Efficient healthcare logistics with a human touch

Despite the long experienced urgency of rapidly increasing healthcare expenditures, there is still a large potential to improve hospitals' logistical efficiency. Operations Research (OR) methodologies may support healthcare professionals in making better decisions concerning planning and capacity issues and improving efficiency in the delivery of healthcare, but appears to be lacking implementation in practice. This thesis displays OR research that focuses both on theoretical results and practical applications, aiming to improve efficiency while improving compliance with patients' and physicians' preferences. Part I Operations Research applied to hospital wards Hospital wards affect patient flow through the hospital, as the availability of beds determines if new patients are admitted or may undergo surgery. Part I focuses on OR methodologies applied to hospital wards and their integration into practice. Chapter 2 first defines the different OR methodologies applied to hospital wards, and the logistical performance measures considered: blocking probability, utilization and throughput. Next, we review the OR literature applied to hospital wards. Based on logistical characteristics and common patient flow problems, we distinguish the following particular ward types: intensive care, acute medical units, obstetric wards, weekday wards, and general wards. We analyze typical trade-offs of performance measures for each ward type, and the common OR models applied to it. Additionally, we provide four modeling examples, discuss reported experiences with implementation of the research, and highlight voids in the literature that may be directions for future research. Main conclusions of the review are: there exist many papers on OR models relating to different types of wards, actual use of the models in practice seems scarce, and the stakeholders of a project play a significant role in the likelihood of implementation. Additionally, researchers should be thorough in their data collection, sensitivity and robustness analyses, and implementation support. Chapter 3 contains an implementation-oriented case study in which we invoke both queueing theory and discrete event simulation, applied to medical wards of the Jeroen Bosch Hospital (JBH) that experienced unbalanced bed occupancies. In this case study, medical specialists proposed interventions. The effects of all possible interventions were quantified in advance using OR methods, which enabled the steering group to choose the most promising intervention to implement. This intervention was supported by all involved staff, and proved to be effective in reality. Part II Online appointment scheduling Online appointment scheduling refers to systems in which customers receive an immediate answer to their appointment request. This answer is either an appointment time, or a refusal. In an online system, a scheduler has no information about the future when he has to decide on the current request. Therefore, offline systems, in which the scheduler collects all requests before making the decisions, often achieve higher utilization and acceptance rates. Chapter 4 provides an extensive literature review of online appointment scheduling for different application areas, not limited to healthcare. The literature is categorized according to the number of appointments each customer requires, the number of resource types at the facility, and the horizon at which the scheduling decisions are made. We provide an overview of the scheduling decisions, the objectives, and the OR methods applied in different application areas. Different application areas focus on different modeling assumptions and goals, but in all application areas, there is a continuous trade-off between customer-friendliness, for example, short access and waiting times, and system efficiency, for example, little idle time and overtime of the server. Most online appointment scheduling literature considers systems with one resource in which customers make a single appointment. We highlight gaps in the literature that represent opportunities for future research. In Chapter 5 we study the effect of introducing advance reservation, i.e., an online appointment system instead of a walk-in system, on the blocking probability for a general queueing model. It appears that introducing advance reservation may increase or decrease the blocking probability, depending on the system parameters. The lower blocking probabilities are achieved because systems with advance reservation tend to accept many elatively short jobs. Part III Patient appointment schedules Patient appointment schedules determine to a large extent patients' access and waiting times. Part III entails two chapters in which we optimize patient appointment schedules to improve patient flow and reduce specialists' idle time and overtime for two JBH departments. At the breast center of the JBH, patients undergo a series of diagnostic tests on different machines to obtain a diagnose. In Chapter 6 we apply a discrete time queueing model to determine the necessary capacity to comply with the national access time norms. The center implemented interventions based on the model, which concern a new patient appointment schedule and an additional multidisciplinary meeting. We show that the interventions reduce both the appointment and diagnostics delay. Additionally, we propose a promising new patient appointment schedule to further reduce patient