Proposed Method for Identifying Emergency Unit Profiles from the Monthly Service Number

Abstract

The indicators are used to quantify how the results of the public emergency departments (ED) can be classified, with regard to the efficiency and quality of the services provided to its users. In this sense, the objective of this work is to identify typical patterns of the curves of the indicator number of monthly attendance of the procedure “Reception with Risk Classification” in a sample of 50 EDs. For this, a quantitative and exploratory research was carried out that adopted the Machine Learning technique known as Cluster Analysis not supervised by the AGNES “AGlomerative NESting” method. 10 profiles (groups) of curves for the Reception with Risk Classification were identified in the selected EDs. Group 1 characterizes the standard profile of these EDs (76% of the total) and the other groups characterize the atypical patterns. The results were obtained using the free software R.