Predicting patient admission using machine learning to enable reduction in emergency department wait times
Bibliographic record
Abstract
Emergency Departments (EDs) serve as critical access points for patients requiring urgent and complex medical care, yet sustained overcrowding continues to threaten the quality, safety, and efficiency of service delivery across Canadian healthcare systems.Urban tertiary care centres, in particular, face mounting operational pressures due to increasing patient volumes and limited inpatient bed availability, resulting in significant delays for patients requiring hospital admission.Existing approaches to mitigate ED overcrowding have yielded limited success, often addressing downstream symptoms rather than upstream causes.Early identification of patients likely to require hospitalization could enable proactive resource allocation, streamline care pathways, and improve patient outcomes.This study aimed to develop and evaluate machine learning (ML) models capable of predicting hospital admission using only the information available after completion of the initial triage.We performed a retrospective cohort analysis of 629,737 adult ED visits to the Montreal General Hospital and Royal Victoria Hospital, two academic sites within the McGill University Health Centre network, spanning eight years from 2015 to 2022.After excluding visits with ambiguous or non-comparable disposition outcomes, 526,145 unique visits were included for model development.Several ML classifiers, including logistic regression, decision trees, random forests, gradient boosting machines (XGBoost, LGBM, CatBoost), and artificial neural networks, were trained on structured triage-level data encompassing demographic variables, presenting complaints, arrival mode, vital signs, and healthcare utilization history.Among all models, XGBoost demonstrated the best overall performance, achieving an AUC-ROC of 0.8608 (95% CI: 0.8580-0.8635),with strong sensitivity (0.6691), specificity (0.8589), and a favourable balance of discrimination and calibration.SHAP-based interpretability analysis identified age, Canadian Triage Acuity Scale (CTAS), arrival mode, vital signs (pulse rate, temperature), and prior hospitalizations as the most influential predictors of admission.These findings suggest that meaningful and clinically relevant predictions can be made within minutes of ED arrival, using only routinely collected triage data.Our results demonstrate the potential for ML-based tools to support early
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".