Machine learning methods applied to triage in emergency services: A systematic review

Can artificial intelligence help ED nurses more accurately triage patients?

ABSTRACT

The Emergency Severity Index (ESI) is the most popular tool used to triage patients in the US and abroad. Evidence has shown that ESI has its limitations in correctly assigning acuity. To address this, AI can be incorporated into the triage process, decreasing the likelihood of assigning an incorrect ESI level.

Drinking from the Holy Grail-Does a Perfect Triage System Exist? And Where to Look for It?

The Emergency Department (ED) is a facility meant to treat patients in need of medical assistance. The choice of triage system hugely impactsed the organization of any given ED and it is important to analyze them for their effectiveness. The goal of this review is to briefly describe selected triage systems in an attempt to find the perfect one. Papers published in PubMed from 1990 to 2022 were reviewed. The following terms were used for comparison: "ED" and "triage system". The papers contained data on the design and function of the triage system, its validation, and its performance. After studies comparing the distinct means of patient selection were reviewed, they were meant to be classified as either flawed or non-ideal. The validity of all the comparable segregation systems was similar. A possible solution would be to search for a new, measurable parameter for a more accurate risk estimation, which could be a game changer in terms of triage assessment. The dynamic development of artificial intelligence (AI) technologies has recently been observed. The authors of this study believe that the future segregation system should be a combination of the experience and intuition of trained healthcare professionals and modern technology (artificial intelligence).

Artificial Intelligence in Emergency Trauma Care: A Preliminary Scoping Review

This study aimed to analyze the use of generative artificial intelligence in the emergency trauma care setting through a brief scoping review of literature published between 2014 and 2024. An exploration of the NCBI repository was performed using a search string of selected keywords that returned N=87 results; articles that met the inclusion criteria (n=28) were reviewed and analyzed. Heterogeneity sources were explored and identified by a significance threshold of P < 0.10 or an I2 value exceeding 50%. If applicable, articles were categorized within three primary domains: triage, diagnostics, or treatment. Findings suggest that CNNs demonstrate strong diagnostic performance for diverse traumatic injuries, but generalized integration requires expanded prospective multi-center validation. Injury scoring models currently experience calibration gaps in mortality quantification and lesion localization that can undermine clinical utility by permitting false negatives. Triage predictive models now confront transparency, explainability, and healthcare ecosystem integration barriers limiting real-world translation. The most significant literature gap centers on treatment-oriented generative AI applications that provide real-time guidance for urgent trauma interventions rather than just analytical support.

Evaluating the accuracy of a state-of-the-art large language model for prediction of admissions from the emergency room

BACKGROUND

Artificial intelligence (AI) and large language models (LLMs) can play a critical role in emergency room operations by augmenting decision-making about patient admission. However, there are no studies for LLMs using real-world data and scenarios, in comparison to and being informed by traditional supervised machine learning (ML) models. We evaluated the performance of GPT-4 for predicting patient admissions from emergency department (ED) visits. We compared performance to traditional ML models both naively and when informed by few-shot examples and/or numerical probabilities.

METHODS

We conducted a retrospective study using electronic health records across 7 NYC hospitals. We trained Bio-Clinical-BERT and XGBoost (XGB) models on unstructured and structured data, respectively, and created an ensemble model reflecting ML performance. We then assessed GPT-4 capabilities in many scenarios: through Zero-shot, Few-shot with and without retrieval-augmented generation (RAG), and with and without ML numerical probabilities.

RESULTS

The Ensemble ML model achieved an area under the receiver operating characteristic curve (AUC) of 0.88, an area under the precision-recall curve (AUPRC) of 0.72 and an accuracy of 82.9%. The naïve GPT-4's performance (0.79 AUC, 0.48 AUPRC, and 77.5% accuracy) showed substantial improvement when given limited, relevant data to learn from (ie, RAG) and underlying ML probabilities (0.87 AUC, 0.71 AUPRC, and 83.1% accuracy). Interestingly, RAG alone boosted performance to near peak levels (0.82 AUC, 0.56 AUPRC, and 81.3% accuracy).

