Machine Learning-Based Prediction of Clinical Outcomes for Children During Emergency Department Triage

Advancing Emergency Department Triage Prediction With Machine Learning to Optimize Triage for Abdominal Pain Surgery Patients

BACKGROUND

The development of emergency department (ED) triage systems remains challenging in accurately differentiating patients with acute abdominal pain (AAP) who are critical and urgent for surgery due to subjectivity and limitations. We use machine learning models to predict emergency surgical abdominal pain patients in triage, and then compare their performance with conventional Logistic regression models.

METHODS

Using 38 214 patients presenting with acute abdominal pain at Zhongnan Hospital of Wuhan University between March 1, 2014, and March 1, 2022, we identified all adult patients (aged ≥18 years). We utilized routinely available triage data in electronic medical records as predictors, including structured data (eg, triage vital signs, gender, and age) and unstructured data (chief complaints and physical examinations in free-text format). The primary outcome measure was whether emergency surgery was performed. The dataset was randomly sampled, with 80% assigned to the training set and 20% to the test set. We developed 5 machine learning models: Light Gradient Boosting Machine (Light GBM), eXtreme Gradient Boosting (XGBoost), Deep Neural Network (DNN), and Random Forest (RF). Logistic regression (LR) served as the reference model. Model performance was calculated for each model, including the area under the receiver-work characteristic curve (AUC) and net benefit (decision curve), as well as the confusion matrix.

RESULTS

Of all the 38 214 acute abdominal pain patients, 4208 underwent emergency abdominal surgery while 34 006 received non-surgical treatment. In the surgery outcome prediction, all 4 machine learning models outperformed the reference model (eg, AUC, 0.899 [95%CI 0.891-0.903] in the Light GBM vs. 0.885 [95%CI 0.876-0.891] in the reference model), Similarly, most machine learning models exhibited significant improvements in net reclassification compared to the reference model (eg, NRIs of 0.0812[95%CI, 0.055-0.1105] in the XGBoost), with the exception of the RF model. Decision curve analysis shows that across the entire range of thresholds, the net benefits of the XGBoost and the Light GBM models were higher than the reference model. In particular, the Light GBM model performed well in predicting the need for emergency abdominal surgery with higher sensitivity, specificity, and accuracy.

CONCLUSIONS

Machine learning models have demonstrated superior performance in predicting emergency abdominal pain surgery compared to traditional models. Modern machine learning improves clinical triage decisions and ensures that critically needy patients receive priority for emergency resources and timely, effective treatment.

Travel Distances for Interhospital Transfers of Critically Ill Children: A Geospatial Analysis

IMPORTANCE

The U.S. pediatric acute care system has become more centralized, placing increasing importance on interhospital transfers.

OBJECTIVES

We conducted a geospatial analysis of critically ill children undergoing interfacility transfer with a specific focus on understanding travel distances between the patient's residence and the hospitals in which they receive care.

DESIGN, SETTING, AND PARTICIPANTS

Retrospective geospatial analysis using five U.S. state-level administrative databases; four states observed from 2016 to 2019 and one state from 2018 to 2019. Participants included 10,665 children who experienced 11,713 episodes of critical illness involving transfer between two hospitals.

MAIN OUTCOMES AND MEASURES

Travel distances and the incidence of "potentially suboptimal triage," in which patients were transferred to a second hospital less than five miles further from their residence than the first hospital.

RESULTS

Patients typically present to hospitals near their residence (median distance from residence to first hospital, 4.2 miles; interquartile range [IQR], 1.8-9.6 miles). Transfer distances are relatively large (median distance between hospitals, 28.9 miles; IQR, 11.2-53.2 miles), taking patients relatively far away from their residences (median distance from residence to second hospital, 30.1 miles; IQR, 12.2-54.9 miles). Potentially suboptimal triage was frequent: 24.2 percent of patients were transferred to a hospital less than five miles further away from their residence than the first hospital. Potentially suboptimal triage was most common in children living in urban counties, and became less common with increasing medical complexity.

CONCLUSIONS AND RELEVANCE

The current pediatric critical care system is organized in a hub-and-spoke model, which requires large travel distances for some patients. Some transfers might be prevented by more efficient prehospital triage. Current transfer patterns suggest the choice of initial hospital is influenced by geography as well as by attempts to match hospital resources with perceived patient needs.

Emergency Severity Index Version 4 and Triage of Pediatric Emergency Department Patients

Importance

Most emergency departments (EDs) across the US use the Emergency Severity Index (ESI) to predict acuity and resource needs. A comprehensive assessment of ESI accuracy among pediatric patients is lacking.

Objective

To assess the frequency of mistriage using ESI (version 4) among pediatric ED visits using automated measures of mistriage and identify characteristics associated with mistriage.

Design, Setting, and Participants

This cohort study used operational measures for each ESI level to classify encounters as undertriaged, overtriaged, or correctly triaged to assess the accuracy of the ESI and identify characteristics of mistriage. Participants were pediatric patients at 21 EDs within Kaiser Permanente Northern California from January 1, 2016, to December 31, 2020. During that time, version 4 of the ESI was in use by these EDs. Visits with missing ESI, incomplete ED time variables, patients transferred from another ED, and those who left against medical advice or without being seen were excluded. Data were analyzed between January 2022 and June 2023.

Exposures

Assigned ESI level.

Main Outcomes and Measures

Rates of undertriage and overtriage by assigned ESI level based on mistriage algorithm, patient and visit characteristics associated with undertriage and overtriage.

Results

This study included 1 016 816 pediatric ED visits; the mean (SD) age of patients was 7.3 (5.6) years, 479 610 (47.2%) were female, and 537 206 (52.8%) were male. Correct triage occurred in 346 918 visits (34.1%; 95% CI, 34.0%-34.2%), while overtriage and undertriage occurred in 594 485 visits (58.5%; 95% CI, 58.4%-58.6%) and 75 413 visits (7.4%; 95% CI, 7.4%-7.5%), respectively. In adjusted analyses, undertriage was more common among children at least 6 years old compared with those younger 6 years; male patients compared with female patients; patients with Asian, Black, or Hispanic or other races or ethnicities compared with White patients; patients with comorbid illnesses compared with those without; and patients who arrived by ambulance compared with nonambulance patients.

Conclusions and Relevance

This multicenter retrospective study found that mistriage with ESI version 4 was common in pediatric ED visits. There is an opportunity to improve pediatric ED triage, both in early identification of critically ill patients (limit undertriage) and in more accurate identification of low-acuity patients with low resource needs (limit overtriage). Future research should include assessments based on version 5 of the ESI, which was released after this study was completed.

Predicting severe outcomes in pediatric trauma patients: Shock index pediatric age-adjusted vs. age-adjusted tachycardia

INTRODUCTION

When an injured patient arrives in the Emergency Department (ED), timely and appropriate care is crucial. Shock Index Pediatric Age-Adjusted (SIPA) has been shown to accurately identify pediatric patients in need of emergency interventions. However, no study has evaluated SIPA against age-adjusted tachycardia (AT). This study aims to compare SIPA with AT in predicting outcomes such as mortality, severe injury, and the need for emergent intervention in pediatric trauma patients.

MATERIAL AND METHODS

This is a retrospective cross-sectional analysis of patient data abstracted from the Trauma Quality Improvement Program Participant Use Files (TQIP PUFs) for years 2013-2020. Patients aged 4-16 with blunt mechanism of injury and injury severity score (ISS) > 15 were included. 36,517 children met this criteria. Sensitivity, specificity, overtriage, and undertriage rates were calculated to compare the effectiveness of AT and elevated SIPA as predictors of severe injuries and need for emergent intervention. Emergent interventions included craniotomy, endotracheal intubation, thoracotomy, laparotomy, or chest tube placement within 24 h of arrival.

RESULTS

AT classified 59% of patients as "high risk," while elevated SIPA identified 26%. Compared to AT patients, a greater proportion of patients with elevated SIPA required a blood transfusion within 24 h (22% vs. 12%, respectively; p < 0.001). In-hospital mortality was higher for the elevated SIPA group than AT (10% vs. 5%, respectively; p < 0.001) as well as the need for emergent operative interventions (43% vs. 32% respectively; p < 0.001). Grade 3 or higher liver/spleen lacerations requiring blood transfusion were also more common among elevated SIPA patients than AT patients (8% vs. 4%, respectively; p < 0.001). AT demonstrated greater sensitivity but lower specificity compared to SIPA across all outcomes. AT showed improved overtriage and undertriage rates compared to SIPA, but this is attributed to identifying a large proportion of the sample as "high risk."

CONCLUSIONS

AT outperforms SIPA in sensitivity for mortality, injury severity and emergent interventions in pediatric trauma patients while the specificity of SIPA is high across these outcomes.

Prediction of post-delivery hemoglobin levels with machine learning algorithms

Predicting postpartum hemorrhage (PPH) before delivery is crucial for enhancing patient outcomes, enabling timely transfer and implementation of prophylactic therapies. We attempted to utilize machine learning (ML) using basic pre-labor clinical data and laboratory measurements to predict postpartum Hemoglobin (Hb) in non-complicated singleton pregnancies. The local databases of two academic care centers on patient delivery were incorporated into the current study. Patients with preexisting coagulopathy, traumatic cases, and allogenic blood transfusion were excluded from all analyses. The association of pre-delivery variables with 24-h post-delivery hemoglobin level was evaluated using feature selection with Elastic Net regression and Random Forest algorithms. A suite of ML algorithms was employed to predict post-delivery Hb levels. Out of 2051 pregnant women, 1974 were included in the final analysis. After data pre-processing and redundant variable removal, the top predictors selected via feature selection for predicting post-delivery Hb were parity (B: 0.09 [0.05-0.12]), gestational age, pre-delivery hemoglobin (B:0.83 [0.80-0.85]) and fibrinogen levels (B:0.01 [0.01-0.01]), and pre-labor platelet count (B*1000: 0.77 [0.30-1.23]). Among the trained algorithms, artificial neural network provided the most accurate model (Root mean squared error: 0.62), which was subsequently deployed as a web-based calculator: https://predictivecalculators.shinyapps.io/ANN-HB . The current study shows that ML models could be utilized as accurate predictors of indirect measures of PPH and can be readily incorporated into healthcare systems. Further studies with heterogenous population-based samples may further improve the generalizability of these models.

