A machine learning model integrating free-text chief complaints with structured triage variables accurately predicted heart failure hospitalization with strong discrimination (AUC = 0.896).
Cohort (n=270,596)
Yes
Does a machine learning model using triage-available data accurately predict heart failure hospitalization in adult ED encounters?
Integrating free-text chief complaints with structured triage variables in a machine learning model accurately predicts heart failure hospitalization at ED presentation.
Effect estimate: AUC 0.896
BACKGROUND: Early identification of patients at risk for heart failure (HF) hospitalization in the emergency department (ED) is challenging because definitive diagnostic tests are often unavailable at triage. Chief complaint narratives contain rich symptom information but are rarely leveraged for early risk stratification. We sought to develop and validate a machine learning model that predicts HF hospitalization using only data available at ED intake, including free-text chief complaints and structured triage variables. METHODS: We conducted a retrospective cohort study of 270,596 adult ED-to-inpatient encounters across a large integrated health system (2016-2021). The primary outcome was HF hospitalization, defined by a primary discharge diagnosis of HF. Predictors were limited to triage-available data: demographics, vital signs, comorbidity burden, and free-text chief complaints. Chief complaint text was transformed using term frequency-inverse document frequency and latent semantic analysis, supplemented by clinically defined symptom phenotypes. Logistic regression and light gradient boosting machine (LGBM) models were trained and evaluated on a held-out test set. Model performance was assessed using discrimination, calibration, and precision-oriented thresholds. RESULTS: HF hospitalization occurred in 7.5% of encounters. Models incorporating both structured and natural language processing-derived features achieved the highest performance. The combined LGBM model demonstrated strong discrimination (AUC = 0.896), recall (0.816), and precision (0.630 at the default threshold), outperforming structured-only and NLP-only models. Symptom clusters related to dyspnea and edema were among the strongest predictors. CONCLUSION: HF hospitalization can be accurately predicted at ED presentation using only triage-available data. Integrating free-text chief complaints with structured variables substantially improves early risk stratification and may support earlier diagnostic evaluation and resource planning in acute care settings.
Emakhu et al. (2026) conducted a cohort in Heart failure (n=270,596). Machine learning model (LGBM) using structured and NLP-derived features vs. Structured-only and NLP-only models was evaluated on Heart failure hospitalization, defined by a primary discharge diagnosis of HF (AUC 0.896). A machine learning model integrating free-text chief complaints with structured triage variables accurately predicted heart failure hospitalization with strong discrimination (AUC = 0.896).