Essays on the interface of health care and operations management: Human judgement and quality

This thesis studies an important aspect of healthcare operations: discretion in task completion due to human judgment, and task discretion's impact on quality. The goal of this thesis is to provide frameworks for modeling this aspect, and understanding its effect on costs and revenue in healthcare operations management. To accomplish this goal, we formulate and analyze two problems in a stochastic setting and a deterministic setting, respectively.;In the first essay, we analyze a stochastic model motivated by a specific type of diagnostic service center. In response to sky-rocketing costs, the current US health care system is moving toward helping patients seek greater engagement in health care choices. This is because there exists a strong belief in practice that helping patients select appropriate services can reduce extra costs incurred when patients present themselves for inappropriate treatments. This raises important questions about managing a special type of diagnostic service center that combines expert opinion, information, and high-quality decision counseling to help patients make more informed, thus more cost-effective treatment decisions. In these service centers, customers are strategic, i.e., they have autonomy to decide whether to use the service center or not, based on their expectation of diagnostic accuracy and the waiting time. Since longer service (greater service depth) typically entails higher accuracy, but also more congestion, the manager must decide on the optimal service depth to guide a diagnostic process between an agent and customer. We model this human decision process (diagnostic process) in detail as a sequential testing process and incorporate strategic customer behavior within a queueing system. This is the first theoretical model studying the tradeoff between quality (accuracy of the process) and the time required to achieve this accuracy, in the presence of strategic customers. Our results have important practical implications: We provide a framework toward factoring congestion into decisions to achieve high quality of care, and we derive insights for managing diagnostic service centers in health care. The results not only confirm previous findings in the literature: Why increasing skill level may actually increase congestion; but also generate new insights such as why it may not be optimal for the manager to capture more demand in certain circumstances, and the role of the copayment between the patient and the service provider in maximizing the potential of the service center.;In the second and the third essays, we explore this tradeoff of quality and efficiency in a deterministic setting in which a given set of tasks are correlated by precedence relationships and constrained by resource capacity. This situation is typically faced by a project manager in the process of generating knowledge products, such as news reporting, intelligence gathering in addition to health care. In general, the quality obtained by performing a given task in such a system will increase with how long it is executed, and the overall quality of the process depends on how effectively time and resources are allocated to various tasks. The decisions are activity start times and activity durations that maximize overall output quality, given the constraints about activity release dates, activity precedence relationships and the project deadline.;In the second essay, we formulate the quality optimization scheduling problem and prove its NP-completeness. We develop and empirically verify the effectiveness a hybrid solution procedure which combines two components: (1) a linear programming (LP) solver for optimally setting the activity durations of a set of temporally related activities, and (2) a precedence constraint posting (PCP) search procedure for resolving resource conflicts and establishing resource feasibility. Within this approach, the dual concerns of resource feasibility and quality optimization are considered in a tightly integrated fashion. Our analysis centered on the design of search control heuristics. The experimental analysis indicates that a good heuristic must strike the right balance between minimizing quality loss at each step and retaining flexibility for future duration reduction.;In the third essay, we further explore the benefit of retaining flexibility in constructing a high quality schedule. In contrast to the integrated scheme proposed in the second essay, we separately address the two principal difficulties in solving the resource constrained quality maximization problem: at the first stage, find a resource feasible solution; and then at the second stage, achieve good quality while maintaining feasibility. In addition, we adopt state-of-the-art techniques in building temporally flexible schedules: We use an "iterative chaining algorithm" to inject flexibility into the schedule before linear optimization in the second step. The results confirm that there exists a potential synergy between the objectives of maintaining temporal flexibility and maximizing quality, which implies that other existent techniques in building flexible schedules can be adapted to solve this new problem. (Abstract shortened by UMI.).