Application of Machine Learning on Emergency Department Management: An Integrative Review

by Roison Andro Narvaez, MSN RN CMCS CLDP LGBH
Mark Angelo Picar Abellera, RN MSN St.
Billie Ann Orofino Boregas, RN MSN St.
Cherry Mae Baganian, RN MSN St.
Joebelle Babiera Escamis, RN MSN St.

St. Paul University Philippines – Graduate School

Citation: Narvaez, R. A., Abellera, M. A., Boregas, B. A., Baganian, C. M.,  Dela Cruz, C. J., & Escamis, J. B. (2024). Application of machine learning on emergency department management: An integrative review. Canadian Journal of Nursing Informatics, 19(3). https://cjni.net/journal/?p=13489

Abstract

Background: Worldwide, many hospitals are investing significantly in and integrating artificial intelligence (AI) models in their electronic system due to their promising advantages. However, the application of AI in Emergency Departments (EDs) are still ambiguous which prompt researchers to seek and analyze the literature in relation to the use of AI models in the ED.

Aim: To get an understanding of the current state of artificial intelligence (AI) applications or tools in EDs and to what extent these applications have impacted how ED clinicians do their everyday duties.

Design: This study used an integrative review design.

Result: Nineteen (19) studies were included based on eligibility criteria. The majority were cohort studies with Level of Evidence IV. Also, employed AI interventions varied, and most AI tools or models discussed in the literature generated positive outcomes in multiple aspects: clinical decision-making, reduced cognitive workloads, reduced ED visits, predicted outcomes, improved ED triage, and reduced triage overcrowding.

Conclusion: AI significantly impacts the roles of emergency clinicians in the ED as they focus on developing clinical decision support tools for the detection or prediction of clinical outcomes.  It is unlikely that they will replace the jobs of clinicians in the near future, despite some studies showing that these tools work better than ED clinicians in some aspects of care. Further study is still recommended for future researchers.

Implications for Practice: Most applied AI significantly improved ED clinicians’ decision-making and simultaneously lessened diagnostic errors and the amount of mental effort necessary, thereby, have the potential to improve productivity and efficiency. AI in ED practice could vastly improve the quality of care rendered to patients.

What is known about the topic?

1. Research into Artificial Intelligence for healthcare primarily focuses on proof-of-theory studies.
2. Widespread application of Artificial Intelligence in ED settings is currently being hampered by several issues.
3. Using specialized technologies, the ED healthcare team will be able to give patients better, more personalized treatment if they are able to overcome the challenges preventing future deployment.

What does this paper add?

1. The researchers found out that most of the literature focuses on improving clinical decision-making of ED clinicians. Most AI enabled tools are centered on prediction and identification of several clinical variables.
2. AI enabled tools have been found to accurately and efficiently predict clinical conditions, risk of mortality, number of hospital visits, admissions and so forth.

Background

The demand for better medical attention has grown rapidly during the past few years. The increasing burden of sickness, multimorbidity, and disability is putting stress on the healthcare system (Panch et al., 2018). This is because of an aging and changing population, an increase in the number of people seeking health services, higher expectations from society, and rising health costs (Dall et al., 2013; de Meijer et al., 2013; Lloyd-Sherlock et al., 2020). Improving the effectiveness of healthcare services is a long-term goal, especially for countries with a low or medium human development index, because the health of the population has a direct effect on other economic factors like productivity, welfare, and social stability (Asandului et al., 2014), but it is not clear if it is still possible for people to meet the high standards set by healthcare professionals (Jalal et al., 2020).

 To deal with these problems, healthcare facilities around the world are using artificial intelligence (AI) more and more in their daily work (Jiang et al., 2017; Shaheen, 2021). The term “artificial intelligence” (AI) is often used to describe computer technologies that act like systems that help humans think, learn, adapt, engage, and understand their senses (Secinaro et al., 2021; Raschka et al., 2020). Intelligent systems and applications, particularly those based on machine learning (ML)-based artificial intelligence (AI), have evidently been widely accepted across all societal levels and sectors (e.g., Siri, Alexa). But there have been questions about how reliable, trustworthy, and fair they are in the face of algorithmic bias and unfairness (Mueller et al., 2022).

When artificial intelligence (AI) systems are taught with incomplete or skewed data (their “worldview”), it can lead to biased “thinking,” which can make prejudice and inequality worse, spread false information, and even cause physical harm (Xu, 2019). These kinds of concerns have prevented some ambitious AI initiatives from ever getting off the ground. The goal of the subfield of computer science called “machine learning,” (ML) is to make algorithms work better by searching through large data sets for useful patterns that can then be used to make reliable calculations or predictions (Stewart et al., 2018; Soun et al., 2020). The speed with which AI can draw conclusions is one of its strengths, which explains why it potentially has applications in emergency treatment. (Berlyand et al. 2018). Since AI has already proven to be fast and accurate, this is a huge plus (Matheny et al., 2020).

According to Weisberg et al. (2020), AI has great possibility of enhancing the efficacy and standard of emergency medical services. Many people are concerned that AI will someday replace human emergency personnel like radiologists (Grant et al., 2020). AI, on the other hand, has the potential to change the role of the emergency doctor because it is smarter than humans (Mazurowski, 2019). Even though there is more and more written about AI’s potential in healthcare settings, no studies have looked at how AI-based clinical decision support tools affect what clinicians do or how long it takes to treat patients in the emergency department (Rendell et al., 2018). The use and impact of AI tools in the EDs are still ambiguous and need to be explored and brought to light. This prompted the researchers to undertake a thorough literature review and understand how AI is changing the way ED clinicians do their daily jobs.

