Machine learning model can predict survival in dialysis patients undergoing continuous renal replacement therapy

Investigation result

A UCLA-led team has developed a machine learning model that can accurately predict short-term survival in dialysis patients undergoing continuous renal replacement therapy (CRRT).

background

CRRT is a treatment used for seriously ill hospitalized patients whose health conditions prevent them from undergoing regular hemodialysis. It is a gentler treatment that provides continuous treatment over a long period of time. However, about half of adults who undergo CRRT do not survive, making the treatment a waste for both the patient and their family.

Method

To help doctors decide whether patients should start CRRT, the researchers developed a machine learning model that uses data from electronic medical records of thousands of patients to predict the chances of survival after treatment.

Impact

The findings provide a data-driven tool to support clinical decision-making that incorporates advanced machine learning techniques to analyze large, complex sets of patient data that have previously been challenging for physicians. The study shows how integrating machine learning models into healthcare can improve treatment outcomes and resource management.

comment

“CRRT is often used as a last resort, but many patients do not survive, leading to wasted resources and false hope for families,” said Dr. Ira Kurtz, chief of UCLA Nephrology and senior author of the study. “By being able to predict which patients will benefit, this model serves as the basis for testing its utility in future clinical trials, with the aim of improving patient outcomes and resource use. As with all machine learning models, this one needs to be tested in the real world to determine whether its predictions are similarly accurate for patients for whom it has not been trained.”

author

Additional authors are Davina Zamanzadeh, Jeffrey Feng, Panayiotis Petousis, Arvind Vepa, Majid Sarrafzadeh and Alex Bui of UCLA, and S. Ananth Karumanchi of Cedars Sinai Medical Center in Los Angeles.

journal

The study has been published in a peer-reviewed journal. Nature Communications.

Funding

This study was supported by funding from the Nephrology, Urology, and Hematology Advanced Research Training Program (NIH NIDDK U2C DK129496), a Medical Imaging Informatics Training Grant (NIH NIBIB T32 EB016640), a UCLA CTSI grant (NIH UL1 TR001881), the Smidt Family Foundation, the Factor Family Foundation, the Davita Allen Nissenson Research Fund, the Ralph Block Family Fund, and the Kleeman Family Fund.