Research Progress of Machine Learning in Emergency Medicine During 2014-2023: A Bibliometric Analysis

Asploro Journal of Biomedical and Clinical Case Reports

ISSN: 2582-0370
Article Type: Mini Review
DOI: 10.36502/2025/ASJBCCR.6399
Asp Biomed Clin Case Rep. 2025 May 27;8(2):105-17

View PDF

Xiaoyan Xian¹, Dan Zhu², Shuyun Xu^3*
¹Department of Emergency Medicine, West China Hospital, Sichuan University, Chengdu, 610041, Sichuan, China

Corresponding Author: Shuyun Xu
Address: Department of Emergency Medicine, West China Hospital, Sichuan University, No.37 Guoxue Lane, Wuhou district, Chengdu 610041, China.
Received date: 07 May 2025; Accepted date: 20 May 2025; Published date: 27 May 2025

Citation: Xian X, Zhu D, Xu S. Research Progress of Machine Learning in Emergency Medicine During 2014-2023: A Bibliometric Analysis. Asp Biomed Clin Case Rep. 2025 May 27;8(2):105-17.

Copyright © 2025 Xian X, Zhu D, Xu S. This is an open-access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium provided the original work is properly cited.

Keywords: Machine Learning, Emergency, Bibliometrics

Abbreviations: WOS: Web of Science; SCI-E: Science Citation Index-Expanded; RRI: Relative Research Interest; IF: Impact Factor; RCTs: Randomized Clinical Trials; AAY: Average Appearing Years; ICU: Intensive Care Unit; ED: Emergency Department; OHCA: Out-of-Hospital Cardiac Arrest; ROSC: Return of Spontaneous Circulation; LR: Logical Regression; XGB: Xgboost

Abstract

Background: Machine learning, as an important branch of artificial intelligence, is more and more widely used in emergency medicine, including triage, disease diagnosis, treatment, prognosis prediction, and more. By analyzing the publication status of the literature related to machine learning in emergency medicine through bibliometric analysis, our purpose was to enlighten trends and hotspots for the future development of machine learning in emergency medicine.
Method: Two researchers retrieved and screened all the literature related to machine learning in emergency medicine from Jan. 2014 to Dec. 2023 on the Web of Science. Excel, VOSviewer, CiteSpace, and the Online Analysis Platform of Literature Metrology were applied to visualize the research trends and study the co-occurring keywords in machine learning in emergency medicine.
Results: 536 publications with a total citation of 5,181 times were identified. The global articles have increased over the past decade, and developed countries contribute the most. The United States contributed the most number of publications (34.7%), the highest number of citations (2,412), and the H-index (12.97). The number of publications from China ranked second, and the citations were 327 times with an H-index of 11, which ranked sixth and tied for second respectively. The American Journal of Emergency Medicine is the journal published most in machine learning in emergency medicine. In the identification research cluster, “mortality” and “risk” are determined to be the hotspot, while “emergency medical services” and “out-of-hospital cardiac arrest (OHCA)” are the new trend in machine learning in emergency medicine.
Conclusion: In the past decade, research on machine learning in emergency medicine has increased gradually and will increase rapidly in the next decade. The United States has an absolute advantage and takes the leading position in this field. The quantity and quality of Chinese articles are inconsistent. In the identification research cluster, “mortality” and “risk” are determined to be the hotspot, while “emergency medical services”, “out-of-hospital cardiac arrest (OHCA)” and “emergency medical services” are the new trend in machine learning in emergency medicine.

Introduction

With the high-speed development of emergency medicine, there is an increasing demand for emergency services, such as rapid and effective triage, early disease prediction, accurate disease diagnosis, and timely therapeutic interventions. Traditional techniques may not be sufficient [1]. Machine learning (ML) is a data analysis method that makes analytical model building automated, based on the theory that systems can identify patterns, learn from data, and make decisions with minimal or no human intervention [2]. Machine learning techniques usually have comparable and even better accuracy than the hospital’s medical staff. It has the potential to capture complex, non-linear, and often subtle relationships between covariates and outcomes or among characteristics that cluster patients or other types of observations into sub-groups that would not otherwise be readily apparent [3]. Moreover, it will also reduce human errors as well as time and expenses and improve the pace of providing services [1]. ML methods can be divided into supervised learning, unsupervised learning, and reinforcement learning.