CONCLUSIONS

The naïve LLM had limited performance but showed significant improvement in predicting ED admissions when supplemented with real-world examples to learn from, particularly through RAG, and/or numerical probabilities from traditional ML models. Its peak performance, although slightly lower than the pure ML model, is noteworthy given its potential for providing reasoning behind predictions. Further refinement of LLMs with real-world data is necessary for successful integration as decision-support tools in care settings.

Triage-clinical reasoning on emergency nursing competency: a multiple linear mediation effect

BACKGROUND

Triage is the first step in providing prompt and appropriate emergency nursing and addressing diagnostic issues. Rapid clinical reasoning skills of emergency nurses are essential for prompt decision-making and emergency care. Nurses experience limitations in emergency nursing that begin with triage. This cross-sectional study explored the mediating effect of perceived triage competency and clinical reasoning skills on the association between Korean Triage and Acuity Scale (KTAS) proficiency and emergency nursing competency.

METHODS

A web-based survey was conducted with 157 emergency nurses working in 20 hospitals in South Korea between mid-May and mid-July 2022. Data were collected utilizing self-administered questionnaires to measure KTAS proficiency (48 tasks), perceived triage competency (30 items), clinical reasoning skills (26 items), and emergency nursing competency (78 items). Data were analyzed using the PROCESS macro (Model 6).

RESULTS

Perceived triage competency indirectly mediate the relationship between KTAS proficiency and emergency nursing competency. Perceived triage competency and clinical reasoning skills were significant predictors of emergency nursing competency with a multiple linear mediating effect. The model was found have a good fit (F = 8.990, P <.001) with, a statistical power of 15.0% (R² = 0.150).

CONCLUSIONS

This study indicates that improving emergency nursing competency requires enhancing triage proficiency as well as perceived triage competency, which should be followed by developing clinical reasoning skills, starting with triage of emergency nurses.

Improved interpretable machine learning emergency department triage tool addressing class imbalance

Objective

The Score for Emergency Risk Prediction (SERP) is a novel mortality risk prediction score which leverages machine learning in supporting triage decisions. In its derivation study, SERP-2d, SERP-7d and SERP-30d demonstrated good predictive performance for 2-day, 7-day and 30-day mortality. However, the dataset used had significant class imbalance. This study aimed to determine if addressing class imbalance can improve SERP's performance, ultimately improving triage accuracy.

Methods

The Singapore General Hospital (SGH) emergency department (ED) dataset was used, which contains 1,833,908 ED records between 2008 and 2020. Records between 2008 and 2017 were randomly split into a training set (80%) and validation set (20%). The 2019 and 2020 records were used as test sets. To address class imbalance, we used random oversampling and random undersampling in the AutoScore-Imbalance framework to develop SERP+-2d, SERP+-7d, and SERP+-30d scores. The performance of SERP+, SERP, and the commonly used triage risk scores was compared.

Results

The developed SERP+ scores had five to six variables. The AUC of SERP+ scores (0.874 to 0.905) was higher than that of the corresponding SERP scores (0.859 to 0.894) on both test sets. This superior performance was statistically significant for SERP+-7d (2019: Z = -5.843, p < 0.001, 2020: Z = -4.548, p < 0.001) and SERP+-30d (2019: Z = -3.063, p = 0.002, 2020: Z = -3.256, p = 0.001). SERP+ outperformed SERP marginally on sensitivity, specificity, balanced accuracy, and positive predictive value measures. Negative predictive value was the same for SERP+ and SERP. Additionally, SERP+ showed better performance compared to the commonly used triage risk scores.

Conclusions

Accounting for class imbalance during training improved score performance for SERP+. Better stratification of even a small number of patients can be meaningful in the context of the ED triage. Our findings reiterate the potential of machine learning-based scores like SERP+ in supporting accurate, data-driven triage decisions at the ED.