waiting times and staff overtime, based on a discrete event simulation. It appears that waiting and overtime may be reduced if patient types are not clustered in the schedule but spread over the day, and idle time is divided equally over the day instead of clustered at the end of the day. Chapter 7 presents an implementation-oriented case study in which we investigate different appointment scheduling policies for the JBH plaster room. At the plaster room patients may either make an appointment or walk-in. Invoking discrete event simulation, we investigate different appointment slot lengths and times, different proportions of patients making appointments instead of walking in, and different arrival rate scenarios. It appears from the simulation results that waiting and overtime are acceptable for most patients and workdays, but exceptionally high values occur for both performance measures. We show that some appointment rules are both easy to implement, and promising to reduce overtime while having similar or less waiting time. Part IV Optimizing doctor schedules Patient appointment schedules build upon specialists' schedules, as appointment can only be scheduled at times that the specialists are available. Part IV focuses on optimizing specialist schedules, to improve compliance with both specialists' preferences and patients' access times norms. In Chapter 8 we optimize the schedule of gynecologists. Gynecologists, and specialists in general, typically have many different tasks, such as seeing outpatients, on call duties, and performing surgeries. Due to specializations, each task may only be performed by a subset of the gynecologists. There are many hard and soft constraints on the sequence and frequency of tasks in a schedule. We invoke mixed integer linear programming to optimize the assignment of tasks to gynecologists and shifts. For practical purposes, we investigate two heuristics that appear to be very effective in swiftly generating good schedules. The resulting roster not only reduces the time to obtain a roster, but also compliance with specialists' preferences and patients' access times. Chapter 9 provides a quantification of the benefits of scheduling outpatient clinic hours dynamically, which entails that part of the capacity is only scheduled in case the access times exceed a certain threshold. The optimal dynamic scheduling policy is obtained through a Markov decision problem. We compare the dynamic schedule to the optimal static schedule, which is obtained invoking an integer linear program. Both methods are applied to the Surgery department of the JBH. Although the precise outcomes are parameter specific, dynamic scheduling significantly improves system utilization and access time norm compliance. Part V Optimizing throughput at the ED Emergency departments (EDs) often experience severe overcrowding, which may put patient lives at risk. Typically, specialists at EDs use multiple rooms in parallel; while one patient awaits test results in a treatment room, the specialist visits other patients. The assignment of rooms among the specialists is often unbalanced, which affects the blocking probability, waiting time and length of stay of patients. In Chapter 10 we analyze patients' expected sojourn times invoking a queueing model in a random environment, for different room assignment policies and working routines of the specialists. We conduct a discrete event simulation to validate our model in case of time-varying arrivals, which are typical for EDs. It appears that a specialist should always be assigned at least two rooms in parallel. Additionally, if a specialist has to choose between seeing a new patient or visiting one of the patients waiting for their test results, the specialist should always take the first option. We extend the research of Chapter 10 in Chapter 11 to incorporate more realistic assumptions, such as time-varying arrivals and patient- and physician-type dependent treatment times. To this end, we invoke a mixed integer programming model in a rolling horizon approach. Additionally, we analyze several extreme cases of the ED system to investigate which policy results in the best performance for which parameter setting. For the realistic ED system, numerical results indicate that giving priority to existing patients results in the best performance. As this policy prescribes the exact opposite compared to the results of Chapter 10, we conclude that it is unlikely that there exists a policy that is optimal for all ED systems. Conclusions Optimizing the efficiency of healthcare logistics often improves both patient-friendliness and quality of care through better accessibility and alignment of the appointments, and may provide the hospital with the possibility to treat more patients with the same capacity. This thesis provides both theoretical and pragmatic applications of OR methodologies that support logistical decision making in hospitals. The research in this thesis contributes to the vast amount of academic papers on OR in healthcare, and additionally provides implementation-oriented case studies and experiences with implementing research results. We focus on inventive, pragmatic mathematical solutions with a human touch, by incorporating both patient and physician preferences in our approaches. With this thesis we intend to bring theory and practice closer together, so academics and practitioners can join forces in the continuous improvement of the efficiency of healthcare logistics.