Stories

Civic Science Fellow Jyoti Madhusoodanan investigates harmful biases in healthcare algorithms 

While interviewing a scientist for a freelance story in 2018, science journalist Jyoti Madhusoodanan heard something unexpected: an offhand mention of a clinical calculation known as the estimated glomerular filtration rate equation, or eGFR, which is used widely to assess kidney function, included a “race adjustment” for Black patients.

“I grew up in India and only began thinking of myself as Asian once I moved to the U.S., an experience that really brought home for me what we mean when we say race is a social construct,” Madhusoodanan says. “So the idea that it could be tied to biology was baffling from the start.”

Intrigued, she asked where to find a list of these race-adjusted equations.

“There isn’t a list,” the researcher replied.

“I remember thinking, ‘How can there be no list?’” she says. “I immediately started compiling my own.”

As a seasoned science journalist whose reporting has appeared widely in publications including Scientific American, Nature, and The New York Times, Madhusoodanan is no stranger to this kind of detective work. Throughout her career, she has been drawn to stories about how science impacts individuals and communities, particularly those that have been traditionally underserved by both medicine and journalism.

This approach makes her a natural fit for civic science journalism, which seeks to not just expose problems, but to illuminate solutions and connect scientists with the diverse populations affected by the products of research. Madhusoodanan’s questions about race-based medical algorithms took exactly this kind of community-centered approach.

That work eventually led her to the Doris Duke Racial Equity in Clinical Equations Civic Science Fellowship at the Association of Health Care Journalists. When she learned about the fellowship, she realized she’d long been doing work that aims to forge the kind of bridges civic science is built on—just without the framework to name it. 

“I’d been asking these questions for years,” she says. “The fellowship gave me the tools and community to pursue them systematically.”

Bringing an invisible problem to light

Until recently, race-corrected clinical algorithms were only known to small circles of experts and advocates working to rectify the problem. In June 2020, a paper in the New England Journal of Medicine published a list of these calculations, bringing them out into broader scientific view for the first time.

“Seeing that list laid out felt validating—a real spark,” Madhusoodanan says. “That’s the piece I’d been waiting for.”

With renewed clarity, she dove deeper into how and why scientists include race in their algorithms. “We all live in what biological anthropologist Alan Goodman beautifully called ‘racial smog,’” Madhusoodanan says, referring to the pervasive cultural mindset that classifies individuals based on arbitrary race categories. “Like breathing polluted air, we take it in constantly, without noticing. Scientists, researchers, clinicians—they’re all breathing that smog, too.”

But the practice has real, if unintended, impacts, Madhusoodanan says. Researchers initially incorporated racial corrections to improve the statistical accuracy of their equations based on limited datasets, not realizing that doing so could profoundly disadvantage entire communities. For example, the eGFR adjustment effectively made Black patients with the same kidney function appear to be less sick than white patients, causing them to begin treatment at a later stage of chronic kidney disease and wait longer on average for kidney transplants.

“These statistical fixes ended up perpetuating inequities, rather than correcting them,” she says.

In April 2024 Madhusoodanan began her Civic Science Fellowship at the Association of Health Care Journalists, the leading professional association for reporters in this field. There she created “Algorithms and Health Disparities,” a reporting series and resource site that helps journalists spot hidden bias, explain terms, and come up with illuminating angles for their stories. The project also translates dense research on pulse-oximeter failures into webinars with real-world impact. All told, she wrote two dozen clear, actionable stories—stories that prompt readers to look twice at their own lab results.

“You might not necessarily think of yourself as a patient who needs to advocate for yourself,” Madhusoodanan says. “You could just be a healthy person in your thirties going for an annual physical and talking to your doctor. But suddenly you might have more awareness, more understanding of what’s in your test results.”

That jump from data to personal insight is already happening: “An editor once told me after reading my story that his wife looked at the lab report from her annual physical and found that her eGFR had a race correction on it,” she says. “These aren’t abstract problems. They’re immediately relevant to people’s real lives.”

Toward structural solutions

Patient awareness is only the first step. For deeper change, Madhusoodanan turns her lens on the structures that produced these biased tools in the first place.

“Journalists are uniquely positioned,” she says. “Often, these biases are invisible until you’re personally affected. But if journalists understand them early, we have a chance to shift public narratives for good.”

This collaborative approach reflects Madhusoodanan’s own scientific training, having earned a Ph.D. in microbiology before becoming a journalist. That background shapes how she views the researchers behind problematic practices: not as villains, but as committed scientists working within flawed systems.

“It’s not helpful to paint scientists creating these algorithms as the bad guys,” she says. “They’re products of the same system, trying sincerely to improve health outcomes.”

Since Madhusoodanan first heard of eGFR’s race adjustments, major changes have already occurred. Kidney scientists have revised equations to eliminate race-based calculations; the same reform happened for vaginal birth calculators and pediatric infection algorithms.

Yet progress remains challenging partly because “there’s no real incentive for people to go back and fix these outdated algorithms and equations,” Madhusoodanan says. “There’s no big payout at the end. They do this purely because they really want to make things better.” 

Often the corrections originated simply because the initial clinical studies excluded diverse populations, she says. “Inclusion needs to start at fundamental research stages, before these biases take root.”

Her hope is that informed journalists will help drive this systemic shift in representation. “We can’t get there unless we keep reporting, connect the dots, and keep illuminating this issue,” she says.

In the end, Madhusoodanan says, “it isn’t about blaming anyone. It’s about clearing the smog of societal assumptions and illuminating the questions we haven’t yet asked.”

Curiosity and community

Working remotely from her home in Portland, Oregon, Madhusoodanan found the weekly virtual meetings alongside other Civic Science Fellows a bit intimidating at first.

“You’re in this virtual room feeling overwhelmed, surrounded by really smart people doing important work,” she recalls. “But the Civic Science folks were intentional and lovely about helping everyone overcome the feeling that everyone else here has it figured out, and eventually you realize everyone felt that exact same uncertainty. That shared realization was very reassuring.”

Asked what advice she’d offer emerging journalists or scientists eager to tackle big societal problems like racial inequity, Madhusoodanan emphasizes humility, curiosity, and openness, traits she says were deeply reinforced by her Fellowship.

“I think about it like parallel play from when my daughter was a toddler,” she says. “We each sit with our own set of blocks, building separately, but close enough to inspire one another.” Scientists, clinicians, journalists—each plays best by building parallel yet intersecting paths.

Along those same lines, Madhusoodanan believes journalism doesn’t have to be adversarial to serve truth. “You can ask hard questions with compassion,” she says. “You can build community instead of walls.”

Jyoti is a member of the 2024-25 Civic Science Fellows cohort. Her Fellowship was supported by the Doris Duke Foundation.