Algorithm Accurately Predicts Heart Failure Survival

The algorithm will allow doctors to carry out more personalized assessments of people awaiting heart transplants, thus enabling health care providers to efficiently use limited life-saving resources and reduce health care costs.

The algorithm, named Trees of Predictors, uses machine learning and takes into consideration 53 data points, including age, gender, body mass index, blood type and blood chemistry, to address the differences between people awaiting heart transplants and the compatibility between potential heart transplant recipients and donors. Using these data points, the algorithm predicts how long people with heart failure will live, depending upon whether they will receive a transplant or not. The algorithm can also analyze different possible risk scenarios for potential transplant candidates in order to assist doctors to more thoroughly assess people who can be candidates for heart transplants, and is flexible enough to incorporate more data as treatments evolve.

The UCLA researchers tested the algorithm using 30 years of data on people registered with the United Network for Organ Sharing, a non-profit organization that matches donors and transplant recipients in the US. The researchers found the algorithm provided significantly better predictions than the prediction models currently used by most doctors to project, which transplant recipients would live for at least three years after a transplant. The UCLA algorithm outperformed the models by 14% by correctly predicting that 2,442 more heart transplant recipients of the 17,441 who had received transplants and lived at least that long after the surgery. According to the researchers, the Trees of Predictors algorithm can also be used to gather insights from medical databases and various other types of complex databases.

“Our work suggests that more lives could be saved with the application of this new machine-learning–based algorithm,” said Mihaela van der Schaar, Chancellor’s Professor of Electrical and Computer Engineering at the UCLA Samueli School of Engineering, who led the study. “It would be especially useful for determining which patients need heart transplants most urgently and which patients are good candidates for bridge therapies such as implanted mechanical-assist devices.”

“Following this method, we are able to identify a significant number of patients who are good transplant candidates but were not identified as such by traditional approaches,” said Dr. Martin Cadeiras, a cardiologist at the David Geffen School of Medicine at UCLA. “This methodology better resembles the human thinking process by allowing multiple alternative solutions for the same problem but taking into consideration the variability of each individual.”

Related Links:
UCLA