AI Science Tracker
EN

An Artificial Intelligence Model for Predicting Trauma Mortality Among Emergency Department Patients in South Korea: Retrospective Cohort Study

Kyung Hee University · 29 Aug 2023 · medicine

Discovery

An Artificial Intelligence Model for Predicting Trauma Mortality Among Emergency Department Patients in South Korea: Retrospective Cohort Study

Source check: Source checked. Not independent scientific confirmation.

Finding: Using Korea’s National Emergency Department Information System data for 6.54 million visits at more than 400 hospitals (2016–2019), an AdaBoost–LightGBM ensemble predicted trauma mortality with AUROC 0.9974 in five-fold cross-validation. It outperformed XGBoost and conventional injury/triage scores in the reported dataset. The authors state that external validation is needed; this does not establish clinical benefit.

Technology: AdaBoost–LightGBM ensemble for emergency-department trauma mortality

Publication: Publication status unspecified in annotation · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
The AdaBoost plus LightGBM ensemble predicted emergency-department mortality from nationwide NEDIS records and identified mortality-associated features; AUROC was 0.9974 in retrospective evaluation.
Test and result
Five-fold cross-validation on 6,536,306 Korean emergency visits from more than 400 hospitals (2016–2019) evaluated mortality prediction; the all-feature AdaBoost–LightGBM ensemble reached AUROC 0.9974, compared with XGBoost and conventional injury/triage scores.
Limitation / still unknown
The authors state that external validation is needed and note potential overfitting. Prediction performance does not show that the public risk calculator improved triage or patient outcomes.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://doi.org/10.2196/49283 · https://europepmc.org/article/MED/37642984 · https://europepmc.org/articles/PMC10498319 · https://www.ncbi.nlm.nih.gov/research/bionlp/RESTful/pmcoa.cgi/BioC_json/PMC10498319/unicode · https://pubmed.ncbi.nlm.nih.gov/37642984/ · https://www.jmir.org/2023/1/e49283

The legacy label “Not reported” does not distinguish information absent from reviewed material from information not assessed. It is not a scientific finding of absence.

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse medicine discoveries.