Bibliometric analysis is a statistical method that attempts to assess articles by their citations, analyzing their frequency and citation pattern, which subsequently gleans direction and guidance for future research [4]. Now the methods of bibliometrics have been widely used in many vital fields, including medicine. Based on the Web of Science (WOS), our article aims to use the methodology of bibliometric analysis to analyze the research trend related to machine learning in emergency medicine and predict its possible research frontiers and hotspots in the future.

Materials and Methods

Data Sources and Search Strategies:

Thomson Reuters’ Web of Science (WoS) is one of the largest and most comprehensive academic information resources globally. Therefore, we chose the Science Citation Index-Expanded (SCI-E) of Thomson Reuters’ Web of Science as an optimal database for bibliometric analysis. All searches were completed on January 8, 2024, to avoid bias in the number of publications and citations due to rapid database renewal. The retrieval strategies are presented as follows:

TS = ((“Machine Learning” OR “neural network” OR “Deep Learning” OR “Computer-aided” OR “Prediction Model” OR “Data Mining” OR “Cross* validation” OR “Regularized logistic” OR “Linear discriminant analysis” OR “LDA” OR “Random forest” OR “Naïve Bayes” OR “Least Absolute selection shrinkage operator” OR “LASSO” OR “elastic net” OR “RVM” OR “relevance vector machine” OR “pattern recognition” OR “Computational Intelligence” OR “Machine SAME Intelligence” OR “Knowledge Representation” OR “support vector” OR “SVM” OR “Pattern classification” OR “Decision Tree” OR “Nearest neighbor” OR “Classification Tree” OR “Regression Tree” OR “Iterative Learning” OR “recursive partitioning” OR “causal forest” OR “generalized random forest” OR “X-learner” OR “R-learner” OR “S-learner” OR “Neural Networks, Computer” OR “Artificial Intelligence” OR “Intelligence, Computational” OR “Computer SAME Reasoning” OR “Knowledge SAME Acquisition (Computer)” OR “Knowledge SAME Representations (Computer)” OR “Support Vector Machine” OR “Decision Trees”)) AND SU = (“Emergency Medicine” OR “Medicine, Emergency”) AND Language = English.

We have only selected standard peer-reviewed articles and reviews, excluding other types of studies. Refer to Fig-1 for the detailed retrieval process.

Fig-1: Flow Diagram of the Inclusion Process

Data Collation:

All the data were downloaded from the Web of Science (WOS) and extracted carefully and independently by two reviewers (XXY and ZD), including titles, authors, origin countries and regions, institutions, keywords, publication dates, published journals, the sum of citations, and H-index. Microsoft Excel 2013 (Redmond, Washington, USA), VOSviewer version 1.6.20 (Leiden University, Leiden, the Netherlands), CiteSpace version 6.2.R6 (Drexel University, Philadelphia, PA, USA), and the Online Analysis Platform of Literature Metrology (https://bibliometric.com/) were used for presenting, analyzing data, and describing figures.

Bibliometric Analysis:

A large amount of research was collected by the Web of Science, especially that focused on biomedicine; therefore, we chose WOS to characterize all eligible publications. Relative research interest (RRI) was defined as the number of publications in a particular research field divided by all-field publications per year. The impact factor (IF) of all journals was demonstrated by inquiring about the Journal Citation Reports (JCRs) published in 2023. It has been universally acknowledged that H-index serves as the scientific research impact of a scholar or a country, reflecting both the number of publications and the number of citations per publication. The H-index refers to H papers published by a scholar/country, each of which has been cited at least H times in other publications. The H-index can provide a robust assessment of the impact of long-term cumulative research outputs by researchers in predicting future scientific achievement.

Additionally, VOSviewer and CiteSpace were applied to visualize the data we collected. VOSviewer is an optimal approach for creating various bibliometrics- based maps, author or journal co-citation diagrams, and keyword co-occurrence diagrams. We used VOSviewer to classify keywords into different clusters based on co-occurrence analysis and color them by time course. Different colors were applied to reflect the relative novelty of keywords. Through the application of clustering and burst word detection of CiteSpace, the hidden patterns and laws of knowledge structures in particular disciplines and fields were explored and mined.

Results

Distribution of Countries of Global Publications:

The contributions of different countries and regions were assessed from the quantity and quality of publications. A total of 536 articles from 2014 to 2023 met our inclusion criteria. The United States ranked first in the number of publications at 186 (34.70%), followed by China at 71 (13.25%) and Australia at 45 (8.40%). Analyzing the number of papers per year showed that publications were the most within 2021 (109, 20.34%) (Fig-2). In addition, we found that global interest in machine learning in emergency medicine increased rapidly from 2019, and the trend was overall upward. The number of publications on the application of machine learning in emergency medicine has been growing exponentially since 2021. Considering the number of all-field publications, the global attention towards this field measured by RRI value fluctuated around 0.005% before 2011, while subsequently going up and reaching 0.020% in 2021 (Fig-2). The United States had an absolute advantage in quantity and the number of citations, which was more than the sum of citations in other places. Chinese scholars began to publish articles in this field in 2018, but did not produce any articles from 2014 to 2017. The quantity of literature published by Chinese scholars remained relatively stable between 2021 and 2023.

Fig-2: Research Contributions of Different Countries/Regions to The Areas Concerning Machine Learning in Emergency Medicine

Citations and H-index Analysis:

The overall number of publications related to machine learning in emergency medicine cited is 5181 since 2014. The average citing frequency per publication was 9.67 times. The United States had the most citations with the number of 2412 (46.55%), an average cited frequency per paper of 12.97, and an H-index of 22. Australia ranked second with a citation number of 403, an average cited frequency per paper of 8.96, and an H-index of 11. Though the number of publications in China ranked second to the United States, the citations were only 327 times with an H-index of 11, which ranked sixth and tied for second, respectively (Fig-3). The number of average cited frequency per paper in China only ranked fourteenth.

Fig-3: Research Contributions of Different Countries/Regions to The Areas Concerning Machine Learning in Emergency Medicine

Journals with Research Publications:

More than three-quarters of the publications in this field were published in the top 15 journals (463, 86.38%). The American Journal of Emergency Medicine (IF = 3.6, 2022) had the highest number of publications with a total of 77, thereby accounting for 14.37% of all published literature on machine learning in emergency medicine. Besides, Resuscitation (IF = 6.5, 2022) ranked second with 68 (12.69%) publications. The third magazine, Injury – International Journal of the Care of the Injured (IF = 2.5, 2022), published 53 articles. Academic Emergency Medicine (IF = 3.451, 2020) has 33 papers, which ranked fourth. There were 30 articles in the Scandinavian Journal of Trauma Resuscitation & Emergency Medicine (IF = 3.3, 2022) and 20 articles in the European Journal of Trauma and Emergency Surgery (IF = 2.1, 2022) in this field. Noticeably, the IF factor of Resuscitation (IF = 6.5, 2022) ranked first in the top 15 journals, and the H-index of which ranked first in the top 15 journals (16). The top fifteen journals that published the most papers were listed in Fig-4A.

Institutions with Research Publications:

There were 1199 institutions involved in related machine learning in emergency medicine literature. Publications from the top 15 institutes accounted for 43.85% of all literature in this field. The University of California system had the highest number of publications with a total of 26 institutions worldwide, accounting for 4.85% of all published literature in this field. Harvard University ranked second (25, 4.67%) and Pennsylvania Commonwealth System of Higher Education (PCSHE) ranked third (20, 3.73%). Within the top 15 institutes, 12 are from the United States, one is from South Korea, one is from Canada, and one is from Australia. The top fifteen institutions that published the most papers were listed in Fig-4B.

Fig-4: Distribution of Journals and Institutions Focusing On Machine Learning in Emergency Medicine

Authors with Research Publications:

A total of 66 publications were from the top 10 authors, accounting for 12.31% of all publications related to the field. The authors with the most published articles are Park JH and Shin SD. Park JH, from the Korea Advanced Institute of Science & Technology, published eight articles with a total citation count of 30 and an average of 3.75 citations per article. Shin SD, from Seoul National University, also published eight articles, with a total of 23 citations and an average of 2.88 citations per article.

Herlitz J, Kim KH, Lee JH, and Taylor RA each had seven publications. Notably, Taylor RA, from Yale University, is the author with the highest citation frequency. Although Taylor RA has only seven publications, they received a total of 496 citations, with an average of 70.86 citations per article.

As shown in Table-1, six of the top 10 authors were from South Korea, two from the United States, one from Australia, and one from Sweden.

Table-1: Top 10 Authors with the Most Publications in the Research Scope of Machine Learning in Emergency Medicine

Analysis of Cited References:

The visualization map of cited references included 370 nodes and 845 links. Among these, the top 10 most frequently cited references are shown in Fig-5A.

The most cited publication was published in Critical Care (Impact Factor: 15.1), titled “Emergency department triage prediction of clinical outcomes using machine learning models” by Raita Y, with a total of 14 citations [5]. The second most cited publication was “The Third International Consensus Definitions for Sepsis and Septic Shock”, which received 13 citations [6]. The third most cited article was titled “A Systematic Review Shows No Benefit of Machine Learning on the Performance of Clinical Prediction Models Compared to Logistic Regression”, with a total of 10 citations [7].

In addition, a timeline view of co-cited reference clusters related to machine learning in emergency medicine was generated using CiteSpace (Fig-5B). The log-likelihood rate (LLR) was applied to identify the distribution of hotspots among the ten clusters.

Specifically, clusters with warmer colors and larger nodes represented more recent publications, indicating current hotspots in the field. As shown in Fig-5B, cluster #1 (rapid assessment) and cluster #2 (in-patient admission) have emerged as ongoing hotspots in recent years. Meanwhile, cluster #0 (septic shock), cluster #3 (out-of-hospital cardiac arrest), cluster #4 (intensive care unit), and cluster #6 (quick COVID-19 severity index) reflect areas that were major hotspots in the recent past.

Fig-5

Articles with High Research Impact:

The article titled “Prediction of In-hospital Mortality in Emergency Department Patients With Sepsis: A Local Big Data-Driven, Machine Learning Approach” is the most highly cited publication in the field of machine learning in emergency medicine, with 271 citations since its publication in 2016.

The second most cited article, published in 2018 by James Levin, Scott, et al., is titled “Machine-Learning-Based Electronic Triage More Accurately Differentiates Patients With Respect to Clinical Outcomes Compared With the Emergency Severity Index” [8].

The top 10 most highly cited papers related to machine learning in emergency medicine are listed in Table-2.

Table-2: Top 10 High-Cited Papers Related to Machine Learning in Emergency Medicine

Analysis of Keywords and Co-occurrence Clusters in Publications on Machine Learning in Emergency Medicine:

A total of 153 keywords occurring more than five times were identified and classified into eight distinct clusters (Fig-6A). Of these, 59 keywords that appeared at least ten times since 2019 were selected for further analysis. Fig-6B shows that keywords represented by more yellow dots are more recent, indicating emerging research trends such as early warning score, sepsis, emergency medical services, cardiac arrest, and COVID-19.

Fig-6: The Analysis of Keywords in Publications of Machine Learning in Emergency Medicine

The largest keyword cluster was labeled “out-of-hospital cardiac arrest” (cluster #0), followed by clusters related to emergency department (cluster #1), deep learning (cluster #2), computed tomography (cluster #3), artificial neural network (cluster #4), hip fracture (cluster #5), C-reactive protein (cluster #6), and machine learning (cluster #7). The top eight keywords with the strongest citation bursts since 2014 included “impact” (burst strength 3.44), “mortality” (3.17), and “emergency medical services” (3.16). Notably, the burst for “emergency medical services” started in 2020 and has continued through 2023 (Fig7A and Fig-7B).

Clusters with warmer colors and larger nodes represented more recent publications, highlighting active research hotspots in this field. Among these, cluster #0 (out-of-hospital cardiac arrest) has become a continuous focus of study, reflecting growing interest in applying machine learning to predict out-of-hospital cardiac arrest, which has attracted considerable attention in emergency medicine (Fig-7C).

Fig-7

Discussion

Research Trends of Machine Learning in Emergency Medicine:

The analysis showed a growing trend in the number of scientific articles published, especially from 2017 through 2023. Machine learning in emergency medicine has gradually attracted more people’s attention. Overall, the United States had the largest number of published papers, accounting for about 34.70% of the total publications, with China and Australia ranking second and third. It is still noteworthy that China ranked second in the number of articles, while it ranked sixth and tied for second for the citation frequency and H-index, respectively. The reasons for the inconsistency between the quantity and quality of Chinese articles might be as follows. The number of articles in China did not increase until 2018, especially in 2021, when the number of articles published accounted for 57.5% of all published articles. China takes a long time to catch up with the citation frequency compared to other countries. Secondly, research in China mostly comes from single-center studies, lacking multi-center participation, so the intensity of literature evidence may be insufficient. Although Australia, England, and Canada had published fewer papers than China from 2012 to 2021, the sum of citations, average citation frequency per paper, and the H-index were higher than those of China. Therefore, Chinese researchers in the future should pay attention to promoting the quality of papers.

In the field of machine learning in emergency medicine, the United States owned 12 of the top 15 related institutions and had the top three institutions for the publication of papers, which fully illustrates the leadership of the United States. In terms of artificial intelligence technology research, development, and application, the United States is at the world’s top level. Its excellent technical research and development and elite laboratories have laid a solid technical foundation for the development of artificial intelligence, which explains why it has made a large number of eye-catching research and development achievements. Besides, Seoul National University (South Korea), Monash University (Australia), and University of Toronto (Canada) were the only non-US institutions in the top 15, ranked fourth, sixth, and eleventh, respectively. At present, there is a lack of corresponding influential institutions in our country, which indicates that creating first-class research institutes is essential for increasing the academic level of a country.

The American Journal of Emergency Medicine (IF 3.6), Resuscitation (IF 6.5), and Injury International Journal of the Care of the Injured (IF 2.5) were the top three journals publishing articles (Fig-5). Of the top 10 articles cited, two were published in Annals of Emergency Medicine, two in Academic Emergency Medicine, two in Resuscitation, two in Injury International Journal of the Care of the Injured, and Pediatric Emergency Care respectively, one in American Journal of Emergency Medicine, and one in Emergency Medicine. This shows that future high-quality articles will likely appear in the journals described above. With regard to the authors, 6 out of the top 10 authors of the published articles were from South Korea, of whom Lee JH had published a total of 8 articles, occupying the top spot in the number of articles. Taylor RA from Yale University ranked second in the number of publications, and the total citation frequency of his papers was the highest on the list.

Research Focused on Machine Learning in Emergency Medicine:

Based on the reading and analysis of the top 10 highly cited literature, machine learning has attracted great attention from researchers in predicting sepsis, trauma, coronary heart disease, COVID-19, and emergency triage [8-17]. These highly cited papers were helpful in identifying the main hotspots in machine learning research in emergency medicine. (Table-2) lists the details of the top 10 articles cited by machine learning in emergency medicine. The paper “Prediction of In-hospital Mortality in Emergency Department Patients With Sepsis: A Local Big Data-Driven, Machine Learning Approach” was the most cited paper, which has been cited 192 times since its publication. The corresponding author was Taylor, R. Andrew, who published the research in Academic Emergency Medicine in 2016. The machine learning approach outperformed existing clinical decision rules (CDRs) and traditional analytic techniques for predicting in-hospital sepsis mortality of ED patients [9]. Keywords reflect the core and theme of an article, and high-frequency keywords represent hot topics in this field to a certain extent. At the same time, the research frontier can be grasped by detecting keywords with rapidly increasing frequency in different periods.

