Nursing home staffing is a perennial issue, even absent disasters such as hurricanes and pandemics that strike the most vulnerable residents and health care workers with particular ferocity.
That is why since 2018, a USF-led team has been working to create a high-tech tool to help provide nursing homes with a way to predict and plan for staffing. Now amid the devastation of COVID-19, their work has taken on heightened importance.
USF College of Engineering Assistant Professor Mingyang Li along with School of Aging Studies Professor Kathryn Hyer and Associate Professor Hongdao Meng are leading the interdisciplinary team working to develop predictive analytics models and tools that would assist nursing home operators in making proactive staffing decisions that meet the diverse needs of nursing home residents.
The $286,000 project was funded by the National Science Foundation’s Grant Opportunities for Academic Liaison with Industry (GOALI) effort, which supports projects that bring academic researchers and industry together to solve problems. Greystone Healthcare Management Corporation, which operates dozens of health care centers in Florida, Illinois and Missouri, is the private-sector partner in the project.
The effort to bring predictive analytics to nursing home staffing issues under crisis scenarios was initially inspired by the nursing home care crises amid 2017’s Hurricane Irma.
School of Aging Studies Professor Kathryn Hyer and USF College of Engineering Assistant Professor Mingyang Li.
NSF recently funded an additional supplemental project for the USF team to examine the impact of the pandemic on nursing home staffing and extend decision support platform to crisis scenarios.
“We want to be able to predict an individual’s absenteeism risk by fusing multi-source data and developing predictive analytics tools,” Li said. “Absenteeism is a critical issue and the outbreak of COVID-19 exacerbated that situation.
“Many are concerned with lack of medical supplies and may exhibit no-shows. The remaining caregivers have a very high workload and are prone to burnout,” he added.
“We want to really investigate these no-shows due to COVID-19 and to further develop a decision support toolkit to help inform nursing home administrators proactively make staffing adjustment decisions that will reduce the service operation disruption associated with nursing home absenteeism.”
COVID-19 has been unlike any disaster modern nursing homes have ever seen.
Over the last three months, more than 40,600 long-term care residents and workers have died of COVID-19 – about 40% of the nation’s death toll attributed to the coronavirus, according to an analysis of state data gathered by USA TODAY and reported in early June.
In the latest reports from federal authorities, there have been more than 95,000 cases of COVID-19, the disease caused by the coronavirus, among nursing home residents, according to the federal Centers for Medicare & Medicaid Services.
In Florida, there have been more than 6,400 cases of COVID-19 in long-term care facilities and more than 1,500 resident deaths. Long-term care facilities account for about half of Florida’s deaths in the pandemic, according to data compiled by the Kaiser Family Foundation, a leading think tank on national health issues.
Nearly 2,400 nursing home workers also tested positive for the virus, and 10 workers have died from COVID-19 by mid-June, according to Florida Department Health data compiled by the Florida COVID-19 Hub at the USF Libraries GIS Unit of the Digital Heritage and Humanities Collections.
The quality of elder care in nursing homes is directly tied to adequate staffing, but also to the level of skill among the health care professionals working there. In times of disaster, disruptions to that careful balance can turn dangerous, USF’s experts in aging policy have found.
“All of these workers face stress and challenges of COVID19. Few of these workers have access to the personal protective equipment ─ gowns, gloves, masks that are needed to protect themselves when working,” Hyer said. “Many care for children or older adults in their own home creating concerns about bringing the virus back to their families.”
The project takes on the problem on two fronts. First, multi-source time-sensitive data related to nursing home staff absenteeism, such as survey data of caregivers, nursing home electronic health records and operational staffing data, will be collected and integrated to improve understanding the many facets of staff absenteeism at both a macro and micro level, Li said.
Predictive analytics will fuse and analyze the data to predict absenteeism risk of caregivers. The staffing decision system will be able to provide nursing home a view of staffing both under normal operating conditions and extreme scenarios.
Given that insight, nursing home operators will be able to plan for staffing shortages by bringing in staffing agencies, budgeting for inventive pay or other measures that would ensure residents receive the care they need. That insight also will help to address the industry’s looming staffing crisis, the team said.
During the last decade, the need for direct care workers has almost doubled from 2.9 million in 2008 to nearly 4.5 million, Hyer said. Over the next 10 years, the long-term care workforce is expected to add 1.3 million jobs and an additional 6.9 million jobs will become vacant as existing workers leave the field or exit the labor force.
The industry, though, leans on a vulnerable population to provide the needed labor. Most of these positions are filled by women, particularly women of color and immigrant women who earn wages of about $10-$12 an hour, Hyer said.
“Our work has the potential to address one of the most serious issues facing the country ─ the shortage of trained, caring direct care workers,” Hyer said.
Learn more about Dr. Li’s Data Science and System Informatics laboratory at http://www.eng.usf.edu/~mingyangli/
Learn more about the USF School of Aging’s Florida Policy Center on Aging at https://www.usf.edu/cbcs/aging-studies/fpeca/
Visit the Florida COVID-19 Hub at https://covid19-usflibrary.hub.arcgis.com