Application of Machine Learning Techniques to Development of Emergency Medical Rapid Triage Prediction Models in Acute Care

Given the critical and complex features of medical emergencies, it is essential to develop models that enable prompt and suitable clinical decision-making based on considerable information. Emergency nurses are responsible for categorizing and prioritizing injuries and illnesses on the frontlines of the emergency room. This study aims to create an Emergency Medical Rapid Triage and Prediction Assistance model using electronic medical records and machine learning techniques. Patient information was retrieved from the emergency department of a large regional teaching hospital in Taiwan, and five supervised learning techniques were used to construct classification models for predicting critical outcomes. Of these models, the model using logistic regression had superior prediction performance, with an F1 score of 0.861 and an area under the receiver operating characteristic curve of 0.855. The Emergency Medical Rapid Triage and Prediction Assistance model demonstrated superior performance in predicting intensive care and hospitalization outcomes compared with the Taiwan Triage and Acuity Scale and three clinical early warning tools. The proposed model has the potential to assist emergency nurses in executing challenging triage assessments and emergency teams in treating critically ill patients promptly, leading to improved clinical care and efficient utilization of medical resources.

Applications of natural language processing at emergency department triage: A narrative review

INTRODUCTION

Natural language processing (NLP) uses various computational methods to analyse and understand human language, and has been applied to data acquired at Emergency Department (ED) triage to predict various outcomes. The objective of this scoping review is to evaluate how NLP has been applied to data acquired at ED triage, assess if NLP based models outperform humans or current risk stratification techniques when predicting outcomes, and assess if incorporating free-text improve predictive performance of models when compared to predictive models that use only structured data.

METHODS

All English language peer-reviewed research that applied an NLP technique to free-text obtained at ED triage was eligible for inclusion. We excluded studies focusing solely on disease surveillance, and studies that used information obtained after triage. We searched the electronic databases MEDLINE, Embase, Cochrane Database of Systematic Reviews, Web of Science, and Scopus for medical subject headings and text keywords related to NLP and triage. Databases were last searched on 01/01/2022. Risk of bias in studies was assessed using the Prediction model Risk of Bias Assessment Tool (PROBAST). Due to the high level of heterogeneity between studies and high risk of bias, a metanalysis was not conducted. Instead, a narrative synthesis is provided.

RESULTS

In total, 3730 studies were screened, and 20 studies were included. The population size varied greatly between studies ranging from 1.8 million patients to 598 triage notes. The most common outcomes assessed were prediction of triage score, prediction of admission, and prediction of critical illness. NLP models achieved high accuracy in predicting need for admission, triage score, critical illness, and mapping free-text chief complaints to structured fields. Incorporating both structured data and free-text data improved results when compared to models that used only structured data. However, the majority of studies (80%) were assessed to have a high risk of bias, and only one study reported the deployment of an NLP model into clinical practice.

CONCLUSION

Unstructured free-text triage notes have been used by NLP models to predict clinically relevant outcomes. However, the majority of studies have a high risk of bias, most research is retrospective, and there are few examples of implementation into clinical practice. Future work is needed to prospectively assess if applying NLP to data acquired at ED triage improves ED outcomes when compared to usual clinical practice.

Enhancing efficiency and capacity of telehealth services with intelligent triage: a bidirectional LSTM neural network model employing character embedding

BACKGROUND

The widespread adoption of telehealth services necessitates accurate online department selection based on patient medical records, a task requiring significant medical knowledge. Incorrect triage results in considerable time wastage for both patients and medical professionals. To address this, we propose an intelligent triage model based on a Bidirectional Long Short-Term Memory (Bi-LSTM) neural network with character embedding to enhance the efficiency and capacity of telehealth services.

METHODS

We gathered a 1.3 GB medical dataset comprising 200,000 records, each including medical history, physical examination data, and other pertinent information found on the electronic medical record homepage. Following data preprocessing, a clinical corpus was established to train character embeddings with a medical context. These character embeddings were then utilized to extract features from patient chief  complaints, and a 2-layer Bi-LSTM neural network was trained to categorize these complaints, enabling intelligent triage for telehealth services.

