Integrating computer simulation and the normalized normal constraint method to plan a temporary hospital for COVID-19 patients

Although Temporary Hospitals (THs) usually present improvised structures, they are considered necessary solutions in extreme situations, such as critical moments of the COVID-19 pandemic. Timely planning of hospital resources is a challenging task for public health decision-makers and essential to caring for patients to ensure the healthcare system's sustainability. For this reason, we propose the integration of Discrete Event Simulation (DES), Response Surface Methodology (RSM), and Normalized Normal Constraint (NNC) for planning a TH in Brazil amid the COVID-19 pandemic. While DES is an outstanding process modelling approach, RSM and NNC are methods for planning and analysing experiments and generating a representative set of optimal solutions for decision-makers. This study defined the number of beds in the Short Stay Unit (SSU) that maximizes the number of patients attended and the patient conformity rate while minimizing the SSU daily cost. Moreover, two uncontrollable input factors are also considered: the number of patients arriving at the TH and the percentage of these patients who have COVID-19. We developed complete and evenly distributed Pareto frontiers considering controllable and uncontrollable factors, which decision-makers can use according to the metrics prioritization and the pandemic status.