Case Study Investigative
Determining the cardiovascular risk profile of heart failure patients and thereby predicting mortality and rehospitalisation.
This was a high impact investigation of the survival and readmissions of heart failure patients in the last 10 years at the Royal Adelaide Hospital. The predictors of readmission and survival were investigated using multivariate survival analysis with recurrent event (readmission), single event (death) and time varying covariates (changing patient covariate values with each admission). Clinical risk models were developed to classify patients into risk groups based on their baseline features and readmission features for survival and readmission. Bootstrapping methods were used to obtain comparative models (compared to normal modelling) and to determine the effect of over fitting the data. The bootstrapped models were used as clinical risk models as they are a better reflection of the general population.
Case Study Investigative