RESULTS

60,000 chief complaint-department data pairs were extracted from clinical corpus and divided into the training, validation, and test sets of 42,000, 9,000, and 9,000, respectively. The character embedding based Bi-LSTM neural network achieved a macro-precision of 85.50% and an F1 score of 85.45%.

CONCLUSION

The telehealth triage model developed in this study demonstrates strong implementation outcomes and significantly improves the efficiency and capacity of telehealth services. Character embedding outperforms word embedding, and future work will incorporate additional features such as patient age and gender into the chief complaint feature to future enhance model performance.

Supporting clinical decision making in the emergency department for paediatric patients using machine learning: A scoping review protocol

INTRODUCTION

Machine learning as a clinical decision support system tool has the potential to assist clinicians who must make complex and accurate medical decisions in fast paced environments such as the emergency department. This paper presents a protocol for a scoping review, with the objective of summarising the existing research on machine learning clinical decision support system tools in the emergency department, focusing on models that can be used for paediatric patients, where a knowledge gap exists.

MATERIALS AND METHODS

The methodology used will follow the scoping study framework of Arksey and O'Malley, along with other guidelines. Machine learning clinical decision support system tools for any outcome and population (paediatric/adult/mixed) for use in the emergency department will be included. Articles such as grey literature, letters, pre-prints, editorials, scoping/literature/narrative reviews, non-English full text papers, protocols, surveys, abstract or full text not available and models based on synthesised data will be excluded. Articles from the last five years will be included. Four databases will be searched: Medline (EBSCO), CINAHL (EBSCO), EMBASE and Cochrane Central. Independent reviewers will perform the screening in two sequential stages (stage 1: clinician expertise and stage 2: computer science expertise), disagreements will be resolved by discussion. Data relevant to the research question will be collected. Quantitative analysis will be performed to generate the results.

DISCUSSION

The study results will summarise the existing research on machine learning clinical decision support tools in the emergency department, focusing on models that can be used for paediatric patients. This holds the promise to identify opportunities to both incorporate models in clinical practice and to develop future models by utilising reviewers from diverse backgrounds and relevant expertise.

Machine learning model identification and prediction of patients' need for ICU admission: A systematic review

BACKGROUND

The emergency department (ED) triage process serves as a crucial first step for patients seeking acute care, This initial assessment holds crucial implications for patient survival and prognosis. In this study, a systematic review of the existing literature was performed to investigate the performance of machine learning (ML) models in recognizing and predicting the need for intensive care among ED patients.

METHODS

Four prominent databases (PubMed, Embase, Cochrane Library and Web of Science) were searched for relevant literature published up to April 28, 2023. The Prediction model study Risk of Bias Assessment Tool (PROBAST) was employed to evaluate the risk of bias and feasibility of prediction models.

RESULTS

In ten studies, the main algorithms used were Gradient Boostin, Logistic Regressio, Neural Network, Support Vector Machines, Random Forest. The performance of each model was as follows: Gradient Boosting had a sensitivity range of 0.3 to 0.96, specificity range of 0.6 to 0.99, accuracy range of 0.37 to 0.99, precision range of 0.3 to 0.96, and AUC value range of 0.68 to 0.93; Logistic Regression had a sensitivity range of 0.46 to 0.97, specificity range of 0.28 to 0.99, accuracy range of 0.66 to 0.97, precision range of 0.27 to 0.63, and AUC value range of 0.72 to 0.97; Neural Networks had a sensitivity range of 0.45 to 0.96, specificity range of 0.58 to 0.99, accuracy range of 0.36 to 0.97, precision range of 0.27 to 0.96, and AUC value range of 0.67 to 0.91; Support Vector Machines had a sensitivity range of 0.49 to 0.83, specificity range of 0.94 to 0.98, accuracy range of 0.33 to 0.97, precision range of 0.53 to 0.94, and AUC values were not reported; Random Forests had a sensitivity range of 0.75 to 0.91, specificity range of 0.77 to 0.94, accuracy range of 0.35 to 0.77, precision range of 0.36 to 0.94, and AUC value of 0.83.

