The peer-reviewed journal article features research into machine learning-powered digital tools to help identify cancer patients' likelihood of making unscheduled emergency department visits
SAN MATEO, CA / ACCESSWIRE / October 3, 2023 / Ronin, an innovator of evidence-based clinical decision support, announced today the publication of an original research paper in Communications Medicine, a journal from Nature Portfolio. The paper details research to develop a robust, reliable, and explainable model that predicts emergency department (ED) admissions for cancer patients within a 30-day timeline. Titled, "Ensuring fair, safe, and interpretable artificial intelligence-based prediction tools in a real-world oncological setting," this paper was co-authored by Ronin associates and experts from The University of Texas, MD Anderson Cancer Center. A related blog post, "Driving Healthcare Innovation: A Three-Pronged Approach to Guarantee Safety and Equity of AI" was also featured.
Individuals diagnosed with cancer often experience side effects from their treatments, leading to frequent visits to hospital emergency departments. The ability to foresee these visits can revolutionize the administration of care for vulnerable cancer patients. The published paper documents the development of a digital tool designed to identify cancer patients likely to require an unscheduled visit to the ED within 30 days.
To develop and evaluate this model, the research team utilized health records from more than 28,000 patients who had received care at a single oncology hospital in the United States. The model demonstrated impressive performance and consistent results across various demographic sectors and disease categories. Researchers monitored the model's performance and established that it remained stable over time. The methodology employed in observing the model's performance could prove invaluable for deploying comparable predictive models in clinical settings, ensuring their reliable operation.
Christine Swisher, Ph.D., chief scientific officer at Ronin and one of the paper's authors, emphasized the potential impact of this technology, stating, "Approximately 50% of cancer patient ED visits and 19% of hospitalizations may be avoidable, but each cancer patient's journey is unique. By harnessing the power of AI, this technology can help alert care teams to symptoms and adverse events which could lead to emergency room visits or worsening conditions -- empowering clinicians to manage such events before they become severe or require emergency care."
"Ensuring fair, safe, and interpretable artificial intelligence-based prediction tools in a real-world oncological setting," was authored by Renee George, Benjamin Ellis, Andrew West, Alex Graff, Stephen Weaver, Michelle Abramowski, Katelin Brown, Lauren Kerr, and Christine Swisher (also of The Lawrence J. Ellison Institute for Transformative Medicine) of Ronin, and Sheng-Chieh Lu and Chris Sidey-Gibbons of the Section of Patient-Centered Analytic, Division of Internal Medicine, The University of Texas MD Anderson Cancer Center. It can be accessed on the Communications Medicine website.
Communications Medicine is a selective open-access journal from Nature Portfolio that publishes high-quality research, reviews, and commentary across all clinical, translational, and public health research fields. The journal aims to foster collaboration across these communities to facilitate discovery that will promote health and improve patients' lives. Primary research papers published by the journal represent significant advances in preventing, diagnosing, or treating human diseases relevant to a specialized field.
About Ronin
Ronin is on a mission to improve cancer care. Every clinical decision should be founded on data, personalized for each patient, and made with confidence. Our groundbreaking cancer intelligence platform helps clinicians make better-informed treatment decisions, predict and preempt adverse patient events, and effectively engage and support patients in their care journey. Learn more about how we are improving patient outcomes, clinician satisfaction, and operational efficiency at projectronin.com.
###
MEDIA CONTACT:
Dave Anderson
Anderson Interactive (for Ronin)
[email protected]
SOURCE: Project Ronin