Comorbidity Trajectory Network in Heart Failure Patients Based on Prefixspan Algorithm

Due to the evolution and improvement of the medicare quality in the world and the improvement of healthcare system in Taiwan over the past few decades, the ageing population is increasing and the global population is getting older. In addition, the westernization of lifestyle and diet has also made people suffer from chronic diseases. Although many chronic diseases including diabetes, hypertension, hyperlipidemia, heart disease and stroke patients can be improved and controlled because of the advanced medical technology, a variety of chronic diseases will lead to heart failure over many years. Currently, the number of heart failure patients has increased significantly. The main reasons, as stated above, are the aging of the population and the advancement of cardiac catheterization technology. Many patients with acute myocardial infarction or coronary artery occlusion disease that might have died in the past can survive because of appropriate first-aid measures and medical treatment present today, but they will still suffer with heart failure in the end. Therefore, the plausibility of heart failure disease has increased. In the future, the elderly population will become the focus of healthcare and heart failure patients will become a burden on healthcare. Therefore, to prevent the progression from chronic diseases to heart failure is an important medical care issue. One of the important and effective methods to reduce the medical burden is to utilize national health insurance to analyze, predict and prevent the occurrence of heart failure and strengthen the prognostic medical care mechanism by using large amounts of health insurance data. In this study, developing new algorithms to improve computing performance and efficiency and achieves two purposes. First is to effectively assist the government in implementing the principles of preventive medicine with three sub-specialty areas and five levels and the other one is to strengthen precise prediction and preventive medicine replacement of the current medical mechanism of disease diagnosis and treatment. Analyze the medical records of heart failure patients in Taiwan and figure out the distribution of disease time courses, common diseases patterns, and explore the odds ratio of different disease trajectories. It is hoped that healthy people can activate the heart failure warning mechanism earlier by providing sequential disease combinations and the relative time course of heart failure.