Abstract:
Operating rooms (ORs) are responsible for a major portion of revenues and costs (Gul et al. 2015). Therefore, careful management of ORs may result in substantial revenue increases and cost reductions. Parallel surgery processing has recently been implemented in some hospitals to improve patient flow process (Marjamaa et al. 2009). Parallel processing is a technique that allows concurrent implementation of anesthesia induction and turnover. This implies that the surgical process for a patient can start by giving an anesthetic agent to the patient in an induction room while the OR is being cleaned and set up for the patient.
Surgery scheduling decisions should be given carefully in ORs that work based on parallel processing principle. The impact of parallel processing on surgery sequencing and patient appointment time setting decisions must be considered in particular. Designing surgery schedules is also complicated due to uncertainty in surgery durations. Uncertainty in induction times makes the process further challenging. Furthermore, the decisions affect the trade-off between patient and provider related performance measures.
In this study, the project team proposes to study the problem of scheduling surgeries under uncertainty for multiple ORs that function based on the parallel processing principle. Decisions in our models include sequencing surgeries, setting patient appointment times, assignment of patients to induction rooms, and setting induction and surgery start times.
We formulated our surgery scheduling problem as a two-stage stochastic mixed integer program. We consider the following criteria in the objective function: expected cost of patient waiting time, OR idle time and induction room idle time. During the project period, we are planning to implement a heuristic that exploits the underlying problem structure to find good approximation for the appointment times.
We will conduct computational experiments to illustrate the performance of our solutions in comparison to the optimal solutions. The experiments will be conducted based on data from Gul et al. (2011). The experiments will reveal how optimal surgery schedules would be affected from parallel processing activities. We will compare our method with scheduling heuristics used earlier to create daily surgery schedules. We will also assess the value of implementing parallel processing based on optimal schedules. Furthermore, we will also investigate the most appropriate value of the ratio of the number of induction rooms to the number of ORs with respect to the project performance measures.