Artificial intelligence-based nurse scheduling in healthcare: A scoping review of functional features and empirical...

Nurse scheduling is a complex workforce management task that directly influences nurses' fatigue, job satisfaction, and retention, as well as patient safety. Traditional rule-based or manually constructed schedules may not fully accommodate increasing workforce complexity, diverse nurse preferences, and organizational constraints. Recent advances in artificial intelligence have enabled the development of intelligent nurse scheduling systems intended to optimize staffing decisions while balancing organizational efficiency and workforce well-being. However, the scope, functional characteristics, and empirical outcomes of artificial intelligence-based nurse scheduling solutions in healthcare remain to be comprehensively synthesized.

This scoping review aimed to systematically map the existing literature on artificial intelligence-based nurse scheduling solutions in healthcare settings by identifying (1) the types of artificial intelligence-driven scheduling approaches and systems; (2) their key functional features and underlying algorithmic foundations; and (3) the empirical outcomes at organizational, workforce, and patient levels.

Five electronic databases (PubMed, Embase, CINAHL, Web of Science, and IEEE) were searched for English-language, peer-reviewed studies published from January 2000 to December 2025.

A scoping review was conducted following the Joanna Briggs Institute methodology and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews guidelines. Studies describing artificial intelligence-based or algorithm-driven nurse scheduling approaches applied in healthcare settings were included. Data were charted and synthesized to capture system characteristics, algorithmic approaches, functional capabilities, and empirical outcomes.

Eleven studies were included, comprising six algorithm-level and five system-level approaches. Most studies employed artificial intelligence-inspired metaheuristic or hybrid optimization methods, predominantly genetic algorithms, with no use of machine learning or data-driven approaches. System-level studies showed greater functional maturity, particularly in automation and explicit modeling of nurse preferences and fairness, whereas real-time adaptability and integration with hospital information systems were rare. Empirical outcomes were mainly assessed using algorithmic or simulation-based metrics, focusing on technical performance and operational efficiency. Nurse-related outcomes increasingly incorporated user-reported measures, whereas organizational and patient-related outcomes were assessed exclusively through simulation-based indicators.

Although artificial intelligence-based nurse scheduling research has advanced technically, most approaches focus on optimization rather than clinical integration. Limited interoperability, real-time functionality, and real-world outcome evaluation suggest that translation into routine practice is still in the early stages. Future research should explore implementation-oriented, human-in-the-loop approaches supported by robust nursing data infrastructure and longitudinal real-world evaluation.

Open Science Framework Registries (Registered on December 26, 2025; https://osf.io/67tsd).

To solve workforce complexity, artificial intelligence-based nurse scheduling must go beyond algorithms to consider organizational factors.