MIAMI – Artificial intelligence models using electronic health records and patient-reported outcomes may help identify cancer survivors at increased risk for emergency department visits, hospitalizations and worsening symptoms after treatment, according to a new study¹ from Sylvester Comprehensive Cancer Center, part of the University of Miami Miller School of Medicine. The findings suggest AI-driven risk forecasting could help care teams intervene earlier with more proactive survivorship support.

Published in JCO–Clinical Cancer Informatics, the study demonstrates how machine learning models applied to clinical data and patient‑reported outcomes, or PROs, can help identify survivors at increased risk for unplanned health care use and elevated symptom burden during the survivorship continuum. By transforming medical records and patient-reported data into predictive signals, the research offers a potential pathway toward more proactive, personalized survivorship care.

Cancer survivorship care is a dynamic, ongoing process, not a single phase of care, explained Frank J. Penedo, Ph.D., director of Sylvester’s Survivorship and Supportive Care Institute, and the study’s senior author, who led the multidisciplinary team. “For many patients, new or evolving challenges arise after treatment ends, just as routine clinical contact often tapers off, raising a critical question: how can we identify those at higher risk earlier, before these concerns intensify and become harder to address?”

Patient-reported outcomes (PROs) capture patient experiences that traditional clinical data often miss or are infrequently assessed, including emotional well-being, fatigue, functional limitations and other practical needs that may interfere with adequate survivorship care. Over the past decade, PROs have become an increasingly important component of cancer care. Yet translating large volumes of patient-reported data and integrating them with vast amounts of medical record data to enable actionable insights, particularly across survivor populations, has remained a persistent challenge.

Led by Akina Natori, M.D., MSPH, an oncologist in the Division of Medical Oncology at Sylvester, the study reframed PROs not as retrospective descriptions of patient experience, but as prospective indicators of future need.

“PROs tell us how patients are actually feeling and functioning,” said Natori, first author of the study. “We wanted to know whether those self‑reported experiences, in combination with clinical data such as cancer and treatment type, could help us identify which survivors might be at higher risk for significant symptom burden or unplanned health care use down the line.”

Unplanned health care use can include emergency department visits or hospitalizations that arise outside of scheduled follow‑up, often signaling unmet needs or gaps in survivorship and supportive care. Being able to forecast that risk could allow care teams to intervene earlier, with targeted symptom management, psychosocial support or closer monitoring.

To explore that possibility, the research team analyzed data from more than 25,000 cancer survivors followed over three years, using machine learning to detect patterns that traditional statistical methods can miss. The advantage of such approaches is their ability to weigh many factors at once – clinical history, treatments, symptoms, emotional well-being, and patterns of healthcare use – and to find the subtle interactions among them that signal which patients are heading toward trouble.

The answers depended on what was being predicted. For acute events like emergency room visits and hospitalizations, recent clinical activity was the strongest signal. What was happening with a patient in the last few months mattered more than where they started. For symptom burden, longer-term trends told a clearer story, and adding patient-reported outcomes nearly doubled how well the models performed compared with clinical data alone. When the researchers flagged the highest-risk 10% of patients, that small group accounted for roughly half of all subsequent health care events and elevated symptom episodes.

“The challenge is developing models that are not only accurate, but also interpretable and meaningful for clinicians making real‑world decisions,” said Jerry R. Bonnell, Ph.D., of the University of Miami’s Frost Institute for Data Science and Computing.

That emphasis on interpretability shaped the study’s design. Rather than treating the models as opaque systems, the team built them to show their reasoning. This surfaced which factors were driving a given patient’s risk score and how these factors shifted over time. The goal is a tool that gives clinicians not just a number, but a starting point for conversation: who needs closer follow-up, what they may need it for and when to step in before a problem escalates.

The project brought together expertise from clinical oncology, psychosocial oncology, population sciences and data science, reflecting the multifaceted nature of survivorship care. Contributors included Vasileios Stathias, Ph.D., assistant director for data science at Sylvester, as well as collaborators across the University of Miami.

“Survivorship sits at the intersection of biology, behavior and health systems,” Stathias said. “By combining patient‑reported and clinical data with advanced analytics, we can begin to see patterns that might otherwise remain invisible and that can inform more proactive care strategies.”

While the authors emphasized that the findings are not intended to immediately change clinical practice, they highlighted the broader implications of the work. As cancer survivorship populations continue to grow, health systems face increasing pressure to deliver long‑term care that is precise, proactive and sustainable.

“This is about shifting from reactive to proactive survivorship care,” Penedo said. “If we can identify patients who are more likely to struggle, we can begin to align supportive resources earlier and more effectively.”

Future research will focus on continuing to refine and validate these models across broader survivor populations, as well as exploring how electronic health records and PRO data-driven risk stratification could be integrated into survivorship standards of care.

“Our long-term goal is to ensure that survivorship care keeps pace with advances in treatment,” said Natori. “That means using data not only to describe outcomes, but to anticipate them, so we can more proactively support patients in the years after cancer.”

Read more about Sylvester’s efforts to advance cancer research through big data on the InventUM blog and follow @SylvesterCancer on X for the latest news on its research and care.

Reference:

1) https://ascopubs.org/doi/10.1200/CCI-25-00389

(Newswise/HG)