Improving Mortality Prediction in People with Cardiovascular Disease: A Random Survival Forests Approach Integrating Frailty Assessment
Description: Introduction: Frailty refers to a condition of increased vulnerability due to the accumulation of various health issues. Individuals with cardiovascular disease (heart problems) often have high levels of frailty and resultantly have worse outcomes. A common method of assessing frailty sums up health problem into a single score called the frailty index. This study examines a more refined approach using Random Survival Forests, a statistical method, to predict mortality by analyzing each health problem separately. Methods: The study included data from 2918 individuals with a history of cardiovascular disease. Frailty was assessed using 46 different health criteria. Two models were trained to predict mortality: one considering each health criterion individually, and the other using the composite score. The process included determining which specific health issues were most predictive of mortality. Results: The study involved individuals with an average age of 66.4, 41.9% of whom were female. The model using individual health criteria was found to be more accurate in predicting mortality compared to the composite score model. Specifically, variations in blood pressure were found to be highly influential in predicting both general and heart-related mortality. Conclusion: Analyzing individual health criteria provided a more accurate mortality prediction for individuals with cardiovascular disease. This approach reveals the potential to optimize predictions by understanding complex relationships between different health issues. By identifying the key factors contributing to mortality, more targeted interventions for frailty can be developed. The study points towards a more precise and individualized method for assessing and managing the health of individuals with heart problems.
Amount: CA$1,500
Funding type: fellowship
Funder scheme: CIHR-ICRH Travel Awards - General pool
Start date: 2023-09-01
End date: 2024-08-31
Investigator: Hoang Quach
Institution: Dalhousie University (Nova Scotia)
Topic: Frailty in Older Adults
Subfield: Geriatrics and Gerontology
Field: Medicine
Domain: Health Sciences
Funded outputs count: 0
Funder award ID: 191127_1
OpenAlex ID: G1886211384