Sign in Subscribe
Machine learning

Artificial Intelligence Improves Patient Care in Emergency Departments

Strategic Insights

— 2 min read

Researchers from Technical University Valencia (TU Valencia) have made a significant breakthrough in using Artificial Intelligence (AI) to optimize patient care in emergency departments. During Seasonal Respiratory Diseases (SRDs), emergency departments often experience nursing shortages, leading to increased patient waiting times, overcrowding, and high infection rates. The study proposes merging AI and Discrete-Event Simulation (DES) to build remedies that reduce waiting times for nursing care.

The AI-based model developed by the researchers achieved a high accuracy of 95.97% in predicting treatment outcomes. The model was tested on a Spanish emergency department and resulted in a significant reduction in median waiting time for respiratory support, ranging from 0.88 to 7.51 hours. The study's findings emphasize the potential of AI in improving patient care and hospital management.

Key Takeaways:

  • Emergency departments during SRDs often experience nursing shortages, leading to increased patient waiting times and overcrowding.
  • The AI-based model developed by the researchers achieved a specificity of 95.97% in predicting treatment outcomes.
  • The model was tested on a Spanish emergency department and resulted in a significant reduction in median waiting time for respiratory support, ranging from 0.88 to 7.51 hours.
  • The study proposes merging AI and DES to build remedies that reduce waiting times for nursing care in emergency departments.
  • The researchers used Extreme Gradient Boosting (XGBoost) to calculate the probability of treatment within the ED wards.
  • The model was evaluated using a simulation model to determine whether the current nurse staff was sufficient to ensure timely treatment of expected respiratory-affected patients.
  • The study pretested three improvement scenarios recommended by hospital administrators to tackle the imbalance problem.
  • The positive and negative predictive scores corresponded to 87.23% and 94.08%, respectively.
  • The Area Under Receiver Operator Characteristic (AU-ROC) curve was 89.00%.
  • The study concluded that the median waiting time for respiratory support use was reduced between 0.88 and 7.51 hours after using a new nurse staffing configuration.

Statistics:

  • Specificity of the AI-based model: 95.97% (CI 95% 93.07% - 97.90%)
  • Specificity of the model: 82.0% (CI 95% 73.05% - 88.96%)
  • Positive and negative predictive scores: 87.23% (CI 95% 78.76% - 93.22%) and 94.08% (95% CI 90.80% - 96.45%), respectively
  • Area Under Receiver Operator Characteristic (AU-ROC) curve: 89.00% (CI 95% 84.46% - 94.78%)
  • Median waiting time for respiratory support use: reduced by 0.88 to 7.51 hours after using a new nurse staffing configuration

Sources:

  • NewsRx. New Artificial Intelligence Data Have Been Reported by Researchers at Technical University Valencia (TU Valencia) (Nurse Staffing Management in the Context of Emergency Departments and Seasonal Respiratory Diseases: An Artificial Intelligence ...). Health & Medicine Week. September 5, 2025; p 2541.
  • Ortiz-Barrios, M., et al. Nurse Staffing Management in the Context of Emergency Departments and Seasonal Respiratory Diseases: An Artificial Intelligence and Discrete-Event Simulation Approach. Journal of Medical Systems, 2025;49(1):106.

Read more