In this thesis, we approach the problem of healthcare staff scheduling to propose a more efficient use of existing capacity to provide timely access in different health services. We present a set of problems related to healthcare staff scheduling. The first problem seeks to reduce the door-to-doctor time in an Emergency Department; the second problem aims to reduce the waiting time of the overall treatment also in an Emergency Department; the third problem aims to provide timely access to both clinic and surgery in a specialized surgical unit. We formulate and solve two-stage stochastic programming models that seek to accurately represent the particular features that are inherent of each problem. An important aspect in healthcare problems is a large number of uncertainties involved in the processes. The incorporation of the uncertainty increases the complexity of the problem, and it, therefore, becomes computationally infeasible to consider all of the possible scenarios. We circumvent this difficulty by relying on Sample Average Approximation (SAA) to address the demand uncertainty. We also use a discrete-event simulation (DES) model to represent the problems. Finally, we apply the framework to real case studies showing that the proposed models are adaptable to different healthcare providers. Throughout the thesis, we efficiently solve the models using real cases of Brazil and USA hospitals.