This dissertation in operations management studies healthcare operations using large-scale data sets and econometric methods. In chapter one, we examine the benefits of step-down units (SDU), which provide an intermediate level of care for semi-critically ill patients. Using data from 10 hospitals, we estimate the clinical and operational impact of SDU care and find that SDU care significantly improves health outcomes among post-intensive care unit patients. However, its benefits are less certain for patients admitted from the emergency department. In chapter two, we utilize data from a field experiment and examine the efficacy of a connected health system that aimed to reduce readmissions through improved medication adherence. Patients in our study received electronic pill bottles that tracked medication adherence and received different types of feedback when they were non-adherent. We find that patients were more likely to become adherent when receiving high levels of intervention involving personalized feedback and when the intervention is escalated quickly and consistently. We also find that long-term adherence to two crucial heart medications reduces readmission risk. Additionally, we develop a dynamic readmission risk-scoring model and use simulation to show that, when using an intervention strategy that prioritizes high-risk patients, we can significantly reduce readmissions while using the same effort level. In chapter three, we investigate the impact of video visit availability at a large academic medical center on in-system care utilization. Through difference-in-differences, we find that video visit availability is associated with increases in overall care utilization and in-person care utilization within the system. As a whole, this dissertation provides healthcare providers and managers with data-driven insight towards how to maximize effectiveness and efficiency in healthcare delivery through changes in the care structure and health technology adoption.