Artificial Intelligence in Pediatric Emergency Medicine: Applications, Challenges, and Future Perspectives

The dawn of Artificial intelligence (AI) in healthcare stands as a milestone in medical innovation. Different medical fields are heavily involved, and pediatric emergency medicine is no exception. We conducted a narrative review structured in two parts. The first part explores the theoretical principles of AI, providing all the necessary background to feel confident with these new state-of-the-art tools. The second part presents an informative analysis of AI models in pediatric emergencies. We examined PubMed and Cochrane Library from inception up to April 2024. Key applications include triage optimization, predictive models for traumatic brain injury assessment, and computerized sepsis prediction systems. In each of these domains, AI models outperformed standard methods. The main barriers to a widespread adoption include technological challenges, but also ethical issues, age-related differences in data interpretation, and the paucity of comprehensive datasets in the pediatric context. Future feasible research directions should address the validation of models through prospective datasets with more numerous sample sizes of patients. Furthermore, our analysis shows that it is essential to tailor AI algorithms to specific medical needs. This requires a close partnership between clinicians and developers. Building a shared knowledge platform is therefore a key step.

Nomogram for preoperative differentiation of benign and malignant breast tumors using contrast-enhanced cone-beam breast CT (CE CB-BCT) quantitative imaging and assessment features

PURPOSE

Breast cancer's impact necessitates refined diagnostic approaches. This study develops a nomogram using radiology quantitative features from contrast-enhanced cone-beam breast CT for accurate preoperative classification of benign and malignant breast tumors.

MATERIAL AND METHODS

A retrospective study enrolled 234 females with breast tumors, split into training and test sets. Contrast-enhanced cone-beam breast CT-images were acquired using Koning Breast CT-1000. Quantitative assessment features were extracted via 3D-slicer software, identifying independent predictors. The nomogram was constructed to preoperative differentiation benign and malignant breast tumors. Calibration curve was used to assess whether the model showed favorable correspondence with pathological confirmation. Decision curve analysis confirmed the model's superiority.

RESULTS

The study enrolled 234 female patients with a mean age of 50.2 years (SD ± 9.2). The training set had 164 patients (89 benign, 75 malignant), and the test set had 70 patients (29 benign, 41 malignant). The nomogram achieved excellent predictive performance in distinguishing benign and malignant breast lesions with an AUC of 0.940 (95% CI 0.900-0.940) in the training set and 0.970 (95% CI 0.940-0.970) in the test set.

CONCLUSION

This study illustrates the effectiveness of quantitative radiology features derived from contrast-enhanced cone-beam breast CT in distinguishing between benign and malignant breast tumors. Incorporating these features into a nomogram-based diagnostic model allows for breast tumor diagnoses that are objective and possess good accuracy. The application of these insights could substantially increase reliability and efficacy in the management of breast tumors, offering enhanced diagnostic capability.

Use of Artificial Intelligence in Triage in Hospital Emergency Departments: A Scoping Review

The integration of artificial intelligence (AI) and machine learning (ML) in healthcare has become a major point of interest and raises the question of its impact on the emergency department (ED) triaging process. AI's capacity to emulate human cognitive processes coupled with advancements in computing has shown positive outcomes in various aspects of healthcare but little is known about the use of AI in triaging patients in ED. AI algorithms may allow for earlier diagnosis and intervention; however, overconfident answers may present dangers to patients. The purpose of this review was to explore comprehensively recently published literature regarding the effect of AI and ML in ED triage and identify research gaps. A systemized search was conducted in September 2023 using the electronic databases EMBASE, Ovid MEDLINE, and Web of Science. To meet inclusion criteria, articles had to be peer-reviewed, written in English, and based on primary data research studies published in US journals 2013-2023. Other criteria included 1) studies with patients needing to be admitted to hospital EDs, 2) AI must have been used when triaging a patient, and 3) patient outcomes must be represented. The search was conducted using controlled descriptors from the Medical Subject Headings (MeSH) that included the terms "artificial intelligence" OR "machine learning" AND "emergency ward" OR "emergency care" OR "emergency department" OR "emergency room" AND "patient triage" OR "triage" OR "triaging." The search initially identified 1,142 citations. After a rigorous, systemized screening process and critical appraisal of the evidence, 29 studies were selected for the final review. The findings indicated that 1) ML models consistently demonstrated superior discrimination abilities compared to conventional triage systems, 2) the integration of AI into the triage process yielded significant enhancements in predictive accuracy, disease identification, and risk assessment, 3) ML accurately determined the necessity of hospitalization for patients requiring urgent attention, and 4) ML improved resource allocation and quality of patient care, including predicting length of stay. The suggested superiority of ML models in prioritizing patients in the ED holds the potential to redefine triage precision.

Triage Accuracy in Pediatrics Using the Emergency Severity Index

INTRODUCTION

Although the Emergency Severity Index is the most widely used tool in the United States to prioritize care for patients who seek emergency care, including children, there are significant deficiencies in the tool's performance. Inaccurate triage has been associated with delayed treatment, unnecessary diagnostic testing, and bias in clinical care. We evaluated the accuracy of the Emergency Severity Index to stratify patient priority based on predicted resource utilization in pediatric emergency department patients and identified covariates influencing performance.

METHODS

This cross-sectional, retrospective study used a data platform that links clinical and research data sets from a single freestanding pediatric hospital in the United States. Chi-square analysis was used to describes rates of over- and undertriage. Mixed effects ordinal logistic regression identified associations between Emergency Severity Index categories assigned at triage and key emergency department resources using discrete data elements and natural language processing of text notes.

RESULTS

We analyzed 304,422 emergency department visits by 153,984 unique individuals in the final analysis; 80% of visits were triaged as lower acuity Emergency Severity Index levels 3 to 5, with the most common level being Emergency Severity Index 4 (43%). Emergency department visits scored Emergency Severity Index levels 3 and 4 were triaged accurately 46% and 38%, respectively. We noted racial differences in overall triage accuracy.

DISCUSSION

Although the plurality of patients was scored as Emergency Severity Index 4, 50% were mistriaged, and there were disparities based on race indicating Emergency Severity Index mistriages pediatric patients. Further study is needed to elucidate the application of the Emergency Severity Indices in pediatrics using a multicenter emergency department population with diverse clinical and demographic characteristics.

Factors predictive of hospital admission for children via emergency departments in Australia and Sweden: an observational cross-sectional study

BACKGROUND

Identifying factors predictive of hospital admission can be useful to prospectively inform bed management and patient flow strategies and decrease emergency department (ED) crowding. It is largely unknown if admission rate or factors predictive of admission vary based on the population to which the ED served (i.e., children only, or both adults and children). This study aimed to describe the profile and identify factors predictive of hospital admission for children who presented to four EDs in Australia and one ED in Sweden.

METHODS

A multi-site observational cross-sectional study using routinely collected data pertaining to ED presentations made by children < 18 years of age between July 1, 2011 and October 31, 2012. Univariate and multivariate analysis were undertaken to determine factors predictive of hospital admission.

RESULTS

Of the 151,647 ED presentations made during the study period, 22% resulted in hospital admission. Admission rate varied by site; the children's EDs in Australia had higher admission rates (South Australia: 26%, Queensland: 23%) than the mixed (adult and children's) EDs (South Australia: 13%, Queensland: 17%, Sweden: 18%). Factors most predictive of hospital admission for children, after controlling for triage category, included hospital type (children's only) adjusted odds ratio (aOR):2.3 (95%CI: 2.2-2.4), arrival by ambulance aOR:2.8 (95%CI: 2.7-2.9), referral from primary health aOR:1.5 (95%CI: 1.4-1.6) and presentation with a respiratory or gastrointestinal condition (aOR:2.6, 95%CI: 2.5-2.8 and aOR:1.5, 95%CI: 1.4-1.6, respectively). Predictors were similar when each site was considered separately.

CONCLUSIONS

Although the characteristics of children varied by site, factors predictive of hospital admission were mostly similar. The awareness of these factors predicting the need for hospital admission can support the development of clinical pathways.

Utilization of Machine Learning Approaches to Predict Mortality in Pediatric Warzone Casualties

BACKGROUND

Identification of pediatric trauma patients at the highest risk for death may promote optimization of care. This becomes increasingly important in austere settings with constrained medical capabilities. This study aimed to develop and validate predictive models using supervised machine learning (ML) techniques to identify pediatric warzone trauma patients at the highest risk for mortality.

METHODS

Supervised learning approaches using logistic regression (LR), support vector machine (SVM), neural network (NN), and random forest (RF) models were generated from the Department of Defense Trauma Registry, 2008-2016. Models were tested and compared to determine the optimal algorithm for mortality.

