A Data Mining Framework for the Analysis of Patient Arrivals into Healthcare Centers

Abstract

We present a data mining framework that can be applied for analyzing patient arrivals into healthcare centers. The sequentially applied methods are association mining, text cloud analysis, Pareto analysis, cross-tabular analysis, and regression analysis. We applied our framework using real-world data from a one of the largest public hospitals in the U.A.E., demonstrating its applicability and possible benefits. The dataset used was eventually 110,608 rows in total for the regression models, covering the most utilized 14 hospital units. The dataset is at least 10-fold larger than datasets used in closely-related research. The developed data mining framework can provide the input for a subsequent optimization model, which can be used to optimally assign appointments for patients, based on their arrival patterns.