Algorithms and estimating equations using laboratory data can help clinicians arrive at a diagnosis for their patient. Recently, the medical community is taking a closer look at examples of algorithms and equations that have a racial component to see if they are scientifically valid or if they lead to incorrect, delayed, or missed diagnoses.

Yesterday’s session, “Laboratory Medicine’s Role in Creating Equitable Clinical Laboratories: A Global Call to Action,” explored examples of race-based calculations and set out a roadmap for laboratorians to determine if these equations or algorithms are used in their institutions and how to discuss their utility with clinicians.

Dr. Octavia Peck-Palmer, University of Pittsburgh Medical Center, and Dr. Darshali Vyas, Massachusetts General Hospital, led a discussion of clinical cases using race-based equations by different clinical services at their institutions.

“As trainees and practitioners of medicine, there was a contradiction in the messages we were getting,” Vyas explained. “From our social science and genetic faculty, we got a clear message that race was a social construct and was not a reliable proxy for genetic difference. But as we moved to our clinical rotations into the hospitals, we saw time and again that race was being used daily as a proxy for genetic difference. The result of this is a conflict between our insights into population genetics and how we are clinically implementing race.”

Estimated glomerular filtration rate (eGFR) is the most well-known equation that incorporates a race-based factor, and there has been a call by both laboratory and nephrology professional groups to abandon the race component. Other algorithms discussed that incorporate a racial component included alpha-fetal protein (AFP) in maternal-fetal screening tests, a calculator to determine the risk of undergoing a vaginal birth after caesarean section (VBAC), and a calculated surgical risk score including age, sex, BMI, and race.

Peck-Palmer queried the attendees to share their opinions of the bias at their institutions via the audience response system and discussed how to assess their test menu and lead the charge for equitable care.

Not all the algorithms are reported directly from the laboratory. But the data that laboratorians have access to is vital to begin the discussion with their clinical colleagues about how they are using the data to treat their patients. “The fact that the laboratory generates such large databases of values means we can provide clinicians retrospective data demonstrating how biased an algorithm is,” Peck-Palmer said. “Laboratorians are really at the intersection of health, between the lab values and the clinicians.”

Laboratorians can work with clinicians and say, “let us use our data to help you help your patients,” she added.