Performance measurement of Brazil Federal University hospitals through principal component analysis: a public health care services’ panorama

Purpose This paper aims to contextualize the process of public hospital providing services, based on the measurement of the performance of Federal University Hospitals (HUFs) of Brazil, using the technique of multivariate statistics of principal component analysis. Design/methodology/approach This research presented a descriptive and quantitative character, as well as exploratory purpose and followed the inductive logic, being empirically structured in two stages, that is, the application of principal component analysis (PCA) in four healthcare performance dimensions; subsequently, the full reapplication of principal component analysis in the most highly correlated variables, in module, with the first three main components (PC1, PC2 and PC3). Findings From the principal component analysis, considering mainly the component I, with twice the explanatory power of the second (PC2) and third components (PC3), it was possible to evidence the efficient or inefficient behavior of the HUFs evaluated, through the production of medical residency, by specialty area. Finally, it was observed the formation of two groups composed of seven and eight hospitals, that is, Groups II and IV, which shows that these groups reflect similarities, with respect to the scores and importance of the variables for both hospitals’ groups. Research limitations/implications Among the main limitations, it was observed incomplete data for some HUFs, which made it impossible to search for information to explain and better contextualize certain aspects. More specifically, a limited number of hospitals with complete information was dealt with for 60% of SIMEC/REHUF performance indicators. Practical implications The use of PCA multivariate technique was of great contribution to the contextualization of the performance and productivity of homogeneous and autonomous units, represented by the hospitals. It was possible to generate a high quantity of information, to contribute with assumptions to complement the decision-making processes in these organizations. Social implications Development of public policies, with emphasis on hospitals linked to teaching centers represented by university hospitals. The projection of improvements in the reach of the efficiency of the services of assistance to the public health, from the qualified formation of professionals, both to academy, as to clinical practice. Originality/value The originality of this paper for the scenarios, Brazilian public health sector and academic area involved the application of a consolidated performance analysis technique, that is, PCA, obtaining a rich work in relation to the extensive exploitation of techniques to support decision-making processes. In addition, the sequence and the way in which the content, formed by object of study and techniques, has been organized, generating a particular scenario for the measurement of performance in hospital organizations.