CONCLUSION

ML models have demonstrated good performance in identifying and predicting critically ill patients in ED triage. However, because of the limited number of studies on each model, further high-quality prospective research is needed to validate these findings.

Triage: A Global Perspective

Triage is a process by which patients are assessed, classified, and sorted based on their presenting complaint and clinical urgency, providing assurance for timely access to emergency care. The goal is to get the right person to the right place, in the right amount of time, for the right reason, and within the context of resource availability. In many countries, a standardized triage system, underpinned through the use of guidelines, is used to provide clinicians with support and guidance. Triage is a globally adopted principle, and although triage guidelines are used in many countries, no single system has been internationally adopted. This paper discusses the importance of how triage process standardization improves patient care, resource management, and benchmarking at local, national, and international levels by applying 5 internationally recognized triage systems to fictional case studies. Evaluation of similarities and differences in severity scores, with a gap analysis, occurs.

Predictive Model for Mortality in Severe COVID-19 Patients across the Six Pandemic Waves

The impact of SARS-CoV-2 infection remains substantial on a global scale, despite widespread vaccination efforts, early therapeutic interventions, and an enhanced understanding of the disease's underlying mechanisms. At the same time, a significant number of patients continue to develop severe COVID-19, necessitating admission to intensive care units (ICUs). This study aimed to provide evidence concerning the most influential predictors of mortality among critically ill patients with severe COVID-19, employing machine learning (ML) techniques. To accomplish this, we conducted a retrospective multicenter investigation involving 684 patients with severe COVID-19, spanning from 1 June 2020 to 31 March 2023, wherein we scrutinized sociodemographic, clinical, and analytical data. These data were extracted from electronic health records. Out of the six supervised ML methods scrutinized, the extreme gradient boosting (XGB) method exhibited the highest balanced accuracy at 96.61%. The variables that exerted the greatest influence on mortality prediction encompassed ferritin, fibrinogen, D-dimer, platelet count, C-reactive protein (CRP), prothrombin time (PT), invasive mechanical ventilation (IMV), PaFi (PaO2/FiO2), lactate dehydrogenase (LDH), lymphocyte levels, activated partial thromboplastin time (aPTT), body mass index (BMI), creatinine, and age. These findings underscore XGB as a robust candidate for accurately classifying patients with COVID-19.

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.

On Scene Injury Severity Prediction (OSISP) model for trauma developed using the Swedish Trauma Registry

BACKGROUND

Providing optimal care for trauma, the leading cause of death for young adults, remains a challenge e.g., due to field triage limitations in assessing a patient's condition and deciding on transport destination. Data-driven On Scene Injury Severity Prediction (OSISP) models for motor vehicle crashes have shown potential for providing real-time decision support. The objective of this study is therefore to evaluate if an Artificial Intelligence (AI) based clinical decision support system can identify severely injured trauma patients in the prehospital setting.

METHODS

The Swedish Trauma Registry was used to train and validate five models - Logistic Regression, Random Forest, XGBoost, Support Vector Machine and Artificial Neural Network - in a stratified 10-fold cross validation setting and hold-out analysis. The models performed binary classification of the New Injury Severity Score and were evaluated using accuracy metrics, area under the receiver operating characteristic curve (AUC) and Precision-Recall curve (AUCPR), and under- and overtriage rates.

RESULTS

There were 75,602 registrations between 2013-2020 and 47,357 (62.6%) remained after eligibility criteria were applied. Models were based on 21 predictors, including injury location. From the clinical outcome, about 40% of patients were undertriaged and 46% were overtriaged. Models demonstrated potential for improved triaging and yielded AUC between 0.80-0.89 and AUCPR between 0.43-0.62.

CONCLUSIONS

AI based OSISP models have potential to provide support during assessment of injury severity. The findings may be used for developing tools to complement field triage protocols, with potential to improve prehospital trauma care and thereby reduce morbidity and mortality for a large patient population.

