A Two-Stage Scheduling Method for Nurse Scheduling and Its Practical Application

TL;DR

This paper introduces a two-stage nurse scheduling method that combines automated optimization with manual adjustments, improving practical usability and adoption in hospitals facing staffing challenges.

Contribution

The paper proposes a novel two-stage scheduling approach that integrates human expertise with optimization, facilitating practical nurse schedule creation in real-world hospital settings.

Findings

Successful deployment in acute and chronic care hospitals.

Enhanced schedule practicality through manual adjustments.

Addressed implementation challenges with tailored solutions.

Abstract

The creation of nurses' schedules is a critical task that directly impacts the quality and safety of patient care as well as the quality of life for nurses. In most hospitals in Japan, this responsibility falls to the head nurse of each ward. The physical and mental burden of this task is considerable, and recent challenges such as the growing shortage of nurses and increasingly diverse working styles have further complicated the scheduling process. Consequently, there is a growing demand for automated nurse scheduling systems. Technically, modern integer programming solvers can generate feasible schedules within a practical timeframe. However, in many hospitals, schedules are still created manually. This is largely because tacit knowledge, considerations unconsciously applied by head nurses, cannot be fully formalized into explicit constraints, often resulting in automatically generated schedules that are not practically usable. To address this issue, we propose a novel "two-stage scheduling method." This approach divides the scheduling task into night shift and day shift stages, allowing head nurses to make manual adjustments after the first stage. This interactive process makes it possible to produce nurse schedules that are suitable for real-world implementation. Furthermore, to promote the practical adoption of nurse scheduling, we present case studies from acute and chronic care hospitals where systems based on the proposed method were deployed. We also discuss the challenges encountered during implementation and the corresponding solutions.