Mortality and risk were the top two keywords. Mortality was an important index to evaluate the prognosis of patients, and the application of machine learning in emergency medicine was mainly focused on trauma, surgery, and sepsis [18-20]. Machine learning has good performance in disease risk factor screening. In discriminating emergency department (ED) patients likely to progress to septic shock, LR and XGB models showed better reclassification than the baseline model with positive NRI [21]. Among patients presenting to the ED with altered mental status, the data-driven decision tool had good predictive accuracy on triage of these patients’ admission risk [22].

Burst words represent keywords often quoted in a certain period, thus indicating cutting-edge features and trends. Combined with the time dynamic evolution of keywords and burst word analysis, we can understand the research field, development, and future research trends of machine learning in emergency medicine. In Figure, we see that since 2018, the focus of machine learning has been related to survival and out-of-hospital cardiac arrest (OHCA). As a new tool, machine learning was widely used in predicting the prognosis of common diseases in the emergency department. For example, nomogram can accurately and efficiently predict the overall survival of SAP patients in the first 24 hours after admission [23]. Using a data-driven approach with a machine learning algorithm, researchers confirmed that machine learning could predict the survival of OHCA [24]. Related to OHCA, machine learning is used in ROSC after cardiac arrest, Utstein-based ROSC, OHCA, cardiac arrest hospital prognosis, and cardiac arrest survival score [25]. In addition, emergency medical services have become a hot topic and frontier since 2020.

Based on machine learning models, severity levels can be predicted and accurately identified, aiding triage by predicting mortality, hospital admission, and critical care requirements [26]. The use of machine-learning models for automated decision-making in vehicle routing and scheduling can improve the mobilisation of vehicles for pre-hospital emergency services in developing or low-income countries. This can help reduce rescue response times and ensure the maximum benefit from available resources for emergency care [27]. Based on over 360,000 structured emergency call center records received by Singapore’s National Emergency Call Centre in 2018-2020, NUS constructed a machine learning model that significantly reduced the overclassification rate, suggesting that this approach can be used in call centers to provide better ambulance dispatch triage and case acuity recommendations to optimise ambulance resource utilisation [28].

Strengths and limitations

Our research is the first bibliometric analysis of machine learning in emergency medicine research activities. There are several limitations to this study. First, we have included only English literature, while some critical but non-English studies on this topic may have been ignored. Second, this research only collects original articles and reviews extracted from the Web of Science database, which is still constantly updating, so our results may be somewhat biased. In the future, we should use more databases and pay attention to non-English studies, which are also important.

Conclusion

In conclusion, our article shows the recent trend of machine learning in emergency medicine. Global publications will continue to rise, and developed countries have contributed more. The United States is the most productive country. The quantity and quality of Chinese articles are inconsistent. The latest progress can be tracked in the American Journal of Emergency Medicine, Resuscitation, and Injury International Journal of the Care of the Injured. Disease prognosis and prediction may become hotspots in the near future. In the identification research cluster, “mortality” and “risk” are determined to be the hotspots, while “emergency medical services,” “out-of-hospital cardiac arrest (OHCA),” and “emergency medical services” are the new trends in machine learning in emergency medicine.

Conflict of Interest

The authors have read and approved the final version of the manuscript. The authors have no conflicts of interest to declare.

