Can algorithms predict frailty?
Data access has been approved for a novel research study which will create and test algorithms to predict frailty using administrative health data and primary care electronic medical record data.
Frailty, a medical syndrome with numerous causes, is characterized by reduced strength, endurance and physiological function which results in increased vulnerability to functional decline, dependence and death. Early identification of frailty can decrease end-of-life costs while increasing quality of life. Frailty can be an early way to triage patients, and an accurate way to assess the degree of frailty can be a prompt for care planning between primary care clinicians and their patient and their patient’s family.
“Independent of age, frailty is predictive of adverse health events including hospitalizations, institutionalization, falls, declining health and death,” says project leader Professor Sabrina Wong at the University of British Columbia’s School of Nursing and Centre for Health Services and Policy Research. “These adverse health events translate to increasing costs for the healthcare system. While costs and frailty may be positively related because those who are frail need care, it may be that costs, particularly of hospital care, are rising because the healthcare system is not serving this population well. Those who are frail could be falling through the cracks or receiving far more aggressive treatment than they may want.”
Currently primary care clinicians usually identify and work with those who are frail. However, more work is needed to consistently and accurately detect frailty, or those at risk of becoming frail, both at a clinical practice and at a population level. Identifying those who are frail in primary care, as well as in communities, could enable targeted communications with patients and families and community-based resources in order to improve patient care, patients’ and caregivers’ quality of life and better use of the healthcare system.
This innovative study will use administrative health data and primary care electronic medical record data to create algorithms to identify frailty in those aged 65 years and older living in the community. The predictive ability of the algorithms will be tested using evidence provided by primary care clinicians and patients.
The results of this study will inform both clinical care and jurisdictional level health services planning. The goal of this work is to support healthy aging and the needs of older adults.
PopData will link data from the BC Ministry of Health and the BC Vital Statistics Agency with the BC Canadian Primary Care Sentinel Surveillance Network (CPCSSN) data for the project. This same linkage has already taken place in the Manitoba Health Policy Centre using the Manitoba CPCSSN and administrative health data files. Investigators on this project are from BC, Alberta and Manitoba.
The project is funded by the Canadian Institutes of Health Research and the Michael Smith Foundation for Health Research.