Financial profiling of public hospitals: an application by data mining

Abstract

This paper presents an application of the data mining method to determine the financial profiles of the public hospitals in Turkey. The study is based on the data compiled in 2004, covering 645 public hospitals run by the Ministry of Health (MoH) as the main provider of primary and secondary health services in Turkey. The public hospitals, currently financed by a mixture of funds allocated from the general budget and individually operated revolving funds, need urgent solutions to their financial problems as a part of an ongoing national reform effort. The analysis adopts the Chi-Square Automatic Interaction Detector (CHAID) decision tree algorithm, as one of the most efficient and up-to-date data mining method used for segmentation. The study has found that the public hospitals could be categorized by the CHAID into 12 different profiles in terms of their financial performance. These profiles have guided us in determining the key financial indicators to be focused upon in the public hospitals and present best practices to improve their respective financial performances. The findings have also allowed policy suggestions as to the financial strategies that may be considered in improving the financial performance of the public hospitals toward a successful health sector reform in Turkey.