RESULTS

A total of 2,007 patients (79% male, median age range 7-12 years old, 62.5% sustaining penetrating injury) met the inclusion criteria. Severe injury (Injury Severity Score > 15) was noted in 32.4% of patients, while overall mortality was 7.13%. The RF and SVM models displayed recall values of .9507 and .9150, while LR and NN displayed values of .8912 and .8895, respectively. Random forest (RF) outperformed LR, SVM, and NN on receiver operating curve (ROC) analysis demonstrating an area under the ROC of .9752 versus .9252, .9383, and .8748, respectively.

CONCLUSION

Machine learning (ML) techniques may prove useful in identifying those at the highest risk for mortality within pediatric trauma patients from combat zones. Incorporation of advanced computational algorithms should be further explored to optimize and supplement the diagnostic and therapeutic decision-making process.

Prediction of major adverse cardiac events in the emergency department using an artificial neural network with a systematic grid search

BACKGROUND

The aim of our research was to design and evaluate an Artificial Neural Network (ANN) model using a systemic grid search for the early prediction of major adverse cardiac events (MACE) among patients presenting to the triage of an emergency department.

METHODS

This is a single-center, cross-sectional study using electronic health records from January 2017 to December 2020. The research population consists of adults coming to our emergency department triage at Aga Khan University Hospital. The MACE during hospitalization was the main outcome. To enhance the architecture of an ANN using triage data, we used a systematic grid search strategy. Four hidden ANN layers were used, followed by an output layer. Following each hidden layer was back normalization and a dropout layer. MACE was predicted using three binary classifiers: ANN, Random Forests (RF), and logistic regression (LR). The overall accuracy, sensitivity, specificity, precision, and recall of these models were examined. Each model was evaluated using the receiver operating characteristic curve (ROC) and an F1-score with a 95% confidence interval.

RESULTS

A total of 97,333 emergency department visits were recorded during the study period, with 33% of patients having cardiovascular symptoms. The mean age was 54.08 (19.18) years old. The MACE was observed in 23,052 (23.7%) of the patients, in-hospital (up to 30 days) mortality in 10,888 (11.2%) patients, and cardiac arrest in 5483 (5.6%) patients. The data used for training and validation were 77,866 and 19,467 in an 80:20 ratio, respectively. The AUC score for MACE with ANN was 0.97, which was greater than RF (0.96) and LR (0.96). Similarly, the precision-recall curve for MACE utilizing ANN was greater (0.94 vs. 0.93 for RF and 0.93 for LR). The sensitivity for MACE prediction using ANN, RF, and LR classifiers was 99.3%, 99.4%, and 99.2%, respectively, with the specificities being 94.5%, 94.2%, and 94.2%, respectively.

CONCLUSION

When triage data is used to predict MACE, death, and cardiac arrest, ANN with systemic grid search gives precise and valid outcomes and will benefit in predicting MACE in emergency rooms with limited resources that have to deal with a substantial number of patients.

Machine learning models to predict success of endoscopic sleeve gastroplasty using total and excess weight loss percent achievement: a multicentre study

BACKGROUND

The large amount of heterogeneous data collected in surgical/endoscopic practice calls for data-driven approaches as machine learning (ML) models. The aim of this study was to develop ML models to predict endoscopic sleeve gastroplasty (ESG) efficacy at 12 months defined by total weight loss (TWL) % and excess weight loss (EWL) % achievement. Multicentre data were used to enhance generalizability: evaluate consistency among different center of ESG practice and assess reproducibility of the models and possible clinical application. Models were designed to be dynamic and integrate follow-up clinical data into more accurate predictions, possibly assisting management and decision-making.

METHODS

ML models were developed using data of 404 ESG procedures performed at 12 centers across Europe. Collected data included clinical and demographic variables at the time of ESG and at follow-up. Multicentre/external and single center/internal and temporal validation were performed. Training and evaluation of the models were performed on Python's scikit-learn library. Performance of models was quantified as receiver operator curve (ROC-AUC), sensitivity, specificity, and calibration plots.

RESULTS

Multicenter external validation: ML models using preoperative data show poor performance. Best performances were reached by linear regression (LR) and support vector machine models for TWL% and EWL%, respectively, (ROC-AUC: TWL% 0.87, EWL% 0.86) with the addition of 6-month follow-up data. Single-center internal validation: Preoperative data only ML models show suboptimal performance. Early, i.e., 3-month follow-up data addition lead to ROC-AUC of 0.79 (random forest classifiers model) and 0.81 (LR models) for TWL% and EWL% achievement prediction, respectively. Single-center temporal validation shows similar results.

CONCLUSIONS

Although preoperative data only may not be sufficient for accurate postoperative predictions, the ability of ML models to adapt and evolve with the patients changes could assist in providing an effective and personalized postoperative care. ML models predictive capacity improvement with follow-up data is encouraging and may become a valuable support in patient management and decision-making.

Explainable artificial intelligence in emergency medicine: an overview

Artificial intelligence (AI) and machine learning (ML) have potential to revolutionize emergency medical care by enhancing triage systems, improving diagnostic accuracy, refining prognostication, and optimizing various aspects of clinical care. However, as clinicians often lack AI expertise, they might perceive AI as a "black box," leading to trust issues. To address this, "explainable AI," which teaches AI functionalities to end-users, is important. This review presents the definitions, importance, and role of explainable AI, as well as potential challenges in emergency medicine. First, we introduce the terms explainability, interpretability, and transparency of AI models. These terms sound similar but have different roles in discussion of AI. Second, we indicate that explainable AI is required in clinical settings for reasons of justification, control, improvement, and discovery and provide examples. Third, we describe three major categories of explainability: pre-modeling explainability, interpretable models, and post-modeling explainability and present examples (especially for post-modeling explainability), such as visualization, simplification, text justification, and feature relevance. Last, we show the challenges of implementing AI and ML models in clinical settings and highlight the importance of collaboration between clinicians, developers, and researchers. This paper summarizes the concept of "explainable AI" for emergency medicine clinicians. This review may help clinicians understand explainable AI in emergency contexts.

Artificial intelligence differentiates abdominal Henoch-Schönlein purpura from acute appendicitis in children

OBJECTIVE

This study aims to construct an artificial intelligence (AI) model capable of effectively discriminating between abdominal Henoch-Schönlein purpura (AHSP) and acute appendicitis (AA) in pediatric patients.

METHODS

A total of 6965 participants, comprising 2201 individuals with AHSP and 4764 patients with AA, were enrolled in the study. Additionally, 53 laboratory indicators were taken into consideration. Five distinct artificial intelligence (AI) models were developed employing machine learning algorithms, namely XGBoost, AdaBoost, Gaussian Naïve Bayes (GNB), MLPClassifier (MLP), and support vector machine (SVM). The performance of these prediction models was assessed through receiver operating characteristic (ROC) curve analysis, calibration curve assessment, and decision curve analysis (DCA).

RESULTS

We identified 32 discriminative indicators (p < .05) between AHSP and AA. Five indicators, namely the lymphocyte ratio (LYMPH ratio), eosinophil ratio (EO ratio), eosinophil count (EO count), neutrophil ratio (NEUT ratio), and C-reactive protein (CRP), exhibited strong performance in distinguishing AHSP from AA (AUC ≥ 0.80). Among the various prediction models, the XGBoost model displayed superior performance evidenced by the highest AUC (XGBoost = 0.895, other models < 0.89), accuracy (XGBoost = 0.824, other models < 0.81), and Kappa value (XGBoost = 0.621, other models < 0.60) in the validation set. After optimization, the XGBoost model demonstrated remarkable diagnostic performance for AHSP and AA (AUC > 0.95). Both the calibration curve and decision curve analysis suggested the promising clinical utility and net benefits of the XGBoost model.

CONCLUSION

The AI-based machine learning model exhibits high prediction accuracy and can differentiate AHSP and AA from a data-driven perspective.

Predicting adverse perinatal outcomes among gestational diabetes complicated pregnancies using neural network algorithm

OBJECTIVE

The primary aim of this study is to utilize a neural network model to predict adverse neonatal outcomes in pregnancies complicated by gestational diabetes (GDM).

DESIGN

Our model, based on XGBoost, was implemented using Python 3.6 with the Keras framework built on TensorFlow by Google. We sourced data from medical records of GDM-diagnosed individuals who delivered at our tertiary medical center between 2012 and 2016. The model included simple pregnancy parameters, maternal age, body mass index (BMI), parity, gravity, results of oral glucose tests, treatment modality, and glycemic control. The composite neonatal adverse outcomes defined as one of the following: large or small for gestational age, shoulder dystocia, fetal umbilical pH less than 7.2, neonatal intensive care unit (NICU) admission, respiratory distress syndrome (RDS), hyperbilirubinemia, or polycythemia. For the machine training phase, 70% of the cohort was randomly chosen. Each sample in this set consisted of baseline parameters and the composite outcome. The remaining samples were then employed to assess the accuracy of our model.

RESULTS

The study encompassed a total of 452 participants. The composite adverse outcome occurred in 29% of cases. Our model exhibited prediction accuracies of 82% at the time of GDM diagnosis and 91% at delivery. The factors most contributing to the prediction model were maternal age, pre-pregnancy BMI, and the results of the single 3-h 100 g oral glucose tolerance test.

CONCLUSION

Our advanced neural network algorithm has significant potential in predicting adverse neonatal outcomes in GDM-diagnosed individuals.

Application of exponential smoothing method and SARIMA model in predicting the number of admissions in a third-class hospital in Zhejiang Province

OBJECTIVE

To establish the exponential smoothing prediction model and SARIMA model to predict the number of inpatients in a third-class hospital in Zhejiang Province, and evaluate the prediction effect of the two models, and select the best number prediction model.