Objectives

The following are the objectives of this systematic review:

1. to know how AI and machine learning are being used in emergency medicine.
2. to know how useful machine learning-based AI is in urgent care settings.
3. to identify ML-based AI effects on ED’s capabilities.
4. to recognize the implications of machine learning-based AI on the organizational framework of emergency medical services

Method

Design

This research is based on a thorough literature review on the roles of machine learning applications in emergency department management, using Whittemore and Knafl’s (2005) integrated review process as a guide. The integrative review process consists of five steps: 1) determining the problem; 2) gathering information or conducting a literature search; 3) evaluating the information gathered; 4) analyzing the information; and 5) determining and presenting the results, which can make a significant contribution to a body of knowledge and thus to practice and research(Russell, 2005).

Search Strategy

The investigation was performed on November 19, 2022. Relevant studies were retrieved from electronic resources including ScienceDirect, PubMed, SAGE Journals, and Google Scholar. The following key phrases were used: machine learning AND/OR artificial intelligence, AND emergency medicine OR emergency departments to search for appropriate articles.

Quality Assessment

The researchers independently and manually evaluated each of the 1,500 articles from the electronic databases. All articles that met the inclusion criteria focused on AI, the potential for AI to affect the work of ED clinicians, and research that advanced the field of AI. The selected literature included in this synthesis must have been published from 2017 to 2022, since these studies are still considered relevant. To ensure the papers’ quality, they underwent peer review. In addition, the articles were written in English, which increased their visibility.

Titles that were duplicates, articles published in non-English and translated into English, protocols, abstracts, and opinions, were eliminated. Information technology evolves rapidly, so it is likely that articles published before 2017 are no longer relevant. Researchers were able to include 19 articles in the refined search. Studies that concluded negative outcomes for the AI-based application and did not perform better than typical methods were also disregarded. A PRISMA flow chart depicts the conclusion of this selection method in Figure 1.

Figure 1.

Prisma Diagram

Data Abstraction and Data Analysis

A matrix table was used to classify the data that had been extracted from the database using the Sparbel and Anderson (2000) tool with the following information: author, year of publication, design, method, sample size, participants, sampling technique, aims, AI method, and the study’s findings. This review utilized the Melnyk and Fineout-Overholt (2015) Rating System for the Hierarchy of Evidence for Intervention and Treatment Questions to establish the level of evidence (LOE) for each study, as illustrated in Table 1.

Table 1.

Summary of Included Studies on Machine Learning in Emergency Department Management

Results

The initial search for credible literature generated 1500 results, wherein, nineteen (19) satisfactorily met the inclusion and exclusion criteria (Table 1) and were assessed for quality (Table 2). The authors illustrated the search process via manual identification and screening (Fig. 1). All (19) of the papers studied were classified as having level of evidence IV. The studies were conducted in the USA (n=7), Canada (n=2), and Israel (n=2). In addition, one article was conducted in each of the following countries: Singapore (n=1), China, Taiwan (n=1), United Kingdom (n=1), South Korea (n=1), Hong Kong (n=1), Turkey (n=1) Australia (n=1), and Iran (n=1). Most studies were conducted in either the United States or Asia, and the majority of studies utilized triage-stage EHR data.

To comprehend the impact of AI on ED clinicians’ jobs, it is necessary to determine for what purposes AI is implemented in the ED. The theme represents a distinct aspect of the problem, methodology, and application of AI investigated in individual studies.

Table 2.

Quality Assessment of the Full-Text Scan

Table 3 illustrates how artificial intelligence was implemented and how it aids the ED’s staff in treating patients.

Table 3.

Goals of using AI in ED

1a. Improving patient outcomes is a vast term that encompasses diverse facets, such as lower morbidity and less complications (Chiew et al., 2019; Goto et al., 2018; Horng et al., 2017;  Jiang et al., 2017).  Improving patient outcomes includes reducing mortality rates (Hunter-Zinck et al., 2019) decreasing lengths of hospital stays, and raising levels of patient satisfaction (Pak et al., 2021).

1b. The overcrowding of emergency departments is a frequently cited issue. The most prominent factor contributing to ED overcrowding is the aging population’s longer life expectancy (Frost et al., 2017). Reducing the number of visits would have many effects on both doctors and patients, such as giving doctors more time with each patient (Goto et al., 2018).

Artificial intelligence (AI) in emergency care intends to improve ED congestion management in more ways than one. Sometimes it just is not possible to cut down on the number of visits. Overcrowding in EDs can be dealt with in various ways, one of which is to increase flow or patient throughput (Kulshrestha et al., 2021; Tahayori et al., 2020).

2a. Many studies try to predict how patients will get sick and what will happen to them so that ED’s can run better. This can help physicians allocate resources to patients at high risk (Yousef et al., 2021).

2b. Prediction and early detection are nearly identical. Early detection means that a patient already has a condition, while prediction can prevent it from happening (Xie et al., 2021). They typically go together. As with prediction, early detection of the outcome may avert a consequential occurrence.

2c. Whether technological advances that can predict and identify future outcomes will render human clinicians obsolete in the future in addition to recognizing and forecasting future outcomes (Kerr, 2020).

