Machine learning helps improve end-of-life care, communication for people with cancer
A machine learning algorithm that triggered electronic reminders to health care providers increased the rate at which providers and patients spoke about end-of-life care preferences, according to randomized trial results.
The algorithm — which predicts mortality risk — prompted “electronic nudges” that resulted in a fourfold increase in serious illness conversations, as well as reduced use of systemic therapy at end of life, researchers wrote in JAMA Oncology.
The findings suggest machine learning, along with the behavioral nudges — can result in “long-lasting improvements in cancer care delivery,” Ravi B. Parikh, MD, medical oncologist and assistant professor of medical ethics and health policy at Perelman School of Medicine at University of Pennsylvania, as well as associate director of Penn Center for Cancer Care Innovation, and colleagues wrote.
Healio spoke with Parikh about the findings and their potential implications.
Collapse