Two University of Florida College of Nursing researchers have been awarded a five-year $2.57 million grant for a project that will use registered nurses’ observation notes to prevent death and injury of hospitalized older adults who are at a higher risk for hospital-acquired falls and hospital-induced delirium.
Through the grant from the National Institute on Aging at the National Institutes of Health, Associate Professor Robert Lucero, Ph.D., M.P.H., R.N., FAAN, and assistant professor Ragnhildur Bjarnadottir, Ph.D., M.P.H., R.N., seek to develop a model that will identify and predict mortality and injury related to iatrogenic conditions, or conditions induced by medical treatment, by evaluating and processing electronic data from patient health and administrative records at UF Health in Gainesville and Jacksonville.
Iatrogenic conditions currently account for the death of 250,000 older adults annually in the United States. Among the most common iatrogenic conditions are hospital-acquired falls and hospital-induced delirium, which have been found to be linked, and account for an estimated direct medical cost of $200 billion per year, according to findings in the Journal of the American Geriatrics Society.
Electronic clinical data were captured for about 80% of the U.S. population and about 75% to 80% of this data is text, or unstructured data, including registered nurses’ progress notes. These data can be a rich source of information because RNs provide narrative observations and assessments of patients and their environment that are typically captured and recorded at least once a shift. The process of analyzing narrative text data is complex but is critical to identifying patterns and attributes associated with patient delirium and falls.
Through this project, Lucero, Bjarnadottir and researchers and clinicians from across UF Health, the university’s academic health center, will develop the UF EHR Clinical Data Infrastructure for Enhanced Patient Safety among the Elderly, or UF-ECLIPSE. They will be analyzing data that were collected over six years from more than 130,000 patients and more than 2 million progress notes. As experts in nursing informatics, text-mining and health services, Lucero and Bjarnadottir will be among the first to implement and test an integrated data repository that uses RN-generated notes to support an academic health center. The outcomes of this project will serve as a model for health care organizations nationwide to increase safe, effective care for the millions of older adult Americans hospitalized every day.
“This NIH award demonstrates the strength of nursing and health informatics at the University of Florida,” said Anna M. McDaniel, Ph.D., R.N., FAAN, UF College of Nursing dean and the Linda Harman Aiken Professor. “It will also continue to chip away at a historically siloed approach to understanding complex clinical problems by having interdisciplinary teams tackling and identifying solutions through nursing science, clinical informatics and data science.”
Lucero said this research is significant because hospital-acquired falls and other iatrogenic conditions continue to be a threat to the lives of hospitalized patients. His interest in hospital-acquired falls suffered by older adults, in particular, originated with the loss of a family member following a fall.
“This project gives us the opportunity to validate findings related to nurses’ assessments, clinic procedures and treatments and nurse staffing. We also think it will demonstrate the critical role of registered nurses, specifically their surveillance role to ensure the safety of the health care system because nurses’ notes capture more than just their interactions with patients. These notes capture observations about the environment and the nurses’ professional judgment.”
The occurrence of an iatrogenic condition is impacted by a patient’s mental status and/or medications; environmental factors, such as unfamiliarity of surroundings or clutter that could pose as a trip hazard; and clinical/organizational interventions, which can either be preventive or increase risk. To process nurses’ progress notes to predict hospital-acquired falls, the researchers will create a text-mining pipeline using algorithms and analytical approaches to identify word clusters or patterns in the data that include these factors. Prior to this project, there was very little research on processing and analyzing these types of data.
“The potential impact of this project is substantial,” Bjarnadottir said. “There is so much that we still do not know about how to predict and prevent iatrogenic conditions, but this project can support countless data-driven interdisciplinary aging studies and become a piece of the data science infrastructure necessary to fuel a learning health system. This is, therefore, a significant building block in our quest to find ways to reduce and perhaps eventually eliminate iatrogenic conditions among the growing number of hospitalized older adults in the U.S. and around the world.”
Another goal of the UF-ECLIPSE project is to determine and evaluate the structural and human resources to support sustained interdisciplinary aging studies. Hardware and software will be installed in the UF Health integrated data repository to ensure functionality and durability of text-mining, as well as its integration and evaluation.
The UF-ECLIPSE interdisciplinary team includes academic and clinical experts in the fields of patient safety, care quality, health outcomes, nursing and health informatics, natural language processing, data science, aging, standardized terminology, clinical decision support, statistics, machine learning and hospital operations. In addition to the researchers, a UF Health nurse-led iatrogenic condition task force in Gainesville and Jacksonville will provide expert knowledge in translating the use of the electronic clinical data. Multidisciplinary graduate research assistants will also be trained to support data-driven aging research and advance the science of aging, which is a strategic direction of the National Institute on Aging.
“UF-ECLIPSE is a significant project not only because of the impact it will have on the health and safety of hospitalized older adults by enhancing the learning health system at UF Health, but also the significance for interdisciplinary research,” said David Nelson, M.D., senior vice president for health affairs at UF and president of UF Health. “I am especially excited for the teamwork taking place between our Gainesville and Jacksonville campuses and the involvement of UF Health’s clinical nursing leadership and staff.”