A Synthetic Data Mining Model for Evaluating Hypotension in Hemodialysis Patients

  Hypotension during Hemodialysis often increases mortality in patients undergoing dialysis for a long time. Hypotension is the most frequent adverse event during hemodialysis; therefore, the present study was conducted to investigate hypotension value of patients and present a predictive model using descriptive data mining. In this cross-sectional study, conducted from May-June 2016, the data were extracted from Ali Ibn Abi Talib hospital in Zahedan and were then analyzed using Clementine 12.0. The model was presented using K-Means, C5.0 and CART algorithms. According to the findings the parameters influencing hypotension were buffer type and blood flow the importance of which was verified through clustering and the extracted rules from the model. The use of new modelling methods to analyze dialysis data and discover the existing relationships among them, changes the attitudes of dialysis personnel towards the process of dialysis and dialysis care. The evaluation of hypotension in hemodialysis patients helps a faster and more precise identification of hypotension. It would also facilitate proper and preventive management which enhances performance in dialysis centers. The study highlighted the importance of buffer type due to its effect on hypotension