References

[1] Shafaf N, Malek H. Applications of Machine Learning Approaches in Emergency Medicine; a Review Article. Arch Acad Emerg Med. 2019 Jun 3;7(1):34. [PMID: 31555764]

[2] Ohu I, Benny PK, Rodrigues S, Carlson JN. Applications of machine learning in acute care research. J Am Coll Emerg Physicians Open. 2020 Jul 2;1(5):766-72. [PMID: 33145517]

[3] Basu S, Johnson KT, Berkowitz SA. Use of Machine Learning Approaches in Clinical Epidemiological Research of Diabetes. Curr Diab Rep. 2020 Dec 3;20(12):80. [PMID: 33270183]

[4] Gu P, Li W, Zhao X, Xu D. The Top 100 Most Cited Articles on Intrauterine Adhesion: a Bibliometric Analysis. Reprod Sci. 2022 Feb;29(2):460-74. [PMID: 34780024]

[5] Raita Y, Goto T, Faridi MK, Brown DFM, Camargo CA Jr, Hasegawa K. Emergency department triage prediction of clinical outcomes using machine learning models. Crit Care. 2019 Feb 22;23(1):64. [PMID: 30795786]

[6] Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, Bellomo R, Bernard GR, Chiche JD, Coopersmith CM, Hotchkiss RS, Levy MM, Marshall JC, Martin GS, Opal SM, Rubenfeld GD, van der Poll T, Vincent JL, Angus DC. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016 Feb 23;315(8):801-10. [PMID: 26903338]

[7] Christodoulou E, Ma J, Collins GS, Steyerberg EW, Verbakel JY, Van Calster B. A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models. J Clin Epidemiol. 2019 Jun;110:12-22. [PMID: 30763612]

[8] Holmes JF, Lillis K, Monroe D, Borgialli D, Kerrey BT, Mahajan P, Adelgais K, Ellison AM, Yen K, Atabaki S, Menaker J, Bonsu B, Quayle KS, Garcia M, Rogers A, Blumberg S, Lee L, Tunik M, Kooistra J, Kwok M, Cook LJ, Dean JM, Sokolove PE, Wisner DH, Ehrlich P, Cooper A, Dayan PS, Wootton-Gorges S, Kuppermann N; Pediatric Emergency Care Applied Research Network (PECARN). Identifying children at very low risk of clinically important blunt abdominal injuries. Ann Emerg Med. 2013 Aug;62(2):107-116.e2. [PMID: 23375510]

[9] Taylor RA, Pare JR, Venkatesh AK, Mowafi H, Melnick ER, Fleischman W, Hall MK. Prediction of In-hospital Mortality in Emergency Department Patients With Sepsis: A Local Big Data-Driven, Machine Learning Approach. Acad Emerg Med. 2016 Mar;23(3):269-78. [PMID: 26679719]

[10] Hess EP, Brison RJ, Perry JJ, Calder LA, Thiruganasambandamoorthy V, Agarwal D, Sadosty AT, Silvilotti ML, Jaffe AS, Montori VM, Wells GA, Stiell IG. Development of a clinical prediction rule for 30-day cardiac events in emergency department patients with chest pain and possible acute coronary syndrome. Ann Emerg Med. 2012 Feb;59(2):115-25.e1. [PMID: 21885156]

[11] Levin S, Toerper M, Hamrock E, Hinson JS, Barnes S, Gardner H, Dugas A, Linton B, Kirsch T, Kelen G. Machine-Learning-Based Electronic Triage More Accurately Differentiates Patients With Respect to Clinical Outcomes Compared With the Emergency Severity Index. Ann Emerg Med. 2018 May;71(5):565-74.e2. [PMID: 28888332]

[12] Peck JS, Benneyan JC, Nightingale DJ, Gaehde SA. Predicting emergency department inpatient admissions to improve same-day patient flow. Acad Emerg Med. 2012 Sep;19(9):E1045-54. [PMID: 22978731]

[13] Haimovich AD, Ravindra NG, Stoytchev S, Young HP, Wilson FP, van Dijk D, Schulz WL, Taylor RA. Development and Validation of the Quick COVID-19 Severity Index: A Prognostic Tool for Early Clinical Decompensation. Ann Emerg Med. 2020 Oct;76(4):442-53. [PMID: 33012378]