Using artificial intelligence, Houston Methodists can predict hospitalization outcomes for geriatric dementia patients on the first or second day of hospital admission. This early assessment of outcomes means more timely intervention, better coordination of care, smarter resource allocation, targeted management, and timely treatment for these more vulnerable, high-risk patients.
Because geriatric patients with dementia stay in the hospital longer and incur higher health care costs than other patients, the team sought to address this problem by identifying modifiable risk factors and the development of an artificial intelligence a model that improves patient outcomes, improves their quality of life and reduces the risk of hospital readmissions, and reduces hospitalization costs after the model is implemented.
The study, which appeared online Sept. 29 at Alzheimer’s disease and dementia: translational research and clinical interventions, reviewed the hospital records of 8,407 geriatric patients with dementia over 10 years in the eight-hospital Houston Methodist system, identifying risk factors for poor outcomes among subgroups of patients with different types of dementia resulting from conditions such as Alzheimer’s disease, Parkinson’s disease, vascular dementia and Huntington, among others. From this data, the researchers developed a machine learning model to rapidly recognize predictive risk factors and their degree of importance for adverse hospitalization outcomes early in these patients’ hospital stay.
With 95.6% accuracy, their model outperformed all other common risk assessment methods for these different types of dementia. The researchers add that no other current method has used artificial intelligence to comprehensively predict hospitalization outcomes in elderly patients with dementia in this way, nor has it identified specific risk factors that can be modified by additional clinical procedures or precautions to reduce risk.
“The study showed that if we can identify geriatric patients with dementia as soon as they are hospitalized and recognize significant risk factors, then we can implement some appropriate interventions right away,” said Eugene S. Lai, MD, PhD. Robert W. Hervey, emeritus chief of Parkinson’s disease research and treatment at the Stanley H. Appel Division of Neurology. “By promptly mitigating and correcting modifiable risk factors for adverse outcomes, we can improve outcomes and shorten hospital stays.”
Lai, a neurologist, has worked with these patients for many years and wanted to explore ways to better understand how they manage and behave during hospitalization so that clinicians can improve their care and quality of life. He approached Stephen TK Wong, Ph.D., a bioinformatics expert and director of the Methodist TT and WF Chao Center for BRAIN in Houston, with the idea because he had previously worked with Wong and knew his team had access to a large repository of Houston Methodist patient clinical data and the ability to use AI to analyze big data.
Risk factors for each type of dementia have been identified, including those that are amenable to intervention. The upper defined result of hospitalization risk factors included encephalopathy, number of medical problems on admission, pressure ulcers, urinary tract infections, falls, source of admission, age, race, and anemia, with some overlap in the multiple dementia groups.
Ultimately, researchers seek to implement mitigation measures to guide clinical interventions to reduce these negative outcomes. Wong says the new strategy of applying powerful AI predictions to trigger the implementation of “smart” clinical pathways in hospitals is novel and will not only improve clinical outcomes and patient experience, but also reduce hospitalization costs.
“Our next step will be to implement a proven AI model into a mobile app for critical care and key hospital staff to alert them to geriatric patients with dementia who are at high risk for poor hospitalization outcomes, and guide them to interventions to reduce those risks,” said Wong, the paper’s corresponding author and the John S. Dunn Chair Emeritus of Biomedical Engineering at the Houston Methodist Research Institute. “We will work with Hospital IT to seamlessly integrate this application into EPIC as part of a system-wide implementation for routine clinical use.”
He said it would follow the same smart clinical pathway strategy they’ve been working on to integrate two new AI programs developed by his team into the EPIC system for routine clinical use to deliver interventions that reduce the risk of falls in an injured patient and better assess breast cancer risk to reduce unnecessary biopsies and overdiagnosis.
Wong and Lai’s collaborators on this study were Xin Wang, Chico F. Ezean, Lin Wang, Mamta Pupala, Yunjie He, Xiaohui Yu, Zheng Yin, and Hong Zhao, all from the TT & WF Chao Brain Center at the Houston Methodist Academic Institute. and Yang-Hsiang Huang of the Far East Memorial Hospital in Taiwan.
Xin Wang et al. Risk factors and a machine learning model for predicting hospitalization outcomes in geriatric patients with dementia, Alzheimer’s disease and dementia: translational research and clinical interventions (2022). DOI: 10.1002/trc2.12351
Citation: How Artificial Intelligence Can Help Improve Hospital Stays and Outcomes for Older Dementia Patients (September 29, 2022) Retrieved September 29, 2022, from https://medicalxpress.com/news/2022-09-ai-hospital- outcomes-older-patients. html
This document is subject to copyright. Except in good faith for the purpose of private study or research, no part may be reproduced without written permission. The content is provided for informational purposes only.