Patient-Aware Logistics in Emergency Medical Services

Emergency Medical Services (EMS) are an integral part of healthcare systems. Their purpose is to respond to medical emergencies, treat patients on scene, and, if necessary, transport them to medical facilities for further care. This thesis studies EMS systems from a logistic perspective. Predictive and prescriptive models from the field of Operations Research can support making logistic decisions in EMS. Such quantitative models require quantitative evaluation criteria for logistic decisions. Evaluation criteria in EMS systems are typically based on response time targets. There is, however, an ongoing debate about how well response time targets can quantify the impact of logistic decisions on the quality of EMS, and specifically on patient outcomes. This debate leads to the fundamental question of this thesis: How can the impact on patient outcomes be considered when evaluating EMS systems from a logistic perspective, and what does this imply for making logistic decisions? This thesis proposes a novel approach to evaluating EMS logistics through utility functions defined for patient categories. It subsequently explores what this approach implies for the operational ambulance dispatching and the strategic ambulance station location and sizing decisions. ... mehrWe find that evaluating EMS systems from a logistic perspective should and can primarily be based on how logistic decisions impact patients. Patients have differing needs, and considering this heterogeneity in logistic decisions can improve patient utilities and reduce required resources.

We structure our research into three topics: Quality evaluation, system behavior, and logistic decisions. Quality evaluation addresses the question of how to evaluate EMS systems from a logistic perspective. System behavior refers to the question of how to assess the impact of logistic decisions on EMS system behavior. The topic of logistic decisions explores how to model logistic decisions in EMS systems. The EMS system in Baden-Württemberg, Germany, serves as the application case for this thesis, enabling us to conduct computational experiments with real-world data.

We approach the topic quality evaluation through a literature review on quality dimensions and quality indicators for EMS logistics. The review highlights the gap in quantifying how logistic decisions impact patient outcomes. We subsequently conduct an expert consensus process to define utility functions for patient categories. Utility functions for heterogeneous patients improve the quantification of how logistic decisions impact patient outcomes compared to evaluating response time targets.

We employ a hybrid simulation approach combining agent-based and discrete-event modeling paradigms to study the topic system behavior. Simulation enables modeling how logistic decisions impact EMS system behavior, and facilitates evaluating logistic decisions before implementing them in practice.

Several logistic decisions need to be made in EMS. We investigate what our findings in the topic quality evaluation imply for the operational ambulance dispatching and the strategic ambulance station location and sizing decisions. We conduct a scoping review of ambulance dispatching strategies that reveals a variety of modeling approaches with differing attributes. A simulation-based comparison of strategies with suitable attributes for our patient categories and application case enables investigating the effects of considering heterogeneous patients. The experiments indicate that differentiated dispatching strategies can significantly reduce driving times to prioritized patients, but appear sensitive to how well prioritized patients can be detected as such during the dispatching decisions. We model the ambulance station location and sizing decision using mixed-integer linear programming. Considering heterogeneous patients affects the validity of existing modeling assumptions beyond integrating patient categories. We formulate three mixed-integer linear programs to relax limiting assumptions and to consider heterogeneous patients. Computational experiments demonstrate that our modeling approach can avoid overcapacities induced by previous modeling approaches and provide insight into the instance-dependent effects of considering heterogeneous patients on required resources and system behavior.

Interdisciplinary approaches facilitate quantifying the connection between medicine and logistics. Patients in EMS have differing needs, and the modeling of logistic decisions should take these needs into account explicitly.