Man vs. machine: Predicting hospital bed demand from an emergency department

Area of Science:

• Healthcare Operations
• Artificial Intelligence in Medicine
• Clinical Decision Support Systems

Background:

• Intelligent systems show promise for healthcare decision-making, including predicting hospital bed demand.
• Accurate prediction of emergency department (ED) bed demand is crucial for resource allocation and reducing hospital strain.
• The autonomy and user-independence of these intelligent systems require further investigation.

Purpose of the Study:

• To compare the predictive performance of a computer-based algorithm against human physicians for hospital bed demand.
• To evaluate predictions based on initial Subjective, Objective, Assessment, Plan (SOAP) records in the ED.
• To assess the feasibility of autonomous operation for intelligent systems in predicting hospital bed demand.

Main Methods:

• A retrospective cohort study involving 9030 patients' electronic medical records (EMR) from a tertiary care hospital.
• A Support Vector Machine Classifier (machine) and four ED physicians (humans) predicted hospital bed demand (admissions/discharges).
• Performance was measured using sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUROC), with a subset of 230 EMRs for testing.

Main Results:

• The machine achieved an AUROC of 0.80 (95% CI: 0.75-0.85), comparable to novice physicians (0.82) and experienced physicians (0.76).
• The intelligent system processed test EMRs in approximately 0.008 seconds, significantly faster than physicians (56.40-156.80 seconds).
• The system demonstrated an overall accuracy of 80% in predicting patient admission or discharge states.

Conclusions:

• The computer-based algorithm demonstrated comparable accuracy to human physicians in predicting hospital bed demand.
• The rapid processing time suggests the algorithm's potential for autonomous and user-independent operation.
• These findings support the use of intelligent systems to enhance efficiency and resource management in emergency departments.