METHODS

The data of hospital admissions from January 2019 to September 2022 were selected to establish the exponential smoothing prediction model and the SARIMA model respectively. Then compare the fitting parameters of different models: R2_adjusted, R2, Root Mean Square Error (RMSE)、Mean Absolute Percentage Error (MAPE)、Mean Absolute Error(MAE) and standardized BIC to select the best model. Finally, the established model was used to predict the number of hospital admissions from October to December 2022, and the prediction effect of the average relative error judgment model was compared.

RESULTS

The best fitting exponential smoothing prediction model was Winters Addition model, whose R2_adjusted was 0.533, R2 was 0.817, MAPE was 6.133, MAE was 447.341. The best SARIMA model is SARIMA(2,2,2)(0,1,1)12 model, whose R2_adjusted is 0.449, R2 is 0.199, MAPE is 8.240, MAE is 718.965. The Winters addition model and SARIMA(2,2,2)(0,1,1)12 model were used to predict the number of hospital admissions in October-December 2022, respectively. The results showed that the average relative error was 0.038 and 0.015, respectively. The SARIMA(2,2,2)(0,1,1)12 model had a good prediction effect.

CONCLUSION

Both models can better fit the number of admissions, and SARIMA model has better prediction effect.

Machine learning-based predictor for neurologic outcomes in patients undergoing extracorporeal cardiopulmonary resuscitation

Background

We investigated the predictors of poor neurological outcomes in extracorporeal cardiopulmonary resuscitation (ECPR) patients using machine learning (ML) approaches.

Methods

This study was a retrospective, single-center, observational study that included adult patients who underwent ECPR while hospitalized between January 2010 and December 2020. The primary outcome was neurologic status at hospital discharge as assessed by the Cerebral Performance Categories (CPC) score (scores range from 1 to 5). We trained and tested eight ML algorithms for a binary classification task involving the neurological outcomes of survivors after ECPR.

Results

During the study period, 330 patients were finally enrolled in this analysis; 143 (43.3%) had favorable neurological outcomes (CPC score 1 and 2) but 187 (56.7%) did not. From the eight ML algorithms initially considered, we refined our analysis to focus on the three algorithms, eXtreme Gradient Boosting, random forest, and Stochastic Gradient Boosting, that exhibited the highest accuracy. eXtreme Gradient Boosting models exhibited the highest accuracy among all the machine learning algorithms (accuracy: 0.739, area under the curve: 0.837, Kappa: 0.450, sensitivity: 0.700, specificity: 0.740). Across all three ML models, mean blood pressure emerged as the most influential variable, followed by initial serum lactate, and arrest to extracorporeal membrane oxygenation (ECMO) pump-on-time as important predictors in machine learning models for poor neurological outcomes following successful ECPR.

Conclusions

In conclusion, machine learning methods showcased outstanding predictive accuracy for poor neurological outcomes in patients who underwent ECPR.

Predicting internalizing symptoms with machine learning: identifying individuals that need care

Objective The current project aims to identify individuals in urgent need of mental health care, using a machine learning algorithm (random forest). Comparison/contrast with conventional regression analyses is discussed. Participants: A total of 2,409 participants were recruited from an anonymous university, including undergraduate and graduate students, faculty, and staff. Methods: Answers to a COVID-19 impact survey, the Patient Health Questionnaire-9 (PHQ-9), and the Generalized Anxiety Disorder-7 (GAD-7) were collected. The total scores of PHQ-9 and GAD-7 were regressed on six composites that were created from the questionnaire items, based on their topics. A random forest was trained and validated. Results: Results indicate that the random forest model was able to make accurate, prospective predictions (R2 = .429 on average) and we review variables that were deemed predictively relevant. Conclusions: Overall, the study suggests that predictive models can be clinically useful in identifying individuals with internalizing symptoms based on daily life disruption experiences.

Development and validation of an artificial neural network prediction model for postpartum hemorrhage with placenta previa

BACKGROUND

Postpartum hemorrhage (PPH) is a leading cause of maternal morbidity worldwide and placenta previa is one of the major risk factors for PPH in overall population. However, the clinical prediction of PPH remains challenging. This study aimed to investigate an ideal machine learning-based prediction model for PPH in placenta previa parturients with cesarean section.

METHODS

The clinical data of 223 placenta previa parturients who underwent cesarean delivery in our hospital from 2016 to 2019 were retrospectively collected for analysis. An artificial neural network model was designed to predict PPH, defined as blood loss exceeding 1000 mL with 24h after delivery. Twenty clinical variables were selected as predictors. We also applied six conventional machine learning methods as reference models, including support vector machine, decision tree, random forest, gradient boosting decision tree, adaboost and logistic regression. All the models were validated using 5-fold cross-validation. The area under the receiver operating characteristic curve (AUC), precision, recall and the prediction accuracy of each model were reported.

RESULTS

A total of 223 pregnant women were enrolled in this study, including 101 cases (45.29%) experienced PPH. The proposed model achieved superior prediction performance with an AUC of 0.917, an accuracy of 0.851, a precision score of 0.829 and a recall score of 0.851, which outperformed other six conventional machine learning methods.

CONCLUSIONS

Compared to the conventional machine learning approaches, artificial neural network model shows discriminative ability in identifying women's risk of PPH with placenta previa during cesarean section.

Drivers of mortality in COVID ARDS depend on patient sub-type

The most common cause of death in people with COVID-19 is Acute Respiratory Distress Syndrome (ARDS). Prior studies have demonstrated that ARDS is a heterogeneous syndrome and have identified ARDS sub-types (phenoclusters). However, non-COVID-19 ARDS phenoclusters do not clearly apply to COVID-19 ARDS patients. In this retrospective cohort study, we implemented an iterative approach, combining supervised and unsupervised machine learning methodologies, to identify clinically relevant COVID-19 ARDS phenoclusters, as well as characteristics that are predictive of the outcome for each phenocluster. To this end, we applied a supervised model to identify risk factors for hospital mortality for each phenocluster and compared these between phenoclusters and the entire cohort. We trained the models using a comprehensive, preprocessed dataset of 2,864 hospitalized COVID-19 ARDS patients. Our research demonstrates that the risk factors predicting mortality in the overall cohort of COVID-19 ARDS may not necessarily apply to specific phenoclusters. Additionally, some risk factors increase the risk of hospital mortality in some phenoclusters but decrease mortality in others. These phenocluster-specific risk factors would not have been observed with a single predictive model. Heterogeneity in phenoclusters of COVID-19 ARDS as well as the drivers of mortality may partially explain challenges in finding effective treatments for all patients with ARDS.

Intelligent Framework for Early Detection of Severe Pediatric Diseases from Mild Symptoms

Children's health is one of the most significant fields in medicine. Most diseases that result in children's death or long-term morbidity are caused by preventable and treatable etiologies, and they appear in the child at the early stages as mild symptoms. This research aims to develop a machine learning (ML) framework to detect the severity of disease in children. The proposed framework helps in discriminating children's urgent/severe conditions and notifying parents whether a child needs to visit the emergency room immediately or not. The model considers several variables to detect the severity of cases, which are the symptoms, risk factors (e.g., age), and the child's medical history. The framework is implemented by using nine ML methods. The results achieved show the high performance of the proposed framework in identifying serious pediatric diseases, where decision tree and random forest outperformed the other methods with an accuracy rate of 94%. This shows the reliability of the proposed framework to be used as a pediatric decision-making system for detecting serious pediatric illnesses. The results are promising when compared to recent state-of-the-art studies. The main contribution of this research is to propose a framework that is viable for use by parents when their child suffers from any commonly developed symptoms.

Development of a machine learning-based acuity score prediction model for virtual care settings

OBJECTIVE

Healthcare is increasingly digitized, yet remote and automated machine learning (ML) triage prediction systems for virtual urgent care use remain limited. The Canadian Triage and Acuity Scale (CTAS) is the gold standard triage tool for in-person care in Canada. The current work describes the development of a ML-based acuity score modelled after the CTAS system.

METHODS

The ML-based acuity score model was developed using 2,460,109 de-identified patient-level encounter records from three large healthcare organizations (Ontario, Canada). Data included presenting complaint, clinical modifiers, age, sex, and self-reported pain. 2,041,987 records were high acuity (CTAS 1-3) and 416,870 records were low acuity (CTAS 4-5). Five models were trained: decision tree, k-nearest neighbors, random forest, gradient boosting regressor, and neural net. The outcome variable of interest was the acuity score predicted by the ML system compared to the CTAS score assigned by the triage nurse.

RESULTS

Gradient boosting regressor demonstrated the greatest prediction accuracy. This final model was tuned toward up triaging to minimize patient risk if adopted into the clinical context. The algorithm predicted the same score in 47.4% of cases, and the same or more acute score in 95.0% of cases.

CONCLUSIONS

The ML algorithm shows reasonable predictive accuracy and high predictive safety and was developed using the largest dataset of its kind to date. Future work will involve conducting a pilot study to validate and prospectively assess reliability of the ML algorithm to assign acuity scores remotely.

Criticality and clinical department prediction of ED patients using machine learning based on heterogeneous medical data

PROBLEM

Emergency triage faces multiple challenges, including limited medical resources and inadequate manual triage nurses, which cause incorrect triage, overcrowding in the emergency department (ED), and long patient waiting time.

OBJECTIVE

This paper aims to propose and validate an accurate and efficient artificial intelligence-based method for effectively ED triage and alleviating the pressure on medical resources.

METHODS

We propose two novel machine learning models, TransNet and TextRNN, for predicting patient severity levels and clinical departments using heterogeneous medical data in ED triage. Our models employ a parallel structure for feature extraction and incorporate an attention mechanism to extract essential information from the fused features, enabling accurate predictions. The models analyze the triage data (2020-2022) from the ED of Beijing University People's Hospital, incorporating variables (demographics, triage vital signs, and chief complaints) to identify patient severity levels and clinical departments. We performed data cleaning, categorization, and encoding first. Then, we divided the available data into a training set (56%), a validation set (24%), and a test set (20%) by random sampling. Finally, our models underwent 5-fold cross-validation and were compared with other state-of-the-art models.

RESULTS

We comprehensively evaluated the proposed models against various Recurrent Neural Networks (RNN), Convolutional Neural Networks (CNN), Traditional Machine Learning (TML), and Transformer-based (TF) models, achieving excellent performance in predicting triage outcomes. Specifically, TextRNN achieved a prediction success rate of 86.23% [85.86-86.70] for severity levels and 94.30% [94.00-94.46] for clinical departments among 161,198 ED visits. Moreover, TransNet demonstrated higher sensitivities of 84.08% and 90.05% for severity levels and clinical departments, respectively, with specificities of 76.48% and 95.16%. The accuracy of our model is 0.87%, 0.18%, 4.29%, and 1.96%, higher than that of the above four family models on average. Furthermore, our method significantly reduced under-triage by 12.06% and over-triage by 17.92% compared to manual triage.

CONCLUSIONS

Experimental results demonstrated that the proposed models fuse heterogeneous medical data in the triage process, successfully predicting patients' triage outcomes. Our models can improve triage efficiency, reduce the under/over-triage rate, and provide physicians with valuable decision-making support.

AI and machine learning in resuscitation: Ongoing research, new concepts, and key challenges

Aim

Artificial intelligence (AI) and machine learning (ML) are important areas of computer science that have recently attracted attention for their application to medicine. However, as techniques continue to advance and become more complex, it is increasingly challenging for clinicians to stay abreast of the latest research. This overview aims to translate research concepts and potential concerns to healthcare professionals interested in applying AI and ML to resuscitation research but who are not experts in the field.

Main text

We present various research including prediction models using structured and unstructured data, exploring treatment heterogeneity, reinforcement learning, language processing, and large-scale language models. These studies potentially offer valuable insights for optimizing treatment strategies and clinical workflows. However, implementing AI and ML in clinical settings presents its own set of challenges. The availability of high-quality and reliable data is crucial for developing accurate ML models. A rigorous validation process and the integration of ML into clinical practice is essential for practical implementation. We furthermore highlight the potential risks associated with self-fulfilling prophecies and feedback loops, emphasizing the importance of transparency, interpretability, and trustworthiness in AI and ML models. These issues need to be addressed in order to establish reliable and trustworthy AI and ML models.

Conclusion

In this article, we overview concepts and examples of AI and ML research in the resuscitation field. Moving forward, appropriate understanding of ML and collaboration with relevant experts will be essential for researchers and clinicians to overcome the challenges and harness the full potential of AI and ML in resuscitation.

Clinical characteristics of hospitalized children with community-acquired pneumonia and respiratory infections: Using machine learning approaches to support pathogen prediction at admission

BACKGROUND

Acute respiratory infections (ARIs) are common in children. We developed machine learning models to predict pediatric ARI pathogens at admission.

METHODS

We included hospitalized children with respiratory infections between 2010 and 2018. Clinical features were collected within 24 h of admission to construct models. The outcome of interest was the prediction of 6 common respiratory pathogens, including adenovirus, influenza virus types A and B, parainfluenza virus (PIV), respiratory syncytial virus (RSV), and Mycoplasma pneumoniae (MP). Model performance was estimated using area under the receiver operating characteristic curve (AUROC). Feature importance was measured using Shapley Additive exPlanation (SHAP) values.

RESULTS

A total of 12,694 admissions were included. Models trained with 9 features (age, event pattern, fever, C-reactive protein, white blood cell count, platelet count, lymphocyte ratio, peak temperature, peak heart rate) achieved the best performance (AUROC: MP 0.87, 95% CI 0.83-0.90; RSV 0.84, 95% CI 0.82-0.86; adenovirus 0.81, 95% CI 0.77-0.84; influenza A 0.77, 95% CI 0.73-0.80; influenza B 0.70, 95% CI 0.65-0.75; PIV 0.73, 95% CI 0.69-0.77). Age was the most important feature to predict MP, RSV and PIV infections. Event patterns were useful for influenza virus prediction, and C-reactive protein had the highest SHAP value for adenovirus infections.

CONCLUSION

We demonstrate how artificial intelligence can assist clinicians identify potential pathogens associated with pediatric ARIs upon admission. Our models provide explainable results that could help optimize the use of diagnostic testing. Integrating our models into clinical workflows may lead to improved patient outcomes and reduce unnecessary medical costs.

Clinical support system for triage based on federated learning for the Korea triage and acuity scale

Background and aims

This study developed a clinical support system based on federated learning to predict the need for a revised Korea Triage Acuity Scale (KTAS) to facilitate triage.

Methods

This was a retrospective study that used data from 11,952,887 patients in the Korean National Emergency Department Information System (NEDIS) from 2016 to 2018 for model development. Separate cohorts were created based on the emergency medical center level in the NEDIS: regional emergency medical center (REMC), local emergency medical center (LEMC), and local emergency medical institution (LEMI). External and temporal validation used data from emergency department (ED) of the study site from 2019 to 2021. Patient features obtained during the triage process and the initial KTAS scores were used to develop the prediction model. Federated learning was used to rectify the disparity in data quality between EDs. The patient's demographic information, vital signs in triage, mental status, arrival information, and initial KTAS were included in the input feature.

Results

3,626,154 patients' visits were included in the regional emergency medical center cohort; 8,278,081 patients' visits were included in the local emergency medical center cohort; and 48,652 patients' visits were included in the local emergency medical institution cohort. The study site cohort, which is used for external and temporal validation, included 135,780 patients visits. Among the patients in the REMC and study site cohorts, KTAS level 3 patients accounted for the highest proportion at 42.4% and 45.1%, respectively, whereas in the LEMC and LEMI cohorts, KTAS level 4 patients accounted for the highest proportion. The area under the receiver operating characteristic curve for the prediction model was 0.786, 0.750, and 0.770 in the external and temporal validation. Patients with revised KTAS scores had a higher admission rate and ED mortality rate than those with unaltered KTAS scores.

Conclusions

This novel system might accurately predict the likelihood of KTAS acuity revision and support clinician-based triage.

Automated analysis of unstructured clinical assessments improves emergency department triage performance: A retrospective deep learning analysis

Objectives

Efficient and accurate emergency department (ED) triage is critical to prioritize the sickest patients and manage department flow. We explored the use of electronic health record data and advanced predictive analytics to improve triage performance.

Methods

Using a data set of over 5 million ED encounters of patients 18 years and older across 21 EDs from 2016 to 2020, we derived triage models using deep learning to predict 2 outcomes: hospitalization (primary outcome) and fast-track eligibility (exploratory outcome), defined as ED discharge with <2 resource types used (eg, laboratory or imaging studies) and no critical events (eg, resuscitative medications use or intensive care unit [ICU] admission). We report area under the receiver operator characteristic curve (AUC) and 95% confidence intervals (CI) for models using (1) triage variables alone (demographics and vital signs), (2) triage nurse clinical assessment alone (unstructured notes), and (3) triage variables plus clinical assessment for each prediction target.

Results

We found 12.7% of patients were hospitalized (n = 673,659) and 37.0% were fast-track eligible (n = 1,966,615). The AUC was lowest for models using triage variables alone: AUC 0.77 (95% CI 0.77-0.78) and 0.70 (95% CI 0.70-0.71) for hospitalization and fast-track eligibility, respectively, and highest for models incorporating clinical assessment with triage variables for both hospitalization and fast-track eligibility: AUC 0.87 (95% CI 0.87-0.87) for both prediction targets.

Conclusion

Our findings highlight the potential to use advanced predictive analytics to accurately predict key ED triage outcomes. Predictive accuracy was optimized when clinical assessments were added to models using simple structured variables alone.

Implementation of Prediction Models in the Emergency Department from an Implementation Science Perspective-Determinants, Outcomes, and Real-World Impact: A Scoping Review

STUDY OBJECTIVE

Prediction models offer a promising form of clinical decision support in the complex and fast-paced environment of the emergency department (ED). Despite significant advancements in model development and validation, implementation of such models in routine clinical practice remains elusive. This scoping review aims to survey the current state of prediction model implementation in the ED and to provide insights on contributing factors and outcomes from an implementation science perspective.

METHODS

We searched 4 databases from their inception to May 20, 2022: MEDLINE (through PubMed), Embase, Scopus, and CINAHL. Articles that reported implementation outcomes and/or contextual determinants under the Reach, Effectiveness, Adoption, Implementation Maintenance (RE-AIM)/Practical, Robust, Implementation, and Sustainability Model (PRISM) framework were included. Characteristics of studies, models, and results of the RE-AIM/PRISM domains were summarized narratively.

RESULTS

Thirty-six reports on 31 implementations were included. The most common prediction models implemented were early warning scores. The most common implementation strategies used were training stakeholders, infrastructural changes, and using evaluative or iterative strategies. Only one report examined ED patients' perspectives, whereas the rest were focused on the experience of health care workers or organizational stakeholders. Key determinants of successful implementation include strong stakeholder engagement, codevelopment of workflows and implementation strategies, education, and usability.

CONCLUSION

Examining ED prediction models from an implementation science perspective can provide valuable insights and help guide future implementations.

The Aspects of Running Artificial Intelligence in Emergency Care; a Scoping Review

Introduction

Artificial Inteligence (AI) application in emergency medicine is subject to ethical and legal inconsistencies. The purposes of this study were to map the extent of AI applications in emergency medicine, to identify ethical issues related to the use of AI, and to propose an ethical framework for its use.

Methods

A comprehensive literature collection was compiled through electronic databases/internet search engines (PubMed, Web of Science Platform, MEDLINE, Scopus, Google Scholar/Academia, and ERIC) and reference lists. We considered studies published between 1 January 2014 and 6 October 2022. Articles that did not self-classify as studies of an AI intervention, those that were not relevant to Emergency Departments (EDs), and articles that did not report outcomes or evaluations were excluded. Descriptive and thematic analyses of data extracted from the included articles were conducted.

Results

A total of 137 out of the 2175 citations in the original database were eligible for full-text evaluation. Of these articles, 47 were included in the scoping review and considered for theme extraction. This review covers seven main areas of AI techniques in emergency medicine: Machine Learning (ML) Algorithms (10.64%), prehospital emergency management (12.76%), triage, patient acuity and disposition of patients (19.15%), disease and condition prediction (23.40%), emergency department management (17.03%), the future impact of AI on Emergency Medical Services (EMS) (8.51%), and ethical issues (8.51%).

Conclusion

There has been a rapid increase in AI research in emergency medicine in recent years. Several studies have demonstrated the potential of AI in diverse contexts, particularly when improving patient outcomes through predictive modelling. According to the synthesis of studies in our review, AI-based decision-making lacks transparency. This feature makes AI decision-making opaque.

A deep neural network framework to derive interpretable decision rules for accurate traumatic brain injury identification of infants

OBJECTIVE

We aimed to develop a robust framework to model the complex association between clinical features and traumatic brain injury (TBI) risk in children under age two, and identify significant features to derive clinical decision rules for triage decisions.

METHODS

In this retrospective study, four frequently used machine learning models, i.e., support vector machine (SVM), random forest (RF), deep neural network (DNN), and XGBoost (XGB), were compared to identify significant clinical features from 24 input features associated with the TBI risk in children under age two under the permutation feature importance test (PermFIT) framework by using the publicly available data set from the Pediatric Emergency Care Applied Research Network (PECARN) study. The prediction accuracy was determined by comparing the predicted TBI status with the computed tomography (CT) scan results since CT scan is the gold standard for diagnosing TBI.

RESULTS

At a significance level of [Formula: see text], DNN, RF, XGB, and SVM identified 9, 1, 2,  and 4 significant features, respectively. In a comparison of accuracy (Accuracy), the area under the curve (AUC), and the precision-recall area under the curve (PR-AUC), the permutation feature importance test for DNN model was the most powerful framework for identifying significant features and outperformed other methods, i.e., RF, XGB, and SVM, with Accuracy, AUC, and PR-AUC as 0.915, 0.794, and 0.974, respectively.

CONCLUSION

These results indicate that the PermFIT-DNN framework robustly identifies significant clinical features associated with TBI status and improves prediction performance. The findings could be used to inform the development of clinical decision tools designed to inform triage decisions.

Factors Predicting Quality of Life Impairment in Adult Patients with Atopic Dermatitis: Results from a Patient Survey and Machine Learning Analysis

INTRODUCTION

Atopic dermatitis (AD) is a chronic, inflammatory skin disorder that impairs patients' quality of life (QoL). Physician assessment of AD disease severity is determined by clinical scales and assessment of affected body surface area (BSA), which might not mirror patients' perceived disease burden.

METHODS

Using data from an international cross-sectional web-based survey of patients with AD and a machine learning approach, we sought to identify disease attributes with the highest impact on QoL for patients with AD. Adults with dermatologist-confirmed AD participated in the survey between July-September 2019. Eight machine learning models were applied to the data with dichotomised Dermatology Life Quality Index (DLQI) as the response variable to identify factors most predictive of AD-related QoL burden. Variables tested were demographics, affected BSA and affected body areas, flare characteristics, activity impairment, hospitalisation and AD therapies. Three machine learning models, logistic regression model, random forest and neural network, were selected on the basis of predictive performance. Each variable's contribution was computed via importance values from 0 to 100. For relevant predictive factors, further descriptive analyses were conducted to characterise those findings.

RESULTS

In total, 2314 patients completed the survey with mean age 39.2 years (standard deviation 12.6) and average disease duration of 19 years. Measured by affected BSA, 13.3% of patients had moderate-to-severe disease. However, 44% of patients reported a DLQI > 10, indicative of a very large to extremely large impact on QoL. Activity impairment was the most important factor predicting high QoL burden (DLQI > 10) across models. Hospitalisation during the past year and flare type were also highly ranked. Current BSA involvement was not a strong predictor of AD-related QoL impairment.

CONCLUSIONS

Activity impairment was the single most important factor for AD-related QoL impairment while current extent of AD did not predict higher disease burden. These results support the importance of considering patients' perspectives when determining the severity of AD.

Cross-sectional analysis reveals autoantibody signatures associated with COVID-19 severity

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection is associated with increased levels of autoantibodies targeting immunological proteins such as cytokines and chemokines. Reports further indicate that COVID-19 patients may develop a broad spectrum of autoimmune diseases due to reasons not fully understood. Even so, the landscape of autoantibodies induced by SARS-CoV-2 infection remains uncharted territory. To gain more insight, we carried out a comprehensive assessment of autoantibodies known to be linked to diverse autoimmune diseases observed in COVID-19 patients in a cohort of 231 individuals, of which 161 were COVID-19 patients (72 with mild, 61 moderate, and 28 with severe disease) and 70 were healthy controls. Dysregulated IgG and IgA autoantibody signatures, characterized mainly by elevated concentrations, occurred predominantly in patients with moderate or severe COVID-19 infection. Autoantibody levels often accompanied anti-SARS-CoV-2 antibody concentrations while stratifying COVID-19 severity as indicated by random forest and principal component analyses. Furthermore, while young versus elderly COVID-19 patients showed only slight differences in autoantibody levels, elderly patients with severe disease presented higher IgG autoantibody concentrations than young individuals with severe COVID-19. This work maps the intersection of COVID-19 and autoimmunity by demonstrating the dysregulation of multiple autoantibodies triggered during SARS-CoV-2 infection. Thus, this cross-sectional study suggests that SARS-CoV-2 infection induces autoantibody signatures associated with COVID-19 severity and several autoantibodies that can be used as biomarkers of COVID-19 severity, indicating autoantibodies as potential therapeutical targets for these patients.

Risk Factors for Pediatric Sepsis in the Emergency Department: A Machine Learning Pilot Study

OBJECTIVE

To identify underappreciated sepsis risk factors among children presenting to a pediatric emergency department (ED).

METHODS

A retrospective observational study (2017-2019) of children aged 18 years and younger presenting to a pediatric ED at a tertiary care children's hospital with fever, hypotension, or an infectious disease International Classification of Diseases (ICD)-10 diagnosis. Structured patient data including demographics, problem list, and vital signs were extracted for 35,074 qualifying ED encounters. According to the Improving Pediatric Sepsis Outcomes Classification, confirmed by expert review, 191 patients met clinical sepsis criteria. Five machine learning models were trained to predict sepsis/nonsepsis outcomes. Top features enabling model performance (N = 20) were then extracted to identify patient risk factors.

RESULTS

Machine learning methods reached a performance of up to 93% sensitivity and 84% specificity in identifying patients who received a hospital diagnosis of sepsis. A random forest classifier performed the best, followed by a classification and regression tree. Maximum documented heart rate was the top feature in these models, with importance coefficients (ICs) of 0.09 and 0.21, which represent how much an individual feature contributes to the model. Maximum mean arterial pressure was the second most important feature (IC 0.05, 0.13). Immunization status (IC 0.02), age (IC 0.03), and patient zip code (IC 0.02) were also among the top features enabling models to predict sepsis from ED visit data. Stratified analysis revealed changes in the predictive importance of risk factors by race, ethnicity, oncologic history, and insurance status.

CONCLUSIONS

Machine learning models trained to identify pediatric sepsis using ED clinical and sociodemographic variables confirmed well-established predictors, including heart rate and mean arterial pressure, and identified underappreciated relationships between sepsis and patient age, immunization status, and demographics.

Machine Learning-Based Prediction of Hospital Admission Among Children in an Emergency Care Center

OBJECTIVES

Machine learning-based prediction of hospital admissions may have the potential to optimize patient disposition and improve clinical outcomes by minimizing both undertriage and overtriage in crowded emergency care. We developed and validated the predictive abilities of machine learning-based predictions of hospital admissions in a pediatric emergency care center.

METHODS

A prognostic study was performed using retrospectively collected data of children younger than 16 years who visited a single pediatric emergency care center in Osaka, Japan, between August 1, 2016, and October 15, 2019. Generally, the center treated walk-in children and did not treat trauma injuries. The main outcome was hospital admission as determined by the physician. The 83 potential predictors available at presentation were selected from the following categories: demographic characteristics, triage level, physiological parameters, and symptoms. To identify predictive abilities for hospital admission, maximize the area under the precision-recall curve, and address imbalanced outcome classes, we developed the following models for the preperiod training cohort (67% of the samples) and also used them in the 1-year postperiod validation cohort (33% of the samples): (1) logistic regression, (2) support vector machine, (3) random forest, and (4) extreme gradient boosting.