Medical emergency department triage data processing using a machine-learning solution

Over the years, artificial intelligence has demonstrated its ability to overcome many challenges in our day-to-day life. The evolution of it inquired more studies about Machine Learning possible solutions for different domains, including health care. The increasing demand for artificial intelligence solutions has brought accessibility to loads of data, including clinical data. The availability of medical records facilitates new opportunities to explore Machine Learning models and their abilities to process a significant amount of data and to identify patterns with the purpose of solving a medical problem. Understanding the applicability of artificial intelligence on this type of data has to be a compelling aim for emergency medicine clinicians. This paper focuses on the general clinical problem of the complex correlation between medical records and later diagnosis and, especially, on the process of emergency department triage which uses the Emergency Severity Index (ESI) as triage protocol. This study presents a comparison between three different Machine Learning models, such as Logistic Regression, Random Forest Tree and NN-Sequentail, with the purpose of classifying patients with an emergency code. We conducted four experiments because of imbalanced data. A web-based application was developed to improve the triage process after our theoretical and exploratory results. Overall, in all experiments, the NN-Sequential model had better results, having, in the first experiment, a ROC-AUC score for each ESI emergency code of: 0.59%, 0.76%, 0.71%, 0.78% 0.64%. After applying methods to balance the data, the model yielded a ROC-AUC score for each emergency code of 0.72%, 0.75%, 0.69%, 0.74%, 0.78%. In the last experiment consisting of a three-class classification problem, the NN-Sequential and Random Forest Tree models had similar metric outcomes, and the NN-Sequential algorithm had a ROC-AUC score for each emergency code of: 0.76%, 0.72%, 0.84%. Without any doubt, our research results presented in this paper endorse this tremendous curiosity in Machine Learning applications to enrich aspects of emergency medical care by applying specific methods for processing both medical data and medical records.

Predicting Hospital Ward Admission from the Emergency Department: A Systematic Review

BACKGROUND

The emergency department (ED) is often overburdened, due to the high influx of patients and limited availability of attending physicians. This situation highlights the need for improvement in the management of, and assistance provided in the ED. A key point for this purpose is the identification of patients with the highest risk, which can be achieved using machine learning predictive models. The objective of this study is to conduct a systematic review of predictive models used to detect ward admissions from the ED. The main targets of this review are the best predictive algorithms, their predictive capacity, the studies' quality, and the predictor variables.

METHODS

This review is based on PRISMA methodology. The information has been searched in PubMed, Scopus and Google Scholar databases. Quality assessment has been performed using the QUIPS tool.

RESULTS

Through the advanced search, a total of 367 articles were found, of which 14 were of interest that met the inclusion criteria. Logistic regression is the most used predictive model, achieving AUC values between 0.75-0.92. The two most used variables are the age and ED triage category.

CONCLUSIONS

artificial intelligence models can contribute to improving the quality of care in the ED and reducing the burden on healthcare systems.

The Use and Structure of Emergency Nurses' Triage Narrative Data: Scoping Review

BACKGROUND

Emergency departments use triage to ensure that patients with the highest level of acuity receive care quickly and safely. Triage is typically a nursing process that is documented as structured and unstructured (free text) data. Free-text triage narratives have been studied for specific conditions but never reviewed in a comprehensive manner.

OBJECTIVE

The objective of this paper was to identify and map the academic literature that examines triage narratives. The paper described the types of research conducted, identified gaps in the research, and determined where additional review may be warranted.

METHODS

We conducted a scoping review of unstructured triage narratives. We mapped the literature, described the use of triage narrative data, examined the information available on the form and structure of narratives, highlighted similarities among publications, and identified opportunities for future research.

RESULTS

We screened 18,074 studies published between 1990 and 2022 in CINAHL, MEDLINE, Embase, Cochrane, and ProQuest Central. We identified 0.53% (96/18,074) of studies that directly examined the use of triage nurses' narratives. More than 12 million visits were made to 2438 emergency departments included in the review. In total, 82% (79/96) of these studies were conducted in the United States (43/96, 45%), Australia (31/96, 32%), or Canada (5/96, 5%). Triage narratives were used for research and case identification, as input variables for predictive modeling, and for quality improvement. Overall, 31% (30/96) of the studies offered a description of the triage narrative, including a list of the keywords used (27/96, 28%) or more fulsome descriptions (such as word counts, character counts, abbreviation, etc; 7/96, 7%). We found limited use of reporting guidelines (8/96, 8%).

CONCLUSIONS

The breadth of the identified studies suggests that there is widespread routine collection and research use of triage narrative data. Despite the use of triage narratives as a source of data in studies, the narratives and nurses who generate them are poorly described in the literature, and data reporting is inconsistent. Additional research is needed to describe the structure of triage narratives, determine the best use of triage narratives, and improve the consistent use of triage-specific data reporting guidelines.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

RR2-10.1136/bmjopen-2021-055132.

Developing a machine learning model to predict patient need for computed tomography imaging in the emergency department

Overcrowding is a well-known problem in hospitals and emergency departments (ED) that can negatively impact patients and staff. This study aims to present a machine learning model to detect a patient's need for a Computed Tomography (CT) exam in the emergency department at the earliest possible time. The data for this work was collected from ED at Thunder Bay Regional Health Sciences Centre over one year (05/2016-05/2017) and contained administrative triage information. The target outcome was whether or not a patient required a CT exam. Multiple combinations of text embedding methods, machine learning algorithms, and data resampling methods were experimented with to find the optimal model for this task. The final model was trained with 81, 118 visits and tested on a hold-out test set with a size of 9, 013 visits. The best model achieved a ROC AUC score of 0.86 and had a sensitivity of 87.3% and specificity of 70.9%. The most important factors that led to a CT scan order were found to be chief complaint, treatment area, and triage acuity. The proposed model was able to successfully identify patients needing a CT using administrative triage data that is available at the initial stage of a patient's arrival. By determining that a CT scan is needed early in the patient's visit, the ED can allocate resources to ensure these investigations are completed quickly and patient flow is maintained to reduce overcrowding.

Association between triage nurses' job satisfaction and professional capability: Results of a mixed-method study

AIM

This study aims to examine factors related to the job satisfaction of triaging nurses and their professional capability in the clinical setting.

BACKGROUND

Triage is a complex process that relies on making decisions in favour of the patient and his treatment. The professional capability of a triaging nurse is an important psychological construct of job satisfaction.

METHODS

The study used a mixed-method methodology, with data collection based on an explanatory research design. The research instrument in the quantitative part was a survey questionnaire, and in the qualitative part, a semi-structured interview. The results were integrated using the 'Pillar Integration Process'.

RESULTS

There are significant relationships between professional capability and job satisfaction. Six main topics were exposed: characteristics and traits, work organization, safety is the key, burdening circumstances, capability and self-evaluation.

CONCLUSION

Professional capability is associated with job satisfaction. The necessary managerial changes should be made to achieve job satisfaction and develop professional competence while focusing on already trained and competent triage nurses, as satisfied triage nurses will stay longer in the institution.

IMPLICATIONS FOR NURSING MANAGEMENT

The manager's job is to be aware of the level of job satisfaction, take care to develop their employee's professional capability and take action in case of disrupted balance.

Use of Artificial Intelligence to Manage Patient Flow in Emergency Department during the COVID-19 Pandemic: A Prospective, Single-Center Study

BACKGROUND

During the coronavirus disease 2019 (COVID-19) pandemic, calculation of the number of emergency department (ED) beds required for patients with vs. without suspected COVID-19 represented a real public health problem. In France, Amiens Picardy University Hospital (APUH) developed an Artificial Intelligence (AI) project called "Prediction of the Patient Pathway in the Emergency Department" (3P-U) to predict patient outcomes.

