Frailty risk detection from primary care electronic medical records

Older adults with frailty are at high-risk of declines in their health and do not bounce back as well as their non-frail counterparts. These patients are among the highest users of health care because they end up in crises and go to emergency, are hospitalized, or die. But, if frailty is detected and managed earlier, before negative events occur, it can improve patient outcomes and decrease health system costs. Currently, frailty detection relies on tools that take added time and resources that family doctors do not have. This project will develop and test an automated tool that will detect these high-risk patients from existing electronic medical records using natural language processing and machine learning techniques. This frailty risk detection tool will help doctors and patients action plan together to reduce their risk, avoid negative events, and remain healthier in their communities.

Intern: 
Na Zhang
Faculty Supervisor: 
Linglong Kong
Province: 
Alberta
Partner University: 
Program: