Association of red blood cell distribution width with hospital admission and in-hospital mortality across all-cause adult emergency department visits.

Objectives: To test the association between the initial red blood cell distribution width (RDW) value in the emergency department (ED) and hospital admission and, among those admitted, in-hospital mortality. Materials and Methods: We perform a retrospective analysis of 210 930 adult ED visits with complete blood count results from March 2013 to February 2022. Primary outcomes were hospital admission and in-hospital mortality. Variables for each visit included demographics, comorbidities, vital signs, basic metabolic panel, complete blood count, and final diagnosis. The association of each outcome with the initial RDW value was calculated across 3 age groups (<45, 45-65, and >65) as well as across 374 diagnosis categories. Logistic regression (LR) and XGBoost models using all variables excluding final diagnoses were built to test whether RDW was a highly weighted and informative predictor for each outcome. Finally, simplified models using only age, sex, and vital signs were built to test whether RDW had additive predictive value. Results: Compared to that of discharged visits (mean [SD]: 13.8 [2.03]), RDW was significantly elevated in visits that resulted in admission (15.1 [2.72]) and, among admissions, those resulting in intensive care unit stay (15.3 [2.88]) and/or death (16.8 [3.25]). This relationship held across age groups as well as across various diagnosis categories. An RDW >16 achieved 90% specificity for hospital admission, while an RDW >18.5 achieved 90% specificity for in-hospital mortality. LR achieved a test area under the curve (AUC) of 0.77 (95% confidence interval [CI] 0.77-0.78) for hospital admission and 0.85 (95% CI 0.81-0.88) for in-hospital mortality, while XGBoost achieved a test AUC of 0.90 (95% CI 0.89-0.90) for hospital admission and 0.96 (95% CI 0.94-0.97) for in-hospital mortality. RDW had high scaled weights and information gain for both outcomes and had additive value in simplified models predicting hospital admission. Discussion: Elevated RDW, previously associated with mortality in myocardial infarction, pulmonary embolism, heart failure, sepsis, and COVID-19, is associated with hospital admission and in-hospital mortality across all-cause adult ED visits. Used alone, elevated RDW may be a specific, but not sensitive, test for both outcomes, with multivariate LR and XGBoost models showing significantly improved test characteristics. Conclusions: RDW, a component of the complete blood count panel routinely ordered as the initial workup for the undifferentiated patient, may be a generalizable biomarker for acuity in the ED.

Generating contextual embeddings for emergency department chief complaints.

OBJECTIVE: We learn contextual embeddings for emergency department (ED) chief complaints using Bidirectional Encoder Representations from Transformers (BERT), a state-of-the-art language model, to derive a compact and computationally useful representation for free-text chief complaints. MATERIALS AND METHODS: Retrospective data on 2.1 million adult and pediatric ED visits was obtained from a large healthcare system covering the period of March 2013 to July 2019. A total of 355 497 (16.4%) visits from 65 737 (8.9%) patients were removed for absence of either a structured or unstructured chief complaint. To ensure adequate training set size, chief complaint labels that comprised less than 0.01%, or 1 in 10 000, of all visits were excluded. The cutoff threshold was incremented on a log scale to create seven datasets of decreasing sparsity. The classification task was to predict the provider-assigned label from the free-text chief complaint using BERT, with Long Short-Term Memory (LSTM) and Embeddings from Language Models (ELMo) as baselines. Performance was measured as the Top-k accuracy from k = 1:5 on a hold-out test set comprising 5% of the samples. The embedding for each free-text chief complaint was extracted as the final 768-dimensional layer of the BERT model and visualized using t-distributed stochastic neighbor embedding (t-SNE). RESULTS: The models achieved increasing performance with datasets of decreasing sparsity, with BERT outperforming both LSTM and ELMo. The BERT model yielded Top-1 accuracies of 0.65 and 0.69, Top-3 accuracies of 0.87 and 0.90, and Top-5 accuracies of 0.92 and 0.94 on datasets comprised of 434 and 188 labels, respectively. Visualization using t-SNE mapped the learned embeddings in a clinically meaningful way, with related concepts embedded close to each other and broader types of chief complaints clustered together. DISCUSSION: Despite the inherent noise in the chief complaint label space, the model was able to learn a rich representation of chief complaints and generate reasonable predictions of their labels. The learned embeddings accurately predict provider-assigned chief complaint labels and map semantically similar chief complaints to nearby points in vector space. CONCLUSION: Such a model may be used to automatically map free-text chief complaints to structured fields and to assist the development of a standardized, data-driven ontology of chief complaints for healthcare institutions.

Predicting 72-hour and 9-day return to the emergency department using machine learning.

