A team led by PC3I Innovation Faculty and Associate Director Ravi Parikh, MD, MPP, FACP, has found that machine learning-triggered behavioral nudges quadrupled the rates of conversations between patients and their clinicians about patients’ end-of-life care preferences, as well as decreased end-of-life chemotherapy by 25%.

While serious illness conversations (SICs) are an opportunity for patients to ­share their preferences and values and can lead to improved quality of life and reduce aggressive end-of-life care, most patients with advanced cancer die without a documented SIC.

Through a stepped-wedge randomized clinical trial, patients at risk for 6-month mortality were identified via a machine learning algorithm. These patients’ clinicians received text messages prompting SICs before their next encounter with the patient, in addition to weekly lists of all high-risk patients under their care. Conversation rates nearly quadrupled from 3.4% to 13.5%.

Further information on the implications of these findings can be found in a press release by Penn Medicine. The study, “Long-term Effect of Machine Learning-Triggered Behavioral Nudges on Serious Illness Conversations and End-of-Life Outcomes Among Patients with Cancer: A Randomized Clinical Trial,” was published in JAMA Oncology on January 12, 2023.