Read more

April 15, 2021
2 min read
Save

Chatbot offers real-time medication, symptom support for patients with cancer

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

Linda A. Jacobs, PhD, CRNP, FAAN, spoke at the virtual Cancer Center Survivorship Research Forum about a telehealth chatbot technology that has shown success in helping patients manage their cancer medications and symptoms at home.

“March 2020, life changed for all of us, with COVID-19 impacting delivery of medical care both inpatient and out,” Jacobs, clinical professor of nursing and director of the Cancer Survivorship Program at the Abramson Cancer Center, University of Pennsylvania, said during her presentation. “Patient care redesign was required, including a rapid need to streamline care and reduce in-person visits.”

Prior to the pandemic, the University of Pennsylvania had already begun trials using a mobile phone-based intervention nicknamed “Penny,” as well as an AI-augmented version of the chatbot for patients with cancer and cancer survivors.

Penny is a text-based conversational agent designed to interact with patients and provide symptom management and medication adherence support. The chatbot delivers real-time directions to the patient including step-by-step guidance for medications, images of current medications and how to manage low-grade symptoms. High-grade symptoms and toxicity issues are escalated to the care team.

Linda A. Jacobs, PhD, CRNP, FAAN
Linda A. Jacobs

“The chatbot processes text presented to it and responds according to a complex series of algorithms built by providers and research teams that identify and interpret what the user said,” she continued. “It can infer what they meant and determine a series of appropriate responses, such as follow-up questions if the patient mentions nausea. Penny even detected cases of incorrect dosing.”

Jacobs noted that the version with AI-augmentation was designed to learn and solve problems on its own, using machine learning from every patient interaction.

Data from a visibility study conducted between 2019 and 2020 using the AI-augmented version showed a self-reported daily adherence of 98% and a reduction in calls to practice and emergency department visits. Additionally, Penny handled 78% of participant texts with high fidelity, reliability and accurate grading of side effects during 85 combined participant months. Jacobs noted that the study was limited to 10 patients due to COVID-19 related complications.

The second ongoing study, iSMART (Implementation Strategies for improving Medication Adherence in Real Time), began in December 2020 with a goal to study algorithmic technology of the chatbot. With an expected patient pool of approximately 170 individuals with advanced non-small cell lung cancer, the researchers aim to test the effects of the chatbot’s intervention compared with standard care. Additionally, the study will look at mixed-method approaches to explore multilevel factors “shaping the acceptability, effectiveness and future implementations of Penny into routine cancer care,” Jacobs said.

“As for the next steps, we’d like to focus on survivorship follow-up care and development of new Penny modules,” Jacobs concluded. “The goal is to decrease the burden on health care providers, promote health behaviors and disease prevention, and engage patients through messaging,” which she said could be optimal for the young adult cancer survivor population.