1- Researcher: Dawood Yahya Mohammed Almalki
2- Researcher: Omar Saud Onizan Algethami
3- Researcher: Adnan Hussain Almansour M
4- Researcher: Faisal Nawwar Althobaiti
5- Researcher: Abdulrahman Mohammed Alsherbi
6- Researcher: Abdullah Khudran Almontashri
7- Researcher: Mashhour Awadh Ruddah Althobaiti
8- Researcher: Saeed Mohammed Khalaf Alharhti
Emergency Medical Technician – Saudi Red Crescent Authority
Background
Emergency Medical Services (EMS) operate in environments where demand varies across time and location, while ambulances, personnel, equipment, and dispatch capacity remain limited. Predictive analytics offers a data-driven approach to this challenge by using historical and real-time information to forecast emergency demand and support decisions about ambulance deployment, relocation, staffing, and dispatch. Recent research has increasingly applied statistical forecasting, machine learning, spatial modelling, and optimization methods to EMS resource planning (Alrawashdeh et al., 2024; Martin et al., 2021; Tluli et al., 2024).
Objective
This systematic review aims to examine how predictive analytics has been used in ambulance deployment and EMS resource management, identify the main predictive approaches, summarize their reported operational benefits, and describe the main limitations that prevent wider use in routine EMS practice.
Methods
This systematic review was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidance (Page et al., 2021). English-language studies published from January 2018 to September 2026 were considered. Searches covered PubMed/MEDLINE, Scopus, Web of Science, CINAHL, and IEEE Xplore, with Google Scholar used for supplementary searching. Eligible studies examined predictive or data-driven methods used to forecast EMS demand or support ambulance deployment, dispatching, relocation, staffing, scheduling, or other resource-allocation decisions. Owing to substantial variation in datasets, algorithms, geographic settings, and outcome measures, findings were synthesized narratively.
Results
The reviewed evidence indicates that predictive analytics is being applied mainly to emergency-demand forecasting, spatial and spatiotemporal prediction, and prediction-informed deployment or relocation decisions. Machine-learning methods, including neural networks, gradient boosting, random forests, XGBoost, and ensemble models, have produced promising forecasting results, although established statistical approaches can remain competitive and may be easier to implement (Butsingkorn et al., 2026; Martin et al., 2021; Monks et al., 2023; Shahidian et al., 2025). Recent studies increasingly combine prediction with optimization or reinforcement learning to translate forecasts into ambulance-allocation decisions (Rautenstrauß & Schiffer, 2026; Tsai et al., 2026). Nevertheless, prospective real-world evaluation remains limited, particularly for patient outcomes, workforce effects, equity, and long-term implementation (Becker et al., 2023).
Conclusion
Predictive analytics has substantial potential to support more flexible ambulance deployment and EMS resource management. Its practical value appears greatest when prediction is linked directly to operational decision-making rather than used only to forecast call volume. Future research should emphasize external validation, prospective real-time implementation, patient outcomes, workforce effects, equity, and comparison with existing EMS practice (Becker et al., 2023; Monks et al., 2023; Tsai et al., 2026).
الكلمات المفتاحية
Emergency Medical Services; ambulance deployment; predictive analytics; demand forecasting; resource allocation; machine learning; ambulance relocation; EMS management.,الصفحات: 47-62