Cluster analysis and an application in healthcare

Aim: The aim of this study is to introduce cluster analysis methods and to show the application of cluster analysis methods according to health indicators of countries. Thus, the correct analysis method can be obtained by the results of the cluster analysis methods applied to the data. Material and Method: In this study, the countries will be clustered according to the variables considered as health indicators and the most recent data published on the World Health Organization and World Bank websites are used as data source. Ward's method, k-means method and two-step clustering method were used in the study. Results: Before Ward's method was applied, the data were standardized and then the method was applied. As a result of the method, two clusters were formed. Then, k-means and two-step method was applied and two clusters were observed. ANOVA results showed that all of the health indicators of the countries in the clusters formed were significantly effective (p <0.05). Conclusion: When applying hierarchical clustering methods, the standardization process of the data should be examined and the method that can eliminate this situation should be used. Otherwise, the clustering method will lead to incorrect results. Also, depending on the number of observations, variables and the data type, it should be determined which method should be applied. Based on the results of this study in which three clustering methods applied, all methods resulted with the same cluster structure. Keywords: Cluster Analysis, Distance Measure, Ward's Method, k-means method