RESULTS

Among 88,283 children who were enrolled, the median age was 3.9 years, with 47,931 (54.3%) boys and 1985 (2.2%) requiring hospital admission. Among the models, extreme gradient boosting achieved the highest predictive abilities (eg, area under the precision-recall curve, 0.26; 95% confidence interval, 0.25-0.27; area under the receiver operating characteristic curve, 0.86; 95% confidence interval, 0.84-0.88; sensitivity, 0.77; and specificity, 0.82). With an optimal threshold, the positive and negative likelihood ratios were 4.22, and 0.28, respectively.

CONCLUSIONS

Machine learning-based prediction of hospital admissions may support physicians' decision-making for hospital admissions. However, further improvements are required before implementing these models in real clinical settings.

Clinical narrative-aware deep neural network for emergency department critical outcome prediction

Since early identification of potential critical patients in the Emergency Department (ED) can lower mortality and morbidity, this study seeks to develop a machine learning model capable of predicting possible critical outcomes based on the history and vital signs routinely collected at triage. We compare emergency physicians and the predictive performance of the machine learning model. Predictors including patients' chief complaints, present illness, past medical history, vital signs, and demographic data of adult patients (aged ≥ 18 years) visiting the ED at Shuang-Ho Hospital in New Taipei City, Taiwan, are extracted from the hospital's electronic health records. Critical outcomes are defined as in-hospital cardiac arrest (IHCA) or intensive care unit (ICU) admission. A clinical narrative-aware deep neural network was developed to handle the text-intensive data and standardized numerical data, which is compared against other machine learning models. After this, emergency physicians were asked to predict possible clinical outcomes of thirty visits that were extracted randomly from our dataset, and their results were further compared to our machine learning model. A total of 4,308 (2.5 %) out of the 171,275 adult visits to the ED included in this study resulted in critical outcomes. The area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC) of our proposed prediction model is 0.874 and 0.207, respectively, which not only outperforms the other machine learning models, but even has better sensitivity (0.95 vs 0.41) and accuracy (0.90 vs 0.67) as compared to the emergency physicians. This model is sensitive and accurate in predicting critical outcomes and highlights the potential to use predictive analytics to support post-triage decision-making.

Elicitation of domain knowledge for a machine learning model for paediatric critical illness in South Africa

Objectives

Delays in identification, resuscitation and referral have been identified as a preventable cause of avoidable severity of illness and mortality in South African children. To address this problem, a machine learning model to predict a compound outcome of death prior to discharge from hospital and/or admission to the PICU was developed. A key aspect of developing machine learning models is the integration of human knowledge in their development. The objective of this study is to describe how this domain knowledge was elicited, including the use of a documented literature search and Delphi procedure.

Design

A prospective mixed methodology development study was conducted that included qualitative aspects in the elicitation of domain knowledge, together with descriptive and analytical quantitative and machine learning methodologies.

Setting

A single centre tertiary hospital providing acute paediatric services.

Participants

Three paediatric intensivists, six specialist paediatricians and three specialist anaesthesiologists.

Interventions

None.

Measurements and main results

The literature search identified 154 full-text articles reporting risk factors for mortality in hospitalised children. These factors were most commonly features of specific organ dysfunction. 89 of these publications studied children in lower- and middle-income countries. The Delphi procedure included 12 expert participants and was conducted over 3 rounds. Respondents identified a need to achieve a compromise between model performance, comprehensiveness and veracity and practicality of use. Participants achieved consensus on a range of clinical features associated with severe illness in children. No special investigations were considered for inclusion in the model except point-of-care capillary blood glucose testing. The results were integrated by the researcher and a final list of features was compiled.

Conclusion

The elicitation of domain knowledge is important in effective machine learning applications. The documentation of this process enhances rigour in such models and should be reported in publications. A documented literature search, Delphi procedure and the integration of the domain knowledge of the researchers contributed to problem specification and selection of features prior to feature engineering, pre-processing and model development.

Data harnessing to nurture the human mind for a tailored approach to the child

Big data in pediatrics is an ocean of structured and unstructured data. Big data analysis helps to dive into the ocean of data to filter out information that can guide pediatricians in their decision making, precision diagnosis, and targeted therapy. In addition, big data and its analysis have helped in the surveillance, prevention, and performance of the health system. There has been a considerable amount of work in pediatrics that we have tried to highlight in this review and some of it has been already incorporated into the health system. Work in specialties of pediatrics is still forthcoming with the creation of a common data model and amalgamation of the huge "omics" database. The physicians entrusted with the care of children must be aware of the outcome so that they can play a role to ensure that big data algorithms have a clinically relevant effect in improving the health of their patients. They will apply the outcome of big data and its analysis in patient care through clinical algorithms or with the help of embedded clinical support alerts from the electronic medical records. IMPACT: Big data in pediatrics include structured, unstructured data, waveform data, biological, and social data. Big data analytics has unraveled significant information from these databases. This is changing how pediatricians will look at the body of available evidence and translate it into their clinical practice. Data harnessed so far is implemented in certain fields while in others it is in the process of development to become a clinical adjunct to the physician. Common databases are being prepared for future work. Diagnostic and prediction models when incorporated into the health system will guide the pediatrician to a targeted approach to diagnosis and therapy.

Exploring the effectiveness of artificial intelligence, machine learning and deep learning in trauma triage: A systematic review and meta-analysis

Background

The development of artificial intelligence (AI), machine learning (ML) and deep learning (DL) has advanced rapidly in the medical field, notably in trauma medicine. We aimed to systematically appraise the efficacy of AI, ML and DL models for predicting outcomes in trauma triage compared to conventional triage tools.

Methods

We searched PubMed, MEDLINE, ProQuest, Embase and reference lists for studies published from 1 January 2010 to 9 June 2022. We included studies which analysed the use of AI, ML and DL models for trauma triage in human subjects. Reviews and AI/ML/DL models used for other purposes such as teaching, or diagnosis were excluded. Data was extracted on AI/ML/DL model type, comparison tools, primary outcomes and secondary outcomes. We performed meta-analysis on studies reporting our main outcomes of mortality, hospitalisation and critical care admission.

Results

One hundred and fourteen studies were identified in our search, of which 14 studies were included in the systematic review and 10 were included in the meta-analysis. All studies performed external validation. The best-performing AI/ML/DL models outperformed conventional trauma triage tools for all outcomes in all studies except two. For mortality, the mean area under the receiver operating characteristic (AUROC) score difference between AI/ML/DL models and conventional trauma triage was 0.09, 95% CI (0.02, 0.15), favouring AI/ML/DL models (p = 0.008). The mean AUROC score difference for hospitalisation was 0.11, 95% CI (0.10, 0.13), favouring AI/ML/DL models (p = 0.0001). For critical care admission, the mean AUROC score difference was 0.09, 95% CI (0.08, 0.10) favouring AI/ML/DL models (p = 0.00001).

Conclusions

This review demonstrates that the predictive ability of AI/ML/DL models is significantly better than conventional trauma triage tools for outcomes of mortality, hospitalisation and critical care admission. However, further research and in particular randomised controlled trials are required to evaluate the clinical and economic impacts of using AI/ML/DL models in trauma medicine.

Echocardiographic evaluation of supracardiac anomalous pulmonary venous connection in children: comparison with multilayer spiral CT

Objective To explore the clinical value of transthoracic echocardiography (TTE) in the differentiation of Supracardiac Anomalous Pulmonary Venous Connection (SAPVC) in children. Materials and methods A total of 118 children with concurrent TTE and CT databases of cases diagnosed with SAPVCs were included. We analyzed the consistency between the two for the ability to diagnose the classification of SAPVC, drainage sites, ectopic pulmonary veins and the segments of superior vena cava (SVC). Results The consistency between TTE and CT in diagnosing the existence of SAPVC and the classification were 88.1% (95% CI: 80.9-93.4%) and 91.0% (95% CI: 84.1-95.6%), respectively. The error rate of partial type diagnosed by TTE was significantly higher than that of total and mixed type (20.5% vs. 2.8%, P = 0.003). The consistency between TTE and CT to determine drainage sites was 91.9% (95% CI: 85.2-96.2%). TTE had a significantly higher error rate in determining pulmonary vein drainage to the SVC than in those draining into the left innominate vein (17.5 vs. 2.5%, P = 0.007). The consistency of TTE and CT in judging the number of veins was 87.4% (95% CI: 79.7-92.9%). The error rate in determining the presence of 2 and 5 ectopic pulmonary veins was significantly higher than those of 1 and 4 veins (P < 0.05). Conclusion TTE for diagnosing partial SAPVC and identifying the drainage site of SVC has a high error rate of misdiagnosis and missed diagnosis. The extra attention should be given to these factors in clinical practice to improve the accuracy of TTE in diagnosing SAPVC.

Predicting recurrent cases of intussusception in children after air enema reduction with machine learning models

PURPOSE

To develop a model to identify risk factors and predict recurrent cases of intussusception in children.

METHODS

Consecutive cases and recurrent cases of intussusception in children from January 2016 to April 2022 were screened. The cohort was divided randomly at a 4:1 ratio to a training dataset and a validation dataset. Three parallel models were developed using extreme gradient boosting (XGBoost), logistic regression (LR), and support vector machine (SVM). Model performance was assessed by the area under the receiver operating characteristic curves (AUC).

