Accurate Triage Research Articles

The incidence, providers involved, and patient factors associated with diagnosis of specific lumbar spine pathology subsequent an initial nonspecific low back pain diagnosis

Background ContextLow back pain (LBP) is an exceedingly common diagnosis with significant societal and healthcare burden. While nonspecific LBP diagnoses are frequently assigned and often resolve soon thereafter, some patients may subsequently be diagnosed with specific underlying lumbar spine pathology. PurposeTo characterize the incidence, providers involved, and factors associated with specific pathology LBP (SP-LBP) diagnosed subsequent an initial nonspecific LBP (NS-LBP) diagnosis. Study DesignRetrospective cohort study. Patient SampleAdult patients diagnosed with NS-LBP between 2010 – Q1 2022 were abstracted from a large national administrative database. Outcome MeasurementsIncidence of subsequent SP-LBP diagnoses, providers involved, and factors associated with SP-LBP diagnoses in the 12-months following initial NS-LBP diagnosis. MethodsAdult patients diagnosed with NS-LBP were identified based on International Classification of Disease (ICD) codes ICD-9-D-724.2 and ICD-10-D-M54.5. These patients diagnosed with new NS-LBP were reported based on incidence per 100,000 covered lives. Exclusion criteria included: patients less than 18 years old, those with any prior NS-LBP diagnosis, any prior SP-LBP diagnoses, and those with trauma, neoplasm, or infection diagnosed within 3-months prior to the initial NS-LBP diagnosis. Patients who subsequently received SP-LBP diagnoses within 12-months following initial NS-LBP were then identified. Patients were characterized based on age, sex, history of cancer, obesity, opioid use disorder, anxiety, and depression and then compared with patients receiving no additional LBP diagnosis by univariable and multivariable analysis. Provider specialties who initially diagnosed NS-LBP and those involved with subsequent SP-LBP diagnoses were identified. ResultsNS-LBP was identified for a robust cohort of 6,861,235 patients, yielding an incidence of 4,258 diagnoses per 100,000 covered lives. Subsequent SP-LBP diagnoses were identified for 562,982 (8.21%) within 12-months, of which 62.57% of these SP-LBP diagnoses occurred within 3-months. The most common of these subsequent diagnoses included radiculopathy (3.01% of the overall study population), disc degeneration (2.84%), and disc herniation (1.23%). Patients receiving a subsequent SP-LBP diagnosis were more likely older (per year increase, odds ratio [OR] 1.02), male (OR 1.10), had a history of cancer (OR 1.32), obesity (OR 1.30), opioid use disorder (OR 1.17), anxiety (OR 1.21), or depression (OR 1.17) (p<0.001 for all). NS-LBP was initially diagnosed predominantly by primary care providers (PCPs) in family medicine (44.09%) and internal medicine (27.49%). These PCP providers also contributed to the majority of SP-LBP diagnoses, while specialty providers contributed to a greater degree in SP-LBP diagnoses than NS-LBP. ConclusionNS-LBP is a common diagnosis that usually, 91.79% of the time, does not subsequently convert to a SP-LBP diagnosis within 12-months. This may reflect improvement of LBP symptoms and/or accurate initial triage, largely by PCPs. Of the minority of patients receiving a subsequent SP-LBP diagnosis, various patient factors were significantly associated, and a greater degree of specialty providers were involved, which highlights potential contributors to diagnostic difficulty of SP-LBP and the multidisciplinary nature of LBP management.

A new prehospital cardiac risk assessment index derived from vital signs by first contact emergency medical service, including prehospital oxygen saturation, for accurate prehospital triage of acute m

Abstract Background: Rapid and accurate triage is essential to improve the prognosis of patients with severe acute myocardial infarction (AMI). However, prehospital triage using vital signs, including prehospital oxygen saturation (SpO2), by first medical contact has not been evaluated. Purpose: The purpose of this study was to develop a new risk stratification index for AMI using vital signs including SpO2 for rapid and accurate prehospital triage. Methods: From January 2013 to June 2022, 3006 AMI patients who were directly transported by ambulance were enrolled from the Mie ACS registry. Based on observed risk relationships in 3,006 AMI patients, we developed and evaluated a prehospital cardiac risk assessment index (P-CRAI) using age, heart rate, and systolic blood pressure (SBP) and SpO2 to predict 30-day mortality: (heart rate×[age/10]2)/SBP＋1.1(100-SpO2). We also analyzed whether P-CRAI predicted the need for mechanical circulatory support (MCS) or ventilator use. Results: In the receiver operating characteristic (ROC) curve analysis, P-CRAI 38.1 has a sensitivity of 84.4%, specificity of 72.7%, and accuracy of 84.3% for predicting 30-day mortality, and P-CRAI was a strong predictor of 30-day mortality compared with the other indices and variables (Figure 1A). When divided into two groups with P-CRAI 38, Kaplan-Meier survival curves showed significant differences in 30-day survival between the high and low groups (83.7% vs. 98.7%, P&amp;lt;0.001, Figure 1B). P-CRAI≥38 was the strongest independent predictor of 30-day mortality with a hazard ratio of 7.64 compared to the other factors by multivariate Cox regression analysis. The spline curve of 30-day mortality by P-CRAI shows a correlation with prognosis (Figure 2A). In this study, we then used the bootstrap method to generate a calibration curve and validate P-CRAI. The results showed an excellent fit and that P-CRAI accurately estimated patient prognosis (Figure 2B). In addition, P-CRAI was a robust predictor of the very early events of death at 24 hours and MCS or ventilator use by ROC curve analysis. (c-statistic 0.835 and 0.726, respectively, P&amp;lt;0.01). Conclusion: The P-CRAI was found to be a useful tool for prehospital triage of AMI patients, accurately predicting 30-day mortality and the need for MCS.Figure 1.Figure 2.