3a. The majority of the time, outcomes were predicted and identified through triage. The triage then determined the course of ED care because it was the first opportunity to classify patients (Levin et al., 2018).

3b. Critical outcomes are not all present during triage; consequently, other literature focuses on enhancing emergency department (ED) practices after triage (De Fauw et al., 2018).

3c. Try to predict or find events that are bad but not life-threatening, like hospitalization or admission (Kuo et al., 2020; Pak et al., 2021).

Table 4 discusses the benefits of using AI in EDs and how it assists healthcare workers in patient management.

Table 4.

Influence of AI on ED Work Design

4a. An AI-based clinical decision support tool (CDST) is used in most published works to predict or find problems. CDSTs have existed for more than ten years and aid clinicians in clinical decision-making. In some instances, artificial intelligence could be applied as more accurate than conventional CDSTs, thereby improving healthcare delivery (Rendell et al., 2018).

4b. According to studies, healthcare is improved using AI-based tools. Their model for predicting cardiac arrest reduces false alarms, thereby reducing alarm fatigue and desensitization(Jang et al., 2021). Thus, healthcare may become better.

4c. AI-based tools have the potential to transform ED management on a more centralized and commercial scale(Wang et al., 2021).

4d. As healthcare delivery gets better, AI-based tools can help emergency departments better use their resources, including their staff. By anticipating hospital demand, it is possible, for instance, to adjust schedules to meet demand (Vollmer et al., 2021).

4e. An improved utilization of available resources may also contribute to increased productivity. There are numerous ways to be efficient, but one that has stood the test of time is cost reduction (Frost et al., 2017).

Table 5 discusses the effectiveness of AI in carrying out skilled work in ED and how it lessens the workload of healthcare professionals in ED.

 Table 5.

Work design of ED clinicians and AI’s impact

5a. With the increased need for healthcare, clinicians’ cognitive workloads likewise increase, but this burden can be alleviated by operating an AI-based tool. Thus, clinicians can return to focusing on clinical care(Jiang et al., 2021).

5b. As part of modifying clinical care, most of the literature in this section recommended the use of an AI-based tool to enhance clinician decision-making and reduce disparities. AI-based CDSTs can alert clinicians to anomalies that are difficult to detect (Smith et al., 2019).

5c. While there was some written material that touches on non-diagnostic errors, the details of what might be done to lessen these errors are left out. Over- or under-triaging, both of which can lead to mistakes, can be avoided (Klug et al., 2019).

5d. Several researchers noted that using AI in the ED facilitated a change from crisis-driven to preventative care. For instance, once wait times are expected to rise, on-call doctors can be quickly paged(Kuo et al., 2020).

5e. There were a few research efforts that have resulted in an AI-based tool that can assist or even replace human doctors. Two researchers in this area have figured out that an AI-based tool is equivalent to a medical professional. None of that was ever intended (Klang et al., 2019; Tahayori et al., 2020).

Discussion

In this section, the authors discuss the main findings of the literature review to answer the research questions. The most important results of this study are shown in Figure 2. Figure 2 graphically illustrates the distinguished process of related causes and possible subsequent effects.

Figure 2.

Possible outcomes of various forms of artificial intelligence use in ED

Nineteen (19) eligible studies from the searched literature were included in this review on the use of AI in the ED. Overall, the studies showed various AI interventions when it comes to work design in ED as well as their purpose. Most AI based tools were used for clinical decision support for ED clinicians. Most studies centered on or were aimed at prediction or identification purposes, as AI has become familiar due to its superior ability in prediction modeling compared to the conventional statistical tools or models used. For instance, the prediction ability of AI tools and models in the reviewed literature can accurately predict the number of visits in ED, likelihood of hospital admission, admission rates and length of stay, patients at high risks of acquiring complications (such as septic shock, acute renal damage, major adverse cardiac events), risks of mortality, outcomes of patients with COPD and Asthma, and real-time patient wait-time.

The power of AI to accurately predict several clinical factors and outcomes are attributed to their ability to quickly process several variables across multiple and large data sets. Some studies showed results that AI enabled tools employed can accurately diagnose scaphoid fractures and identify patients who need Head CT scans; hence it is also useful not just in the ED alone but also the Radiology Department. Other studies compared AI enabled tools to other clinical tools as well as clinicians of which one resulted in the same performance, one study found that AI tools were superior to the other tools and over clinicians (which made other researchers conclude that AI may possibly replace clinicians at some point). Subsequently, the literature showed positive outcomes on enhanced ED triage and risk stratification to reduce overcrowding and prioritize patient care effectively. This can then improve decision making, reduce diagnostic errors and reduce workloads of ED clinicians.

According to the study results, the goal of AI-based solutions in the ED is to enhance both the quality of care delivered to patients and the efficiency with which doctors perform their daily jobs. The best working environment benefits both the physicians and the patients. Improving patient outcomes is a broad term that incorporates several elements, such as reduced morbidity and fewer problems (Chiew et al., 2019; Goto et al., 2018; Hong et al., 2017; Jiang et al., 2017). As the quality of treatment improves, clinicians may have more job satisfaction.

An AI-based CDST, or clinical decision support tool, is used in most published works to predict or find problems. CDSTs have existed for more than ten years and aid clinicians in clinical decision-making. In some instances, AI could be programmed to be more accurate than conventional CDSTs, thereby improving healthcare delivery (Rendell et al., 2018).

AI-powered CDSTs are intended to assist physicians in formulating clinical judgments, not replace them (Klug et al., 2019). Some literature indicated that AI-based technologies can possibly replace physicians. However, this is not the objective of the research analyzed in this study. Ozkaya et al. (2022) stated that their instruments are not superior to physicians but could augment them. For instance, it may assist hospitals that lack a certain kind of professional in making sound judgments about patient care. According to Farahmand et al. (2017), this technology cannot totally replace an expert physician, but it helps expedite decision-making when there are too many individuals to attend to. Even if the need for a doctor were reduced, human care is still critical.

Moreover, the technological singularity (TS) is a possible future in which artificial intelligence (AI) exceeds human intellect. In healthcare, TS would replace doctors with AI-guided robots and peripheral equipment. Given the rate at which AI technology is expanding and being incorporated into health care systems, TS in health care may arise in the near future, and AI-enabled services may significantly complement physicians’ knowledge (Shuaib et al., 2020).

In contrast, however, these findings appear to be the opposite of several studies. Schwartz (1970) predicted the augmenting function and, at the same time, the possibility that these AI enabled machines could largely take over the cognitive functions of the physicians. Parikh (2019) as well as other authors concluded in their respective studies that there is no data suggesting actual development in patient outcomes and that studies on AI in healthcare were merely exaggerated.

Some researchers also acknowledged the potential hazards of integrating AI in healthcare, such as socio-ethical, technological, and clinical concerns for both professionals and patients (Ellahham et al., 2020; Gerke & Cohen, 2020; Kantheti & Manne, 2021). Furthermore, Lekadir et al. (2022) detailed several risks of AI in their study, including: 1) AI errors and patient harm, 2) improper use, 3) human biases, inequalities, and inequities in most countries as affected by socioeconomic status, ethnicity, discrimination, and so on, and 4) the issue of inadequate transparency. 5) a lack of data privacy, confidentiality, and patient rights protection; 6) concerns with accountability and inadequacies in AI; and 7) impediments to healthcare adoption. Since AI-based CDSTs can learn, they are likely to surpass doctors in the future (Ozkaya, et al., 2022). Despite the fact that some studies expressly state that AI-based CDSTs are not intended to replace clinical judgment, this does happen on occasion (Klang, et al., 2019). As a consequence, some occupations may become obsolete over time. However, technology will take years to reach that position, and only time will tell what happens when it does.

The great majority of studies done revealed that AI is used to triage to either decrease the number of patients in congested emergency rooms or better manage them. These challenges often necessitated the deployment of a CDST powered by AI that could predict or detect probable future consequences (Levin et al., 2018). According to Martinez et al. (2020), the increased number of patients treated by doctors necessitates the introduction of AI-based CDSTs. It is visibly not possible to keep up with the demand while the ED and its physicians’ workload continues to rise (Klang et al., 2019). AI-based CDSTs may, for instance, expedite therapy and detect patients who are at high risk for death and under-treatment (Klug et al., 2019). The majority of CDSTs aim to improve emergency department (ED) triage via the development of prediction or identification tools. Triage is the very first opportunity to classify patients. As a result, it serves as a standard for emergency department treatment (Levin et al., 2018).

The results of this integrative review support previous literature reviews on AI use in EDs, such as the study by Stewart et al (2018) who explained that AI had multifaceted applications in the ED, for instance, analyzing clinical images, predicting outcomes and monitoring patients. Another author, Jiang et al. (2021) concluded that ML models could be used to differentiate between low-risk and high-risk patients as a strategy to improve efficiency and resource allocation. Nonetheless, AI can influence the work design of ED clinicians by assisting their clinical decisions and reduce workload to alleviate their rising clinical burdens.

It is not necessary to assume that a tool will work for all patients simply because it is effective for a particular patient population. On the other hand, one of ML’s strengths is that it is straightforward to retrain (Vollmer et al., 2021). and that it never stops learning (Smith et al., 2019). The authors are conscious that the studies that were chosen for this review primarily investigated the efficacy of AI in ED and that they were able to have a wider understanding of  AI applications and conditions that promote and inhibit the effective use of AI.

Implications for Practice

AI-based interventions pose several outcomes but the most significant implications for clinicians are that they effectively improve clinical decision-making while simultaneously lessening diagnostic errors and lessening the amount of mental effort necessary. AI improves the ED clinicians’ capacities and aids in the innovation of clinicians’ decisions with far-greater results. In addition, AI has the potential to improve ED productivity and efficiency. It also helps manage the intricate system of interconnected areas inside the hospital. While there are ongoing debates on the degree to which AI can affect the ED, it can undeniably change the role of clinicians. Moreover, working with AI tools or models will prompt the clinicians to collaboratively work with the engineers and manufacturers from time to time to improve systems even more. Lastly, ED practice with the assistance of AI would vastly improve clinicians’ ability to provide quality patient care, as it saves time and fosters good patient-clinician relationships. As a result, it affects both patients and clinicians’ satisfaction.

Limitations and Recommendations

The following are three limitations of this systematic review:

First, the studies that were looked at for this review did not look at how AI-based tools might change the way clinicians do their work. These studies primarily focused on the clinical utility of AI and what it can contribute to ED in general, making it complicated for their specific job design.

Second, this study does not explore comprehensively the various types of AI-based tools used, even though they are discussed in the literature. Third, most studies emphasized the presence of implementation barriers because AI use in ED research is still in its early stages. A tool that works for one patient may not work for all patients.

Based on the findings and conclusions collected, the authors recommend the following:

a.) The authors highly advocate the need for further study on AI use in the EDs, as more studies are needed to elucidate and evaluate the outcomes

b.) Future researchers are also recommended to study the different barriers that have an influence on implementing AI enabled models in the ED. 

c.) Other researchers should explore further the experiences of ED clinicians on handling and using AI applications; and

d.) Hospitals and Medical Schools should integrate discussion and training on the use of AI-enabled tools for both medical staff and students.

Conclusion

This systematic review of the literature was mostly focused on how artificial intelligence changes the roles of emergency clinicians in the ED. The vast majority of published works focused on developing clinical decision support tools for detection or prediction. Even though some research shows that these tools work better than clinicians, it is unlikely that they will replace the jobs of clinicians soon but they can augment the clinicians’ work to make the ED work more efficiently.

References

All with asterisk (*) are the included studies in the review

Asandului, L., Roman, M., & Fatulescu, P. (2014). The efficiency of healthcare systems in Europe: A data envelopment analysis approach. Procedia Economics and Finance, 10, 261–268. https://doi.org/10.1016/s2212-5671(14)00301-3

Berlyand, Y., Raja, A. S., Dorner, S. C., Prabhakar, A. M., Sonis, J. D., Gottumukkala, R. V., Succi, M. D., & Yun, B. J. (2018). How artificial intelligence could transform emergency department operations. The American Journal of Emergency Medicine, 36(8), 1515–1517. https://doi.org/10.1016/j.ajem.2018.01.017

Beyea, S. C., & Nicoll, L. H. (1998). Writing an integrative review. AORN Journal, 67(4), 877–880. https://doi.org/10.1016/s0001-2092(06)62653-7

*Chiew, C. J., Liu, N., Tagami, T., Wong, T. H., Koh, Z. X., & Ong, M. E. H. (2019). Heart rate variability-based machine learning models for risk prediction of suspected sepsis patients in the emergency department. Medicine, 98(6), e14197 https://doi.org/10.1097/md.000000000001419

Dall, T. M., Gallo, P. D., Chakrabarti, R., West, T., Semilla, A. P., & Storm, M. V. (2013). An aging population and growing disease burden will require a large and specialized health care workforce by 2025. Health Affairs, 32(11), 2013–2020. https://doi.org/10.1377/hlthaff.2013.0714

De Fauw, J., Ledsam, J. R., Romera-Paredes, B., Nikolov, S., Tomasev, N., Blackwell, S., Askham, H., Glorot, X., O’Donoghue, B., Visentin, D., van den Driessche, G., Lakshminarayanan, B., Meyer, C., Mackinder, F., Bouton, S., Ayoub, K., Chopra, R., King, D., Karthikesalingam, A., Hughes, C. O., … Ronneberger, O. (2018). Clinically applicable deep learning for diagnosis and referral in retinal disease. Nature Medicine, 24(9), 1342–1350. https://doi.org/10.1038/s41591-018-0107-6

De Meijer, C., Wouterse, B., Polder, J., & Koopmanschap, M. (2013). The effect of population aging on health expenditure growth: a critical review. European Journal of Ageing, 10(4), 353–361. https://doi.org/10.1007/s10433-013-0280-x

Ellahham, S., Ellahham, N., & Simsekler, M. C. E. (2020). Application of artificial intelligence in the health care safety context: opportunities and challenges. American Journal of Medical Quality, 35(4), 341-348. https://doi.org/10.1177/1062860619878515

*Farahmand, S., Shabestari, O., Pakrah, M., Hossein-Nejad, H., Arbab, M., & Bagheri-Hariri, S. (2017). Artificial intelligence-based triage for patients with acute abdominal pain in emergency department; a diagnostic accuracy study. Advanced Journal of Emergency Medicine, 1(1). https://doi.org/10.22114/AJEM.v1i1.11

*Fernandes, M., Mendes, R., Vieira, S. M., Leite, F., Palos, C., Johnson, A., Finkelstein, S., Horng, S., & Celi, L. A. (2020). Risk of mortality and cardiopulmonary arrest in critical patients presenting to the emergency department using machine learning and natural language processing. PLOS ONE, 15(4), e0230876. https://doi.org/10.1371/journal.pone.0230876                                          

Fineout-Overholt, E., Melnyk, B. M., Stillwell, S. B., & Williamson, K. M. (2010). Evidence-Based Practice, Step by Step: Critical Appraisal of the Evidence Part III. AJN, American Journal of Nursing, 110(11), 43–51. https://doi.org/10.1097/01.NAJ.0000390523.99066.b5

*Frost, D. W., Vembu, S., Wang, J., Tu, K., Morris, Q., & Abrams, H. B. (2017). Using the electronic medical record to identify patients at high risk for frequent emergency department visits and high system costs. The American Journal of Medicine, 130(5), 601.e17-601.e22. https://doi.org/10.1016/j.amjmed.2016.12.008

Gerke, S., Minssen, T., & Cohen, G. (2020). Ethical and legal challenges of artificial intelligence-driven healthcare. In Artificial intelligence in healthcare (pp. 295-336). Academic Press. https://doi.org/10.1016/B978-0-12-818438-7.00012-5

*Goto, T., Camargo, C. A., Faridi, M. K., Yun, B. J., & Hasegawa, K. (2018). Machine learning approaches for predicting disposition of asthma and COPD exacerbations in the ED. The American Journal of Emergency Medicine, 36(9), 1650–1654.                   https://doi.org/10.1016/j.ajem.2018.06.062

Grant, K., McParland, A., Mehta, S., & Ackery, A. D. (2020b). Artificial intelligence in emergency medicine: Surmountable barriers with revolutionary potential. Annals of Emergency Medicine, 75(6), 721–726. https://doi.org/10.1016/j.annemergmed.2019.12.024

*Greenbaum, N. R., Jernite, Y., Halpern, Y., Calder, S., Nathanson, L. A., Sontag, D. A., & Horng, S. (2019). Improving documentation of presenting problems in the emergency department using a domain-specific ontology and machine learning-driven user interfaces. International Journal of Medical Informatics, 132, 103981.  https://doi.org/10.1016/j.ijmedinf.2019.103981

*Hong, W. S., Haimovich, A. D., & Taylor, R. A. (2018b). Predicting hospital admission at emergency department triage using machine learning. PLOS ONE, 13(7), e0201016. https://doi.org/10.1371/journal.pone.0201016

Horng, S., Sontag, D. A., Halpern, Y., Jernite, Y., Shapiro, N. I., & Nathanson, L. A. (2017). Creating an automated trigger for sepsis clinical decision support at emergency department triage using machine learning. PLOS ONE, 12(4), e0174708. https://doi.org/10.1371/journal.pone.0174708

Hunter-Zinck, H. S., Peck, J. S., Strout, T. D., & Gaehde, S. A. (2019). Predicting emergency department orders with multilabel machine learning techniques and simulating effects on length of stay. Journal of the American Medical Informatics Association, 26(12), 1427–1436. https://doi.org/10.1093/jamia/ocz171

Jalal, S., Parker, W., Ferguson, D., & Nicolaou, S. (2020). Exploring the role of Artificial Intelligence in an emergency and trauma radiology department. Canadian Association of Radiologists Journal, 72(1), 167–174. https://doi.org/10.1177/0846537120918338

Jang, D. H., Kim, J., Jo, Y. H., Lee, J. H., Hwang, J. E., Park, S. M., Lee, D. K., Park, I., Kim, D., & Chang, H. (2021). Developing neural network models for early detection of cardiac arrest in emergency department. The American Journal of Emergency Medicine, 38(1), 43–49. https://doi.org/10.1016/j.ajem.2019.04.006

Jiang, F., Jiang, Y., Zhi, H., Dong, Y., Li, H., Ma, S., Wang, Y., Dong, Q., Shen, H., & Wang, Y. (2017). Artificial intelligence in healthcare: past, present and future. Stroke and Vascular Neurology, 2(4), 230–243. https://doi.org/10.1136/svn-2017-000101

*Jiang, H., Mao, H., Lu, H., Lin, P., Garry, W., Lu, H., Yang, G., Rainer, T. H., & Chen, X. (2021). Machine learning-based models to support decision-making in emergency department triage for patients with suspected cardiovascular disease. International Journal of Medical Informatics, 145, 104326. https://doi.org/10.1016/j.ijmedinf.2020.1

Kantheti, S. C., & Manne, R. (2021). Application of artificial intelligence in healthcare: chances and challenges. Current Journal of Applied Science and Technology, 40(6), 78-89. https://doi.org/10.9734/CJAST/2021/v40i631320

Kerr, R. S. (2020). Surgery in the 2020s: Implications of advancing technology for patients and the workforce. Future Healthcare Journal, 7(1), 46–49. https://doi.org/10.7861/fhj.2020-0001

* Kim, J., Chang, H., Kim, D., Jang, D. H., Park, I., & Kim, K. (2020). Machine learning for   prediction of septic shock at initial triage in the emergency department. Journal of Critical Care, 55, 163–170. https://doi.org/10.1016/j.jcrc.2019.09.024

*Klang, E., Barash, Y., Soffer, S., Bechler, S., Resheff, Y. S., Granot, T., Shahar, M., Klug, M., Guralnik, G., Zimlichman, E., & Konen, E. (2019). Promoting head CT exams in the emergency department triage using a machine learning model. Neuroradiology, 62(2), 153–160. https://doi.org/10.1007/s00234-019-02293-y

*Klug, M., Barash, Y., Bechler, S., Resheff, Y. S., Tron, T., Ironi, A., Soffer, S., Zimlichman, E., & Klang, E. (2019). A gradient boosting machine learning model for predicting early mortality in the emergency department triage: Devising a nine-point triage score. Journal of General Internal Medicine, 35(1), 220–227.  doi.org/10.1007/s11606-019-05512-7

*Kulshrestha, S., Dligach, D., Joyce, C., Baker, M. S., Gonzalez, R., O’Rourke, A. P., Glazer, J.  M., Stey, A., Kruser, J. M., Churpek, M. M., & Afshar, M. (2021). Prediction of severe chest injury using natural language processing from the electronic health record. Injury, 52(2), 205–212. https://doi.org/10.1016/j.injury.2020.10.094

*Kuo, Y. H., Chan, N. B., Leung, J. M., Meng, H., So, A. M. C., Tsoi, K. K., & Graham, C. A. (2020). An integrated approach of machine learning and systems thinking for waiting time prediction in an emergency department. International Journal of Medical Informatics, 139, 104143. https://doi.org/10.1016/j.ijmedinf.2020.104143

*Levin, S., Toerper, M., Hamrock, E., Hinson, J. S., Barnes, S., Gardner, H., Dugas, A., Linton, B., Kirsch, T., & Kelen, G. (2018). Machine-learning-based electronic triage more accurately differentiates patients with respect to clinical outcomes compared with the emergency severity index. Annals of Emergency Medicine, 71(5), 565-574.e2.  https://doi.org/10.1016/j.annemergmed.2017.08.00

Lekadir, K., Quaglio, G., Garmendia, A., & Gallin, C. (2022). Artificial Intelligence in Healthcare: Applications, risks, and ethical and societal impacts. Panel for the Future of Science and Technology, https://www.europarl.europa.eu/RegData/etudes/STUD/2022/729512/EPRS_STU(2022)729512_EN.pdf?fbclid=IwAR1Kni82xKqv9Oc13WJxrTKGQBUEoVuginHthGCTW1XkkzehlsJ8cHTXFnA

Lloyd-Sherlock, P., Beard, J., Minicuci, N., Ebrahim, S., & Chatterji, S. (2014). Hypertension among older adults in low- and middle-income countries: Prevalence, awareness and control. International Journal of Epidemiology, 43(1), 116–128. https://doi.org/10.1093/ije/dyt215

*Martinez, D. A., Levin, S. R., Klein, E. Y., Parikh, C. R., Menez, S., Taylor, R. A., & Hinson, J. S. (2020). Early prediction of acute kidney injury in the emergency department with machine-learning methods applied to electronic health record data. Annals of Emergency Medicine, 76(4), 501–514. https://doi.org/10.1016/j.annemergmed.2020.05.026

Matheny, M. E., Whicher, D., & Thadaney Israni, S. (2020). Artificial Intelligence in Health Care. JAMA, 323(6), 509. https://doi.org/10.1001/jama.2019.21579                                                                          

Mazurowski, M. A. (2019b). Artificial intelligence may cause a significant disruption to the radiology workforce. Journal of the American College of Radiology, 16(8), 1077–1082. https://doi.org/10.1016/j.jacr.2019.01.026

Melnyk, B.M. & Fineout-Overholt, E. (2015). “Box 1.3: Rating system for the hierarchy of evidence for intervention/treatment questions” in Evidence-based practice in nursing & healthcare: A guide to best practice (3rd ed.) (pp. 11). Wolters Kluwer Health.

*Mowbray, F., Zargoush, M., Jones, A., de Wit, K., & Costa, A. (2020). Predicting hospital admission for older emergency department patients: Insights from machine learning. International Journal of Medical Informatics, 140, 104163.  https://doi.org/10.1016/j.ijmedinf.2020.104163

Mueller, B., Kinoshita, T., Peebles, A., Graber, M. A., & Lee, S. (2022). Artificial intelligence and machine learning in emergency medicine: A narrative review. Acute Medicine &Amp; Surgery, 9(1). https://doi.org/10.1002/ams2.740

*Ozkaya, E., Topal, F. E., Bulut, T., Gursoy, M., Ozuysal, M., & Karakaya, Z. (2020). Evaluation of an artificial intelligence system for diagnosing scaphoid fracture on direct radiography. European Journal of Trauma and Emergency Surgery, 48(1), 585–592. https://doi.org/10.1007/s00068-020-01468-0

Pak, A., Gannon, B., & Staib, A. (2021a). Predicting waiting time to treatment for emergency department patients. International Journal of Medical Informatics, 145, 104303. https://doi.org/10.1016/j.ijmedinf.2020.104303

Panch, T., Szolovits, P., & Atun, R. (2018). Artificial intelligence, machine learning and health systems. Journal of Global Health, 8(2). https://doi.org/10.7189/jogh.08.020303\

Parikh, R.B., Teeple, S., Navathe, A.S. Addressing bias in artificial intelligence in health care. JAMA, 322(24), 2377–2378. https://doi.org/10.1001/jama.2019.18058

Raschka, S., Patterson, J., & Nolet, C. (2020). Machine learning in Python: Main developments and technology trends in data science, machine learning, and artificial intelligence. Information, 11(4), 193. https://doi.org/10.3390/info11040193

*Rendell, K., Koprinska, I., Kyme, A., Ebker?White, A. A., & Dinh, M. M. (2018). The Sydney Triage to Admission Risk Tool (START2) using machine learning techniques to support disposition decision?making. Emergency Medicine Australasia, 31(3), 429–435. https://doi.org/10.1111/1742-6723.13199

Rezaei, F., Yarmohammadian, M., Haghshenas, A., & Tavakoli, N. (2017). Overcrowding in emergency departments: A review of strategies to decrease future challenges. Journal of Research in Medical Sciences, 22(1), 23. https://doi.org/10.4103/1735-1995.200277

Russell, C. L. (2005c). An overview of the integrative research review. Progress in Transplantation, 15(1), 8–13. https://doi.org/10.1177/152692480501500102

Schwartz W. B. (1970). Medicine and the computer. The promise and problems of change. The New England journal of medicine, 283(23), 1257–1264. https://doi.org/10.1056/NEJM197012032832305

Secinaro, S., Calandra, D., Secinaro, A., Muthurangu, V., & Biancone, P. (2021). The role of artificial intelligence in healthcare: A structured literature review. BMC Medical Informatics and Decision Making, 21(1). https://doi.org/10.1186/s12911-021-01488-9                                                                                

Shaheen, M. Y. (2021). AI in Healthcare: medical and socio-economic benefits and challenges. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3932277

