MedStar Health Awarded $3 Million Grant to Study How Artificial Intelligence Technology Like Amazon Alexa and ChatGPT Can Reduce Health Disparities in Post-Operative Recovery

The NIH-funded research will examine the effectiveness of using artificial intelligence and natural language processing to improve outcomes after GI cancer surgery

COLUMBIA, Md. – MedStar Health has received a $3 million dollar grant from the National Institute on Minority Health and Health Disparities (NIMHD) to study how a voice-assisted, remote patient symptom monitoring system (VARSMS) developed with artificial intelligence and natural language processing technology can improve outcomes for patients as they recover from gastrointestinal (GI) cancer surgery. Designed to recognize accents and a variety of speaking cadences, and provide real-time, tailored responses, VARSMS aims to improve post-surgical recovery by proactively encouraging patients to follow discharge recommendations and improving communication with their healthcare providers.

Remote patient symptom monitoring may be a powerful tool to prevent post-discharge complications, particularly among patient populations who historically experience worse outcomes. Black and Hispanic patients are twice as likely as White patients to experience complications after GI cancer surgery, and earlier research indicates that these disparities may be driven by adverse events when patients are recovering at home, potentially under difficult socioeconomic conditions. These patients with the most risk also may face additional barriers to follow-up care including poor provider communication or low health literacy. The project will use artificial intelligence tools like BERT, a language representation model from Google, and ChatGPT to enhance comprehension of patient feedback, particularly focusing on the linguistic habits of diverse minority groups.

“Ensuring open and timely communication between patients and providers is paramount in addressing disparities in health outcomes. We firmly believe that leveraging artificial intelligence can effectively tackle the challenges faced by our most vulnerable patients,” said Nawar Shara, Ph.D., director, Center for Biostatistics, Informatics, and Data Science (CBIDS), MedStar Health Research Institute and Principal Investigator for the study. This study builds on her research examining how voice-assisted devices like Amazon Alexa can improve care management for ethnoracial minority patients with heart failure. “Our pilot study showed that AI technology can be used successfully to improve patient engagement. We hope to continue to see promising results examining its use in the immediate postoperative period and in the future, explore how these tools can be integrated to improve the patient-provider relationship throughout the continuum of care.”

The study will include a randomized clinical trial with patients enrolled from high-risk populations from urban, suburban, and rural settings. Patients who have had surgery for GI cancer will be assigned either the standard of care that includes traditional follow-up communication with their provider or the VARSMS system for a 40-day period following discharge. The research team will then analyze how using the VARSMS system impacts outcomes and adherence to treatment protocols, as well as evaluate the surgical team’s experience.

“Ethno-racial minorities have complex social conditions and significant underlying multi-morbidity. Consistent examples of these challenges among these vulnerable populations have under- or no insurance, experience difficulties in equitable and quality access to cancer care and are one of most vulnerable populations with access to optimal and rigorous digital access across America. Together, these social factors place them at higher risk for postoperative complication and even death,” said Waddah Al-Refaie, MD, FACS, Chair of Department of Surgery at Creighton University and Chair of Surgery at CHI Health, and co-Principal Investigator for the $3M grant. “The ability to collaborate with Creighton University and CHI Health presents a unique opportunity to also examine another subset of under-studied population, which is rural dwellers with surgical cancers. The COVID-19 pandemic has worsened disparities among ethno-racial minorities and rural populations, and this will examine whether AI and ChatGPT can help reduce those inequities.”

“Dr. Shara and Dr. Al-Refaie have a long history of productive collaboration in research that uses data science to achieve health equity, and it’s exciting that this work will continue that partnership as both of our organizations seek to advance the health of our communities,” said Neil J. Weissman, MD, FACC, FASE, chief scientific officer at MedStar Health and president of MedStar Health Research Institute.

The cross-disciplinary research team also involves other experts in human-centered design and evaluation, including Guodong (Gordon) Gao, Ph.D., and Ritu Agarwal, Ph.D., both professors at Johns Hopkins Carey School of Business and Co-Directors of the Center for Digital Health and Artifical Intelligence (CDHAI), Dakuo Wang, Ph.D., associate professor, College of Arts, Media, and Design, and Khoury College of Computer Sciences at Northeastern University.

Research reported in this publication was supported by the National Institute on Minority Health and Health Disparities of the National Institutes of Health under Award Number R01MD018424. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Learn more about this work and more at MedStarResearch.org.

Learn more about Creighton University at Creighton.edu

Learn more about CHI Health at CHIhealth.com