Three-Way Multiblock Methods for Decision-Making in Health Services: A COSTATIS Approach to Improving Management at a Private Clinical Laboratory

Abstract

In the last decade, the clinical laboratory sector in Portugal has witnessed an increase in competition through the integration of small private laboratories in multinational organizations. In this competition scenario, it is imperative that decisions related to management are made based on the analysis of varied and detailed information data in order to propose strategies and prioritize management. Thus, the growing need for data analysis, especially in dynamic contexts, to detect, identify and characterize the most relevant issues for a more efficient decision-making process, is essential. This research was based on data for the period 2014-2017, from a private clinical laboratory located in the Algarve. The use of two interrelated information structures, namely 14 collection points and the subsequent 10 clinical tests with the highest billing volume in 2017, led to the application of the COSTATIS method (Thioulouse et al., 2011) in order to capture a common co-structure that analyses the stability or instability of the two data structures. An exploratory three-way multiblock method highlights the relations between two data structures as a whole. Therefore, the interpretation of the influence relationships of the production of clinical tests on the performance of collection points can reveal particularly useful discrepancies and/or coincidences for the management. With this detailed information, it is possible to propose initiatives for improvement or correction and to promote functional strategies with the objective of assisting the laboratory management process and suggesting strategies that are more efficient.