Research Will Employ Data Science to Determine Suicide Risk in Elderly

Oct. 24, 2018
Research Will Employ Data Science to Determine Suicide Risk in Elderly

 A new research program will harness machine learning and data science to sift through tens of millions of records of U.S. nursing home and assisted living residents to identify risk factors for suicide. The project will be led by Yue Li, Ph.D., in the University of Rochester Medical Center (URMC) Department of Public Health Sciences and Xueya Cai, Ph.D., in the Department of Biostatistics and Computational Biology and is supported by a $1.6 million grant from the National Institute of Mental Health.

Suicide is among the 10 leading causes of death in the U.S. and suicide risk increases substantially after age 65. As approximately 70 percent of individuals over the age of 65 will require long-term care, research indicates that suicide deaths are concentrated in nursing homes and assisted living facilities.

Residents of long-term care facilities are often socially isolated, physically and cognitively disabled, and diagnosed with multiple mental and medical conditions, all of which are associated with suicide in older adults. As a result, it is speculated that nursing home residents have an elevated risk of taking their own lives compared to older adults who live in the community, despite the fact that long-term care facilities are capable of closely monitoring the daily activities and health of their residents.

The new research will examine tens of millions of nursing home and assisted living records from the Centers for Disease Control and Prevention (National Violent Death Reporting System, National Death Index) and the Centers for Medicare and Medicaid Services.

Medical Center faculty will collaborate with researchers in the University of Rochester Goergen Institute for Data Sciences to employ advanced machine learning applications that will sift through the immense sets of data and build algorithms that can identify risk and predictive factors – such as depression, physical conditions, status of social and family connections, and cognitive function – for suicide deaths and attempts. 

The researchers will also attempt to identify characteristics of nursing homes, such as the safety culture, staffing levels, and quality of care, that contribute to elevated suicide risk. The information developed by the project will serve to inform potential future organizational and policy interventions to prevent suicide among elders receiving long-term care.

Additional investigators for the project include Helena Temkin-Greener, Ph.D. with the URMC Department of Public Health Sciences; Yeates Conwell, M.D. with the URMC Department of Psychiatry; and Sheryl Zimmerman, M.S.W., Ph.D. with the University of North Carolina at Chapel Hill.