Doctorate – Appointment Systems in Outpatient Healthcare Services: Using Heterogeneity for Performance Gain
Advisor: Prof. Dr. Marcelo Caldeira Pedroso
Comission: Profs. Drs. Jorge Luiz Biazzi, Marco Aurélio de Mesquita and Júlio Cesar Bastos de Figueiredo
Appointment systems are the connection between efficiency and quality in outpatient health services. Maintaining competitive costs is challenging as resources become increasingly scarce. At the same time, the context of growing demand, driven by demographic aspects and technological advances, imposes complexity for the quality of care. Several appointment systems in the literature seek this balance and the heterogeneity of patients has been applied in their models with the objective of gaining performance. In order to obtain this gain, however, appointment time restrictions are imposed on patients, thus reducing their flexibility of choice, an important component of the quality of care. Composed of three interrelated articles, this study applied heterogeneity in outpatient appointment systems with the objective of gaining performance without losing flexibility in choosing appointment times. Based on gaps identified in the first article, which consisted of a systematic review of the literature on the topic, two appointment systems were developed, for which discrete event simulations were conducted and their performance was measured. For the system of the second article, modeled as sequential, a heuristic was developed to recalculate the remaining appointment times for each appointment request, starting from an initial schedule built with models extracted from the literature. These recalculations were based on the patients' no-show rate, this being the heterogeneity factor adopted. There was a performance gain in terms of total cost (TC) varying between 0.46% and 5.94%, among the 18 environments simulated, with the lowest costs being obtained in scenarios with lower cost ratio of doctor’s time to patients’ time (CR), as well as lower coefficients of variation in service times (Cv). It was also found that the proposed heuristic is more efficient when patients with a higher no-show rate predominate in the clinic session. In the model of the third article, also designed as sequential, heterogeneity was characterized by different no-show probabilities associated with periods of the clinic session. Applying appointment rules developed for the heterogeneity of patients identified in the literature, an improvement in performance was observed in 57 of the 72 scenarios, with an average gain in TC ranging from 0% to 9.54%. It was observed that this gain grows with the increase in the difference between the no-show rates of the session periods. Among the limitations, this study was restricted to only a portion of the wide range of possible combinations of environmental factors. It can also be highlighted the use of only one variable, the probability of no-show, as a factor of heterogeneity. As for future studies, this work points out directions for improving the heuristics addressed. It is also recommended that studies use the heterogeneity associated with periods of the clinic session to build the initial schedule in the system with the heuristic for recalculations, thus combining both proposed systems in a single model.
*Abstract provided by the author