OBJECTIVES: To predict 72-h and 9-day emergency department (ED) return by using gradient boosting on an expansive set of clinical variables from the electronic health record. METHODS: This retrospective study included all adult discharges from a level 1 trauma center ED and a community hospital ED covering the period of March 2013 to July 2017. A total of 1500 variables were extracted for each visit, and samples split randomly into training, validation, and test sets (80%, 10%, and 10%). Gradient boosting models were fit on 3 selections of the data: administrative data (demographics, prior hospital usage, and comorbidity categories), data available at triage, and the full set of data available at discharge. A logistic regression (LR) model built on administrative data was used for baseline comparison. Finally, the top 20 most informative variables identified from the full gradient boosting models were used to build a reduced model for each outcome. RESULTS: A total of 330 631 discharges were available for analysis, with 29 058 discharges (8.8%) resulting in 72-h return and 52 748 discharges (16.0%) resulting in 9-day return to either ED. LR models using administrative data yielded test AUCs of 0.69 (95% confidence interval [CI] 0.68-0.70) and 0.71(95% CI 0.70-0.72), while gradient boosting models using administrative data yielded test AUCs of 0.73 (95% CI 0.72-0.74) and 0.74 (95% CI 0.73-0.74) for 72-h and 9-day return, respectively. Gradient boosting models using variables available at triage yielded test AUCs of 0.75 (95% CI 0.74-0.76) and 0.75 (95% CI 0.74-0.75), while those using the full set of variables yielded test AUCs of 0.76 (95% CI 0.75-0.77) and 0.75 (95% CI 0.75-0.76). Reduced models using the top 20 variables yielded test AUCs of 0.73 (95% CI 0.71-0.74) and 0.73 (95% CI 0.72-0.74). DISCUSSION AND CONCLUSION: Gradient boosting models leveraging clinical data are superior to LR models built on administrative data at predicting 72-h and 9-day returns.

Predicting hospital admission at emergency department triage using machine learning.

OBJECTIVE: To predict hospital admission at the time of ED triage using patient history in addition to information collected at triage. METHODS: This retrospective study included all adult ED visits between March 2014 and July 2017 from one academic and two community emergency rooms that resulted in either admission or discharge. A total of 972 variables were extracted per patient visit. Samples were randomly partitioned into training (80%), validation (10%), and test (10%) sets. We trained a series of nine binary classifiers using logistic regression (LR), gradient boosting (XGBoost), and deep neural networks (DNN) on three dataset types: one using only triage information, one using only patient history, and one using the full set of variables. Next, we tested the potential benefit of additional training samples by training models on increasing fractions of our data. Lastly, variables of importance were identified using information gain as a metric to create a low-dimensional model. RESULTS: A total of 560,486 patient visits were included in the study, with an overall admission risk of 29.7%. Models trained on triage information yielded a test AUC of 0.87 for LR (95% CI 0.86-0.87), 0.87 for XGBoost (95% CI 0.87-0.88) and 0.87 for DNN (95% CI 0.87-0.88). Models trained on patient history yielded an AUC of 0.86 for LR (95% CI 0.86-0.87), 0.87 for XGBoost (95% CI 0.87-0.87) and 0.87 for DNN (95% CI 0.87-0.88). Models trained on the full set of variables yielded an AUC of 0.91 for LR (95% CI 0.91-0.91), 0.92 for XGBoost (95% CI 0.92-0.93) and 0.92 for DNN (95% CI 0.92-0.92). All algorithms reached maximum performance at 50% of the training set or less. A low-dimensional XGBoost model built on ESI level, outpatient medication counts, demographics, and hospital usage statistics yielded an AUC of 0.91 (95% CI 0.91-0.91). CONCLUSION: Machine learning can robustly predict hospital admission using triage information and patient history. The addition of historical information improves predictive performance significantly compared to using triage information alone, highlighting the need to incorporate these variables into prediction models.

Inferring the Origin of Metastases from Cancer Phylogenies.

Determining the evolutionary history of metastases is a key problem in cancer biology. Several recent studies have presented inferences regarding the origin of metastases based on phylogenies of cancer lineages. Many of these studies have concluded that the observed monophyly of metastatic subclones favored metastasis-to-metastasis spread ("a metastatic cascade" rather than parallel metastases from the primary tumor). In this article, we argue that identifying a monophyletic clade of metastatic subclones does not provide sufficient evidence to unequivocally establish a history of metastatic cascades. In the absence of a complete phylogeny of the subclones within the primary tumor, a scenario of parallel metastatic events from the primary tumor is an equally plausible interpretation. Future phylogenetic studies on the origin of metastases should obtain a complete phylogeny of subclones within the primary tumor. This complete phylogeny may be obtainable by ultra-deep sequencing and phasing of large sections or by targeted sequencing of many small, spatially heterogeneous sections, followed by phylogenetic reconstruction using well-established molecular evolutionary models. In addition to resolving the evolutionary history of metastases, a complete phylogeny of subclones within the primary tumor facilitates the identification of driver mutations by application of phylogeny-based tests of natural selection.