Predicting Stroke in Hemodialysis Patients Using Data Mining

Data analysis has a fundamental role in decision making process and one effective way to do this is through data mining. In the field of health, by performing data mining on patients with acute conditions such as hemodialysis (HD), their future trends can be partially understood which helps to reduce the severity of their complications. Given that recent studies indicate an increased risk of stroke in HD patients, the aim of this study is to investigate a conventional surgical procedure in hemodialysis and its relationship with stroke. The method of the present study is the use of supervised data mining algorithms and with the focus on the "decision tree", which is one of the applied algorithms in the classification of variables and helps to identify the influential factors. The input data included 468 of hemodialysis patients who were studied over a five-year period and consisted of 324 females and 144 males. This study showed that the risk of stroke in patients whose vascular access surgery was performed by catheterization before fistula was 84.21%, while there was no significant relationship between the age of dialysis patients and their stroke, but in addition to catheter implantation, history of blood pressure or diabetes were also linked to stroke.