Echocardiography Appointment Scheduling Through Better Utilization of Resources

Background and Objectives: Appointment scheduling systems are applied in a broad variety of healthcare environments to reduce costs, increase resource utilization, and facilitate patients’ access to care. This study strives to present efficient scheduling models for the Echocardiography Department of Tehran Heart Center (THC). These models seek to optimize both patient and hospital utility by maximizing the weighted number of performed echos and minimizing overtime. Methods: There are two major problems in developing such models: shift scheduling problem and capacity allocation problem. In this paper, two mixed integer linear programming (MILP) models are presented based on two different sets of assumptions. The first model is developed according to the current routines of the hospital. In this model, it is assumed that the assignment of specialists to echocardiography laboratories in different shifts is predetermined. Thus this model merely allocates the available capacity of specialists and labs to different types of patients. However, the second model is more comprehensive, as it schedules the shifts of the specialists and allocates the capacity to the patients simultaneously. Findings: The efficiency of the proposed models is evaluated using the real data of the Echocardiography Department of THC. The results showed that both models increased the utility (12.35% and 19.14%, respectively) in comparison with the current status of the department. The first model improved the performance of the department significantly through better utilization of resources; however, the second model improved the performance much more than the first one through creating more capacity and utilizing the capacity efficiently. Conclusions: Although both models showed significant improvements, the second model was found to be more efficient. The reason is that the first model assumes the specialists’ shift assignment to be predetermined, while the second model finds the best shift assignment itself.