Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Industrial Engineering

Committee Chair/Advisor

Dr. Tugce Isik

Committee Member

Dr. Sandra Eksioglu

Committee Member

Dr. Kevin Taaffe

Committee Member

Dr. Amin Khademi


One of the critical challenges in healthcare operations management is to efficiently utilize the expensive resources needed while maintaining the quality of care provided. Simulation and optimization methods can be effectively used to provide better healthcare services. This can be achieved by developing models to minimize patient waiting times, minimize healthcare supply chain and logistics costs, and maximize access. In this proposal, we study some of the important problems in healthcare operations management. More specifically, we focus on perioperative services and study scheduling of operating rooms (ORs) and management of necessary resources such as staff, equipment, and surgical instruments. We develop optimization and simulation methods to coordinate material handling decisions, inventory management, and OR scheduling.

In Chapter 1 of this dissertation, we investigate material handling services to improve the flow of surgical materials in hospitals. The ORs require timely supply of surgical materials such as surgical instruments, linen, and other additional equipment required to perform the surgeries. The availability of surgical instruments at the right location is crucial to both patient safety and cost reduction in hospitals. Similarly, soiled material must also be disposed of appropriately and quickly. Hospitals use automated material handling systems to perform these daily tasks, minimize workforce requirements, reduce risk of contamination, and reduce workplace injuries. Most of the literature related to AGV systems focuses on improving their performance in manufacturing settings. In the last 20 years, several articles have addressed issues relevant to healthcare systems. This literature mainly focuses on improving the design and management of AGV systems to handle the specific challenges faced in hospitals, such as interactions with patients, staff, and elevators; adhering to safety standards and hygiene, etc.

In Chapter 1, we focus on optimizing the delivery of surgical instrument case carts from material departments to ORs through automated guided vehicles (AGV). We propose a framework that integrates data analysis with system simulation and optimization. We test the performance of the proposed framework through a case study developed using data from a partnering hospital, Greenville Memorial Hospital (GMH) in South Carolina. Through an extensive set of simulation experiments, we investigate whether performance measures, such as travel time and task completion time, improve after a redesign of AGV pathways. We also study the impact of fleet size on these performance measures and use simulation-optimization to evaluate the performance of the system for different fleet sizes. A pilot study was conducted at GMH to validate the results of our analysis. We further evaluated different policies for scheduling the material handling activities to assess their impact on delays and the level of inventory required. Reducing the inventory level of an instrument may negatively impact the flexibility in scheduling surgeries, cause delays, and therefore, reduce the service level provided. On the other hand, increasing inventory levels may not necessarily eliminate the delays since some delays occur because of inefficiencies in the material handling processes. Hospitals tend to maintain large inventories to ensure that the required instruments are available for scheduled surgery. Typically, the inventory level of surgical instruments is determined by the total number of surgeries scheduled in a day, the daily schedule of surgeries that use the same instrument, the processing capacity of the central sterile storage division (CSSD), and the schedule of material handling activities. Using simulation-optimization tools, we demonstrate that integrating decisions of material handling activities with inventory management has the potential to reduce the cost of the system.

In Chapter 2 we focus on coordinating OR scheduling decisions with efficient management of surgical instruments. Hospitals pay more attention to OR scheduling. This is because a large portion of hospitals' income is due to surgical procedures. Inventory management of decisions follows the OR schedules. Previous work points to the cost savings and benefits of optimizing the OR scheduling process. However, based on our review of the literature, only a few articles discuss the inclusion of instrument inventory-related decisions in OR schedules. Surgical instruments are classified as (1) owned by the hospital and (2) borrowed from other hospitals or vendors. Borrowed instruments incur rental costs that can be up to 12-25\% of the listed price of the surgical instrument. A daily schedule of ORs determines how many rental instruments would be required to perform all surgeries in a timely manner. A simple strategy used in most hospitals is to first schedule the ORs, followed by determining the instrument assignments. However, such a strategy may result in low utilization of surgical instruments owned by hospitals. Furthermore, creating an OR schedule that efficiently uses available surgical instruments is a challenging problem. The problem becomes even more challenging in the presence of material handling delays, stochastic demand, and uncertain surgery duration. In this study, we propose an alternative scheduling strategy in which the OR scheduling and inventory management decisions are coordinated. More specifically, we propose a mixed-integer programming model that integrates instrument assignment decisions with OR scheduling to minimize costs. This model determines how many ORs to open, determines the schedule of ORs, and also identifies the instrument assignments for each surgery. If the level of instrument inventory cannot meet the surgical requirements, our model allows instruments to be rented at a higher cost. We introduce and evaluate the solution methods for this problem. We propose a Lagrangean decomposition-based heuristic, which is an iterative procedure. This heuristic separates the scheduling problem from the inventory assignment problem. These subproblems are computationally easier to solve and provide a lower bound on the optimal cost of the integrated OR scheduling problem. The solution of the scheduling subproblem is used to generate feasible solutions in every iteration. We propose two alternatives to find feasible solutions to our problem. These alternatives provide an upper bound on the cost of the integrated scheduling problem. We conducted a thorough sensitivity analysis to evaluate the impact of different parameters, such as the length of the scheduling horizon, the number of ORs that can be used in parallel, the number of surgeries, and various cost parameters on the running time and quality of the solution. Using a case study developed at GMH, we demonstrate that integrating OR scheduling decisions with inventory management has the potential to reduce the cost of the system.

The objective of Chapter 3 is to develop quick and efficient algorithms to solve the integrated OR scheduling and inventory management problem, and generate optimal/near-optimal solutions that increase the efficiency of GMH operations. In Chapter 2, we introduced the integrated OR scheduling problem which is a combinatorial optimization problem. As such, the problem is challenging to solve. We faced these challenges when trying to solve the problem directly using the Gurobi solver. The solutions obtained via construction heuristics were much farther from optimality while the Lagrangean decomposition-based heuristics take several hours to find good solutions for large-sized problems. In addition, those methods are iterative procedures and computationally expensive. These challenges have motivated the development of metaheuristics to solve OR scheduling problems, which have been shown to be very effective in solving other combinatorial problems in general and scheduling problems in particular. In Chapter 3, we adopt a metaheuristic, Tabu search, which is a versatile heuristic that is used to solve many different types of scheduling problems. We propose an improved construction heuristic to generate an initial solution. This heuristic identifies the number if ORs to be used and then the assignment of surgeries to ORs. In the second step, this heuristic identifies instrument-surgery assignments based on a first-come, first-serve basis. The proposed Tabu search method improves upon this initial solution. To explore different areas of the feasible region, we propose three neighborhoods that are searched one after the other. For each neighborhood, we create a preferred attribute candidate list which contains solutions that have attributes of good solutions. The solutions on this list are evaluated first before examining other solutions in the neighborhood. The solutions obtained with Tabu search are compared with the lower and upper bounds obtained in Chapter \ref{Ch2}. Using a case study developed at GMH, we demonstrate that high-quality solutions can be obtained by using very little computational time.

Author ORCID Identifier




To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.