, along with X-rays, lab work and family histories.

Vyas said race-based algorithms have already been changed in some diagnostic software, including one for pregnant women that predicts risks of vaginal delivery if the mom previously had a cesarean section.