RESULTS

A total of 2469 cases of intussusception were included, where 225 were recurrent cases. The XGBoost (AUC = 0.718) models showed the best performance in the validation dataset, followed by the LR model (AUC = 0.652), while the SVM model (AUC = 0.613) performed worst among the three models. Based on the Shapley Additive exPlanation values, the most important variables in the XGBoost models were air enema pressure, mass size, age, duration of symptoms, and absence of vomiting.

CONCLUSIONS

Machine learning models, especially XGBoost, could be used to predict recurrent cases of intussusception in children. The most important contributing factors to the models are air enema pressure, mass size, age, duration of symptoms and absence of vomiting.

Prediction of postoperative cardiopulmonary complications after lung resection in a Chinese population: A machine learning-based study

Background

Approximately 20% of patients with lung cancer would experience postoperative cardiopulmonary complications after anatomic lung resection. Current prediction models for postoperative complications were not suitable for Chinese patients. This study aimed to develop and validate novel prediction models based on machine learning algorithms in a Chinese population.

Methods

Patients with lung cancer receiving anatomic lung resection and no neoadjuvant therapies from September 1, 2018 to August 31, 2019 were enrolled. The dataset was split into two cohorts at a 7:3 ratio. The logistic regression, random forest, and extreme gradient boosting were applied to construct models in the derivation cohort with 5-fold cross validation. The validation cohort accessed the model performance. The area under the curves measured the model discrimination, while the Spiegelhalter z test evaluated the model calibration.

Results

A total of 1085 patients were included, and 760 were assigned to the derivation cohort. 8.4% and 8.0% of patients experienced postoperative cardiopulmonary complications in the two cohorts. All baseline characteristics were balanced. The values of the area under the curve were 0.728, 0.721, and 0.767 for the logistic, random forest and extreme gradient boosting models, respectively. No significant differences existed among them. They all showed good calibration (p > 0.05). The logistic model consisted of male, arrhythmia, cerebrovascular disease, the percentage of predicted postoperative forced expiratory volume in one second, and the ratio of forced expiratory volume in one second to forced vital capacity. The last two variables, the percentage of forced vital capacity and age ranked in the top five important variables for novel machine learning models. A nomogram was plotted for the logistic model.

Conclusion

Three models were developed and validated for predicting postoperative cardiopulmonary complications among Chinese patients with lung cancer. They all exerted good discrimination and calibration. The percentage of predicted postoperative forced expiratory volume in one second and the ratio of forced expiratory volume in one second to forced vital capacity might be the most important variables. Further validation in different scenarios is still warranted.

Association between hospital legal constructions and medical disputes: A multi-center analysis of 130 tertiary hospitals in Hunan Province, China

Background

Medical disputes are common in hospitals and a major challenge for the operations of medical institutions. However, few studies have looked into the association between medical disputes and hospital legal constructions. The purpose of the study was to investigate the relationship between hospital legal constructions and medical disputes, and it also aimed to develop a nomogram to estimate the likelihood of medical disputes.

Methods

Between July and September 2021, 2,716 administrators from 130 hospitals were enrolled for analysis. The study collected seventeen variables for examination. To establish a nomogram, administrators were randomly split into a training group (n = 1,358) and a validation group (n = 1,358) with a 50:50 ratio. The nomogram was developed using data from participants in the training group, and it was validated in the validation group. The nomogram contained significant variables that were linked to medical disputes and were identified by multivariate analysis. The nomogram's predictive performance was assessed utilizing discriminative and calibrating ability. A web calculator was developed to be conducive to model utility.

Results

Medical disputes were observed in 41.53% (1,128/2,716) of participants. Five characteristics, including male gender, higher professional ranks, longer length of service, worse understanding of the hospital charters, and worse construction status of hospital rule of law, were significantly associated with more medical disputes based on the multivariate analysis. As a result, these variables were included in the nomogram development. The AUROC was 0.67 [95% confident interval (CI): 0.64-0.70] in the training group and 0.68 (95% CI: 0.66-0.71) in the validation group. The corresponding calibration slopes were 1.00 and 1.05, respectively, and intercepts were 0.00 and -0.06, respectively. Three risk groups were created among the participants: Those in the high-risk group experienced medical disputes 2.83 times more frequently than those in the low-risk group (P < 0.001).

Conclusion

Medical dispute is prevailing among hospital administrators, and it can be reduced by the effective constructions of hospital rule of law. This study proposes a novel nomogram to estimate the likelihood of medical disputes specifically among administrators in tertiary hospitals, and a web calculator can be available at https://ymgarden.shinyapps.io/Predictionofmedicaldisputes/.

Effect of Applying a Real-Time Medical Record Input Assistance System With Voice Artificial Intelligence on Triage Task Performance in the Emergency Department: Prospective Interventional Study

BACKGROUND

Natural language processing has been established as an important tool when using unstructured text data; however, most studies in the medical field have been limited to a retrospective analysis of text entered manually by humans. Little research has focused on applying natural language processing to the conversion of raw voice data generated in the clinical field into text using speech-to-text algorithms.

OBJECTIVE

In this study, we investigated the promptness and reliability of a real-time medical record input assistance system with voice artificial intelligence (RMIS-AI) and compared it to the manual method for triage tasks in the emergency department.

METHODS

From June 4, 2021, to September 12, 2021, RMIS-AI, using a machine learning engine trained with 1717 triage cases over 6 months, was prospectively applied in clinical practice in a triage unit. We analyzed a total of 1063 triage tasks performed by 19 triage nurses who agreed to participate. The primary outcome was the time for participants to perform the triage task.

RESULTS

The median time for participants to perform the triage task was 204 (IQR 155, 277) seconds by RMIS-AI and 231 (IQR 180, 313) seconds using manual method; this difference was statistically significant (P<.001). Most variables required for entry in the triage note showed a higher record completion rate by the manual method, but in the recording of additional chief concerns and past medical history, RMIS-AI showed a higher record completion rate than the manual method. Categorical variables entered by RMIS-AI showed less accuracy compared with continuous variables, such as vital signs.

CONCLUSIONS

RMIS-AI improves the promptness in performing triage tasks as compared to using the manual input method. However, to make it a reliable alternative to the conventional method, technical supplementation and additional research should be pursued.

A Machine Learning Algorithm to Estimate the Probability of a True Scaphoid Fracture After Wrist Trauma

PURPOSE

To identify predictors of a true scaphoid fracture among patients with radial wrist pain following acute trauma, train 5 machine learning (ML) algorithms in predicting scaphoid fracture probability, and design a decision rule to initiate advanced imaging in high-risk patients.

METHODS

Two prospective cohorts including 422 patients with radial wrist pain following wrist trauma were combined. There were 117 scaphoid fractures (28%) confirmed on computed tomography, magnetic resonance imaging, or radiographs. Eighteen fractures (15%) were occult. Predictors of a scaphoid fracture were identified among demographics, mechanism of injury and examination maneuvers. Five ML-algorithms were trained in calculating scaphoid fracture probability. ML-algorithms were assessed on ability to discriminate between patients with and without a fracture (area under the receiver operating characteristic curve), agreement between observed and predicted probabilities (calibration), and overall performance (Brier score). The best performing ML-algorithm was incorporated into a probability calculator. A decision rule was proposed to initiate advanced imaging among patients with negative radiographs.

RESULTS

Pain over the scaphoid on ulnar deviation, sex, age, and mechanism of injury were most strongly associated with a true scaphoid fracture. The best performing ML-algorithm yielded an area under the receiver operating characteristic curve, calibration slope, intercept, and Brier score of 0.77, 0.84, -0.01 and 0.159, respectively. The ML-derived decision rule proposes to initiate advanced imaging in patients with radial-sided wrist pain, negative radiographs, and a fracture probability of ≥10%. When applied to our cohort, this would yield 100% sensitivity, 38% specificity, and would have reduced the number of patients undergoing advanced imaging by 36% without missing a fracture.

CONCLUSIONS

The ML-algorithm accurately calculated scaphoid fracture probability based on scaphoid pain on ulnar deviation, sex, age, and mechanism of injury. The ML-decision rule may reduce the number of patients undergoing advanced imaging by a third with a small risk of missing a fracture. External validation is required before implementation.

TYPE OF STUDY/LEVEL OF EVIDENCE

Diagnostic II.

Development of a low-dimensional model to predict admissions from triage at a pediatric emergency department

Objectives

This study aims to develop and internally validate a low-dimensional model to predict outcomes (admission or discharge) using commonly entered data up to the post-triage process to improve patient flow in the pediatric emergency department (ED). In hospital settings where electronic data are limited, a low-dimensional model with fewer variables may be easier to implement.

Methods

This prognostic study included ED attendances in 2017 and 2018. The Cross Industry Standard Process for Data Mining methodology was followed. Eligibility criteria was applied to the data set, splitting into 70% train and 30% test. Sampling techniques were compared. Gradient boosting machine (GBM), logistic regression, and naïve Bayes models were created. Variables of importance were obtained from the model with the highest area under the curve (AUC) and used to create a low-dimensional model.

Results

Eligible attendances totaled 72,229 (15% admission rate). The AUC was 0.853 (95% confidence interval [CI], 0.846-0.859) for GBM, 0.845 (95% CI, 0.838-0.852) for logistic regression and 0.813 (95% CI, 0.806-0.821) for naïve Bayes. Important predictors in the GBM model used to create a low-dimensional model were presenting complaint, triage category, referral source, registration month, location type (resuscitation/other), distance traveled, admission history, and weekday (AUC 0.835 [95% CI, 0.829-0.842]).

Conclusions

Admission and discharge probability can be predicted early in a pediatric ED using 8 variables. Future work could analyze the false positives and false negatives to gain an understanding of the implementation of these predictions.