JMIR Publications

ABSTRACT

Background:

In healthcare settings, nurse scheduling presents a significant challenge that impacts both patient care quality and nurse well-being. Traditional scheduling methods often neglect individual preferences and constraints, leading to dissatisfaction, burnout, and high turnover rates. The negative effects of inadequate scheduling practices, including restricted autonomy and lack of transparency, can adversely affect nurse morale and patient outcomes. Research indicates that flexible, participative scheduling approaches can mitigate these issues by incorporating nurse preferences and enhancing job satisfaction. Mathematical and AI-based scheduling methods offer potential solutions for optimizing scheduling practices and addressing these challenges.

Objective:

The aim is to develop a comprehensive framework for integrating nurses' preferences into nursing scheduling methods. This involves gathering detailed insights from nursing staff as well as supervisors and mapping these findings to mathematical and AI-based scheduling techniques.

Methods:

This study utilizes and summarizes results of focus group interviews in Swiss healthcare institutions that gain in-depth insights into nurses' experiences and preferences regarding staff scheduling. Interviews provided qualitative data, which were analyzed through open and axial coding to identify key themes. These themes are mapped here to relevant AI methodologies, including Mixed-Integer Programming, Constraint Programming, Genetic Programming, and Reinforcement Learning.

Results:

The interview study identified a strong demand for fair, participative scheduling systems that accommodate individual preferences and ensure transparency. AI-based scheduling is anticipated to improve efficiency and fairness, though concerns about system reliability and the loss of human oversight persist. Mapping key issues to AI methods indicates that these techniques could effectively address specific scheduling challenges, provided they are integrated holistically rather than applied individually, as is still frequently the case.

Conclusions:

Our work underscores the critical role of staff scheduling in healthcare settings and highlights the need for improvements. AI-based scheduling presents a promising solution, offering potential benefits such as increased accuracy, efficiency, and reduced administrative workload. However, it is essential to address concerns related to the limitations of AI, including its ability to fully accommodate individual preferences and the risk of over-reliance on technology.