Study on the Causal Relationship of Cardiovascular Disease Influencing Factors Based on Bayesian Causal Network

The WHO MONICA dataset was used as an example, and a logistic regression model was used to statistically analyze the data, and then a Bayesian causal network model was constructed using the MMHC hybrid algorithm to analyze the causal relationships among cardiovascular disease risk factors, and Bayesian estimation was used to learn the conditional probabilities of each node of the network so as to predict the survival of patients, and to compare the Bayesian causal network model with respect to logistic regression model in the field of chronic diseases. Bayesian causal network model results showed that hospitalization status, age at diagnosis, and angina status were direct causes of cardiovascular mortality, while previous myocardial infarction, sex, and smoking had indirect effects on cardiovascular mortality through other variables. Compared to logistic regression models, Bayesian causal networks based on the MMHC algorithm are more applicable in clinical research because they can intuitively and effectively identify and define the complex causal relationships between survival outcomes and cardiovascular disease and cardiovascular disease with each other. By analyzing this relationship, we are able to implement timely and targeted preventive and therapeutic measures and avoid possible mortality outcomes in high-risk populations.