Dynamic capacity planning of hospital resources under COVID-19 uncertainty using approximate dynamic programming

COVID-19 pandemic has resulted in an inflow of patients into the hospitals and overcrowding healthcare resources. Healthcare managers increased capacities reactively by utilizing expensive but quick methods. Instead this reactive capacity expansion approach, we propose a proactive approach considering different realizations demand uncertainties future due to COVID-19. For purpose, stochastic dynamic model is developed find right amount increase most critical hospital Due problem size, solved with Approximate Dynamic Programming. Based on data collected large tertiary Turkey, experiments show that ADP performs better than benchmark myopic heuristic. Finally, sensitivity analysis performed explore impact epidemic dynamics cost parameters results.