This woman had to campaign to find a liver donor. Transplant docs say there should be a better way

It's pretty unusual in the medical system for a patient to be tasked with finding their own cure. But that’s what happened to Stephanie Azzarello in 2023 when she was told she needed a liver transplant or she would die. 

Azzarello has primary sclerosing cholangitis, a rare chronic liver disease that causes damage to the bile ducts and the liver.

A decade after she was diagnosed, there was so much damage to her liver she went on the waiting list to get a liver from a deceased donor. But the Toronto woman was told it was very unlikely she’d ever get an organ that way. 

Instead, she needed to find someone who was willing to undergo a major surgery to donate part  of their liver to her. No one in her family or close circle of friends was a match, so she was forced to go public. 

“I mean, I'm not asking for directions. I'm asking for a vital organ,” recalls Azzarello, now 41. “It was terrifying, but I knew I didn't have a choice because I was going to die on that list.”

Dr. Mamatha Bhat is trying to change the system that puts Azarello and so many other patients in that position. She’s a clinician scientist at the University Health Network Ajmera Transplant Centre in Toronto. It’s the biggest transplant hospital in North America. 

Patients are prioritized based on the Model for End Stage Liver Disease, also known as the MELD score. This method takes the results of a few blood tests to determine how urgently someone needs a liver. 

But Bhat noticed the model put women and people with rare conditions like PSC at a disadvantage. Even though they could be very sick, the results from the blood tests used for the MELD score didn’t put them high on the waiting list.

For example, one indicator of liver dysfunction is a high level of creatinine in the blood. But the MELD score isn’t adjusted for the fact that female bodies have less muscle mass and therefore lower levels of creatinine to begin with, says Bhat. Likewise, men are more likely to die suddenly in circumstances that lead to organ donation, and some male livers are a poor fit for smaller female bodies.  

“Each patient is a bit unique," she said. A linear statistical model like MELD that only considers a few blood tests doesn’t reflect that complexity, Bhat says.

So starting in 2023 Bhat and her team got funding from the Canadian Institutes for Health Research to develop an artificial intelligence system that would evaluate many different factors and eventually replace the MELD score. 

“You can really bring together multiple parameters, like blood test results, historical changes in those blood test results, changes in the clinical condition of the patient over time,” said Bhat. 

She says early results show the AI system can do a better job of prioritizing patients based on need. Bhat says she hopes it will be used in hospitals across the country in a year or two and cut down the mortality rate among people waiting.

Data from the Canadian Institute for Health Information shows there were 655 liver transplants in Canada in 2024. Another 609 remained on the waitlist during that time, and 89 people died while waiting.