Assessment of patient flow and optimized use of lean thinking transformation from the perspective of graph theory and spectral graph theory: A case study

Abstract

BACKGROUND: Hospital re-engineering initiatives aiming to meet the requirement for patient-centered care often face significant barriers. Opportunities from the optimization of patient flow logistics are often overlooked due to the perception that patient transport related services are ancillary. OBJECTIVES: To reorganize patient pathways by optimizing inpatient assignment and outpatient unit relocation. METHODS: Our analysis was conducted in a campus-based hospital hosting 1694 inpatient beds. Patient flow data was used for algorithm-based optimization to minimize the sum of the distances due to visits to outpatient units and visits by consulting physicians. Inpatients were reordered and outpatient units were relocated to minimize transport need. Optimized schemes were analyzed using graph- and spectral graph theory. RESULTS: Both optimizations yielded an altered hospital layout in which the need for patient transfers decreased (over 30% and 23% in terms of total distance and transfer episodes, respectively). The optimized systems gave rise to buildings with greater specialization, higher importance in terms of contributing to the network architecture, greater synchronization and robustness. CONCLUSIONS: The top-down algorithm-based optimization scheme yielded a system in which the need for cross-building patient transfer decreased. We suggest that network analysis may be a useful tool for capacity planning.