Shuaib, A., Arian, H., & Shuaib, A. (2020). The increasing role of artificial intelligence in health care: Will robots replace doctors in the future? International Journal of General Medicine, 13, 891.https://doi.org/10.2147/IJGM.S268093

Smith, S. W., Walsh, B., Grauer, K., Wang, K., Rapin, J., Li, J., Fennell, W., & Taboulet, P. (2019). A deep neural network learning algorithm outperforms a conventional algorithm for emergency department electrocardiogram interpretation. Journal of Electrocardiology, 52, 88–95. https://doi.org/10.1016/j.jelectrocard.2018.11.013

Soun, J., Chow, D., Nagamine, M., Takhtawala, R., Filippi, C., Yu, W., & Chang, P. (2020). Artificial intelligence and acute stroke imaging. American Journal of Neuroradiology, 42(1), 2–11. https://doi.org/10.3174/ajnr.a6883

Sparbel, K. J., & Anderson, M. A. (2000). Integrated literature review of continuity of care: Part 1, conceptual issues. Journal of Nursing Scholarship, 32(1), 17-24. https://doi.org/10.1111/j.1547-5069.2000.00017.x

Stewart, J., Sprivulis, P., & Dwivedi, G. (2018). Artificial intelligence and machine learning in emergency medicine. Emergency Medicine Australasia, 30(6), 870–874. https://doi.org/10.1111/1742-6723.13145

Stewart, L. A., Clarke, M., Rovers, M., Riley, R. D., Simmonds, M., Stewart, G., & Tierney, J. F. (2015). Preferred reporting items for a systematic review and meta-analysis of individual participant data.JAMA, 313(16), 1657. https://doi.org/10.1001/jama.2015.3656

Tahayori, B., Chini?Foroush, N., & Akhlaghi, H. (2020). Advanced natural language processing
technique to predict patient disposition based on emergency triage notes. Emergency
Medicine Australasia, 33(3), 480-484. https://doi.org/10.1111/1742-6723.13656

*Vollmer, M. A., Glampson, B., Mellan, T., Mishra, S., Mercuri, L., Costello, C., Klaber, R., Cooke, G., Flaxman, S., & Bhatt, S. (2021). A unified machine learning approach to time series forecasting applied to demand at emergency departments. BMC Emergency Medicine, 21(1). https://doi.org/10.1186/s12873-020-00395-y

Wang, X., Blumenthal, H. J., Hoffman, D., Benda, N., Kim, T., Perry, S., Franklin, E. S., Roth, E. M., Hettinger, A. Z., & Bisantz, A. M. (2021). Modeling patient-related workload in the emergency department using electronic health record data. International Journal of Medical Informatics, 150, 104451. https://doi.org/10.1016/j.ijmedinf.2021.104451

Weisberg, E. M., Chu, L. C., & Fishman, E. K. (2020). The first use of artificial intelligence (AI) in the ER: Triage not diagnosis. Emergency Radiology, 27(4), 361–366. https://doi.org/10.1007/s10140-020-01773-6

Whittemore, R., & Knafl, K. (2005). The integrative review: updated methodology. Journal of Advanced Nursing, 52(5), 546–553. https://doi.org/10.1111/j.1365-2648.2005.03621.x

Xie, F., Ong, M. E. H., Liew, J. N. M. H., Tan, K. B. K., Ho, A. F. W., Nadarajan, G. D., Low, L. L., Kwan, Y. H., Goldstein, B. A., Matchar, D. B., Chakraborty, B., & Liu, N. (2021b). Development and assessment of an interpretable machine learning triage tool for estimating mortality after emergency admissions. JAMA Network Open, 4(8), e2118467. https://doi.org/10.1001/jamanetworkopen.2021.18467

Xu, W. (2019). Toward human-centered AI. Interactions, 26(4), 42-46. https://doi.org/10.1145/3328485

Yousef, M. H., Alhalaseh, Y. N., Mansour, R., Sultan, H., Alnadi, N., Maswadeh, A., Al-Sheble, Y. M., Sinokrot, R., Ammar, K., Mansour, A., & Al-Hussaini, M. (2021). The fair allocation of scarce medical resources: A comparative study from Jordan. Frontiers in Medicine, 7, Article 603406. https://doi.org/10.3389/fmed.2020.603406

* Zhang, P. I., Hsu, C. C., Kao, Y., Chen, C. J., Kuo, Y. W., Hsu, S. L., Liu, T. L., Lin, H. J., Wang, J. J., Liu, C. F., & Huang, C. C. (2020). Real-time AI prediction for major adverse cardiac events in emergency department patients with chest pain. Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine, 28(1).  https://doi.org/10.1186/s13049-020-00786-x

Acknowledgments

The researchers acknowledge St. Paul University Philippines for their moral support and guidance in the manuscript.

Author Notes

1. Roison Andro Narvaez                               https://orcid.org/0000-0001-7555-5420             

2. Mark Angelo Picar Abellera                      https://orcid.org/0000-0001-6226-0172

3. Billie Ann Orofino Boregas                       https://orcid.org/0000-0001-7567-1915

4. Cherry Mae Baganian                                 https://orcid.org/0000-0003-0911-4971

5. Christine Joyce Obeña Dela Cruz              https://orcid.org/0000-0003-2951-5182

6. Joebelle Babiera Escamis                           https://orcid.org/0000-0002-4932-4167