MATERIALS

Using the 3P-U model, we performed a prospective, single-center study of patients attending APUH's ED in 2020 and 2021. The objective was to determine the minimum and maximum numbers of beds required in real-time, according to the 3P-U model. Results A total of 105,457 patients were included. The area under the receiver operating characteristic curve (AUROC) for the 3P-U was 0.82 for all of the patients and 0.90 for the unambiguous cases. Specifically, 38,353 (36.4%) patients were flagged as "likely to be discharged", 18,815 (17.8%) were flagged as "likely to be admitted", and 48,297 (45.8%) patients could not be flagged. Based on the predicted minimum number of beds (for unambiguous cases only) and the maximum number of beds (all patients), the hospital management coordinated the conversion of wards into dedicated COVID-19 units.

DISCUSSION AND CONCLUSIONS

The 3P-U model's AUROC is in the middle of range reported in the literature for similar classifiers. By considering the range of required bed numbers, the waste of resources (e.g., time and beds) could be reduced. The study concludes that the application of AI could help considerably improve the management of hospital resources during global pandemics, such as COVID-19.

eXtreme Gradient Boosting-based method to classify patients with COVID-19

Different demographic, clinical and laboratory variables have been related to the severity and mortality following SARS-CoV-2 infection. Most studies applied traditional statistical methods and in some cases combined with a machine learning (ML) method. This is the first study to date to comparatively analyze five ML methods to select the one that most closely predicts mortality in patients admitted with COVID-19. The aim of this single-center observational study is to classify, based on different types of variables, adult patients with COVID-19 at increased risk of mortality. SARS-CoV-2 infection was defined by a positive reverse transcriptase PCR. A total of 203 patients were admitted between March 15 and June 15, 2020 to a tertiary hospital. Data were extracted from the electronic medical record. Four supervised ML algorithms (k-nearest neighbors (KNN), decision tree (DT), Gaussian naïve Bayes (GNB) and support vector machine (SVM)) were compared with the eXtreme Gradient Boosting (XGB) method proposed to have excellent scalability and high running speed, among other qualities. The results indicate that the XGB method has the best prediction accuracy (92%), high precision (>0.92) and high recall (>0.92). The KNN, SVM and DT approaches present moderate prediction accuracy (>80%), moderate recall (>0.80) and moderate precision (>0.80). The GNB algorithm shows relatively low classification performance. The variables with the greatest weight in predicting mortality were C reactive protein, procalcitonin, glutamyl oxaloacetic transaminase, glutamyl pyruvic transaminase, neutrophils, D-dimer, creatinine, lactic acid, ferritin, days of non-invasive ventilation, septic shock and age. Based on these results, XGB is a solid candidate for correct classification of patients with COVID-19.

AI Models for Predicting Readmission of Pneumonia Patients within 30 Days after Discharge

A model with capability for precisely predicting readmission is a target being pursued worldwide. The objective of this study is to design predictive models using artificial intelligence methods and data retrieved from the National Health Insurance Research Database of Taiwan for identifying high-risk pneumonia patients with 30-day all-cause readmissions. An integrated genetic algorithm (GA) and support vector machine (SVM), namely IGS, were used to design predictive models optimized with three objective functions. In IGS, GA was used for selecting salient features and optimal SVM parameters, while SVM was used for constructing the models. For comparison, logistic regression (LR) and deep neural network (DNN) were also applied for model construction. The IGS model with AUC used as the objective function achieved an accuracy, sensitivity, specificity, and area under ROC curve (AUC) of 70.11%, 73.46%, 69.26%, and 0.7758, respectively, outperforming the models designed with LR (65.77%, 78.44%, 62.54%, and 0.7689, respectively) and DNN (61.50%, 79.34%, 56.95%, and 0.7547, respectively), as well as previously reported models constructed using thedata of electronic health records with an AUC of 0.71–0.74. It can be used for automatically detecting pneumonia patients with a risk of all-cause readmissions within 30 days after discharge so as to administer suitable interventions to reduce readmission and healthcare costs.