1064 — Predicting Adverse Outcomes in Heart Failure Patients Using Different Frailty Status Measures
Lead/Presenter: Yan Cheng,
All Authors: Cheng Y (George Washington University)
Shao Y (George Washington University)
Weir CR (University of Utah)
Shah RU (University of Utah)
Bray BE (University of Utah)
Garvin JH (University of Utah)
Zeng-Treitler Q (Washington DC VA Medical Center)
Frailty commonly occurs in older adults and is an important determinant of health outcomes. However, frailty measurements are rarely collected in a quantitative, reliable fashion in routine patient care. Our prior work has examined the association between the number of frailty topics and adverse outcomes. This study is to use existing frailty assessment instruments for ontological guidance, and to create four different measures of frailty status to evaluate their role in predictive modeling.
We used the Veterans Administration Informatics and Computing Infrastructure (VINCI) as the data source, and randomly sampled 12,000 veterans with heart failure diagnosed in 2010. The topic modeling method was applied to identify frailty-related topics from the clinical notes in the electronic medical records. The frailty topics were classified into five deficit areas including physical functioning (PF), role-physical (RP), general health (GH), social functioning (SF), and mental health (MH). We used logistic regression models to analyze the association of frailty and the outcome at the individual level and experimented with different covariates and four different frailty measures: individual frailty topics, number of distinct frailty topics, a dichotomous deficit category, and the number of distinct deficits, respectively.
A total of 8,531 (71.1%) patients had at least one frailty topic. The prevalence of GH, PF, MH, SF, and RP deficits were 89.0%, 61.3%, 56.9%, 40.6%, and 9.5%, respectively. We created four sets of predictive models with each set of models using the same covariates and one of the four frailty measures as predictors, respectively. For each set of models, when the same covariates were used, the accuracy were very similar across the models, with AUC around 0.66, 0.80, 0.81, and 0.86 for set #1, #2, #3 and #4 of models, respectively. The number of distinct deficits was almost consistently positively associated with the outcome and the association was always significant. Among individual deficit variables, physical functioning was the only one deficit that was consistently associated with increased risk of outcome.
Aggregate frailty deficit measurements created based on frailty topics and ontology knowledge significantly and strongly predicted adverse outcome among heart failure patients.
Frailty ontology knowledge based on frailty topics identified using natural language processing may help advance VA research and improve health and outcomes of Veterans.