Comparative Analysis of Burn Injuries in Toddler and Preschool Children: Implications for Triage and Outcome Assessment.

This study aims to compare the demographic, clinical characteristics, and outcomes of burn injuries in toddler and preschool children, and to validate the American Burn Association (ABA) Burn Triage Decision Matrix in the Turkish pediatric population. A retrospective analysis was conducted on 684 pediatric burn patients (494 toddlers, 190 preschoolers) admitted to our burn center over a 5-year period. Variables including gender, burn etiology, burn area, depth, treatment modalities, complications, length of hospital stay, and mortality were analyzed. The performance of the ABA Burn Triage Decision Matrix was evaluated in both groups. Scalding was the predominant cause of burns in both groups, with a significant difference in the involvement of anterior trunk (p = 0.027). The mean Total Body Surface Area (TBSA) was comparable between the groups (p = 0.286). There was no significant difference in mortality rates (p = 0.385), treatment modalities, and complications. The ABA Burn Triage Decision Matrix demonstrated consistency in triaging the severity of burn injuries, with a notable discrepancy observed in the moderate risk category of toddler group. This study highlights the distinct characteristics and outcomes of burn injuries in different pediatric age groups. The ABA Burn Triage Decision Matrix's validation suggests its utility in enhancing triage accuracy and resource allocation in pediatric populations, especially in disaster-prone regions.

How large language model-powered conversational agents influence decision making in domestic medical triage contexts

IntroductionEffective delivery of healthcare depends on timely and accurate triage decisions, directing patients to appropriate care pathways and reducing unnecessary visits. Artificial Intelligence (AI) solutions, particularly those based on Large Language Models (LLMs), may enable non-experts to make better triage decisions at home, thus easing the healthcare system's load. We investigate how LLM-powered conversational agents influence non-experts in making triage decisions, further studying different persona profiles embedded via prompting.MethodsWe designed a randomized experiment where participants first assessed patient symptom vignettes independently, then consulted one of the two agent profiles—rational or empathic—for advice, and finally revised their triage ratings. We used linear models to quantify the effect of the agent profile and confidence on the weight of advice. We examined changes in confidence and accuracy of triage decisions, along with participants' perceptions of the agents.ResultsIn a study with 49 layperson participants, we found that persona profiles can be differentiated in LLM-powered conversational agents. However, these profiles did not significantly affect the weight of advice. Notably, less confident participants were more influenced by LLM advice, leading to larger adjustments to initial decisions. AI guidance improved alignment with correct triage levels and boosted confidence in participants' decisions.DiscussionWhile LLM advice improves triage recommendations accuracy, confidence plays an important role in its adoption. Our findings raise design considerations for human-AI interfaces, highlighting two key aspects: encouraging appropriate alignment with LLMs' advice and ensuring that people are not easily swayed in situations of uncertainty.

Unraveling the impact of abdominal arterial calcifications on kidney transplant waitlist mortality through ensemble machine learning

The scarcity of organ donors relative to the number of patients with End Stage Kidney Disease (ESKD) has led to prolonged waiting times for kidney transplants, contributing to elevated cardiovascular mortality risk. Transplant professionals are tasked with the complex allocation of limited organs to a vulnerable patient group facing heightened morbidity and mortality risk. The need for continuous re-evaluation of waitlisted patients is evident due to the significant number who perish while awaiting transplantation. Among individuals with ESKD, vascular calcification, particularly Abdominal Arterial Calcifications (AAC), holds predictive value for cardiovascular events and mortality. However, a standardized method to quantify AAC’s prognostic potential remains lacking, especially for kidney transplant evaluations. This study presents an ensemble machine learning (ML) approach to study the relationship between AAC score and mortality in patients on the waitlist and triage patients needing transplantation. Using the AAC score, the proposed ML model can predict kidney transplant waitlist morality with an accuracy of 78% while its accuracy is 68% without using this score. This study leverages explainable ML to explore the relationship between predictors and mortality in waitlisted patients, aiming to improve patient triage accuracy.

Teledermoscopic triage of melanoma-suspicious skin lesions is safe: A retrospective comparative diagnostic accuracy study with multiple assessors.

The rising incidence of melanoma and the high number of benign lesions excised due to diagnostic uncertainty highlight the need for effective patient triage. This study assesses the safety and accuracy of teledermoscopic triage on a high-prevalence case set with pre-triaged, challenging, melanoma-suspicious lesions. Five dermatologists independently reviewed 250 retrospectively extracted patient cases. Teledermoscopy assessments were simulated for panels of 1, 2, 3 and 5 assessors using two distinct consensus strategies, Caution Protocol and Majority Vote, and the sensitivity and specificity of the patient triages were calculated. Triage by a single teledermatologist showed a sensitivity of 92.3% and a specificity of 58.7%. Sensitivity improved with the number of assessors, particularly when using the Caution Protocol, though with a considerable drop in specificity. The Majority Vote showed a more balanced improvement in sensitivity and specificity. Safety analyses indicated that diagnostic accuracy decreased with poor image quality and increased case difficulty. Expert teledermoscopic triage of melanocytic skin lesions is highly sensitive and lowers the need for unnecessary excision procedures by half while dismissing as few as 0.4% (95% confidence interval 0-0.6%) of melanomas, even when applied to a high-prevalence pre-triaged subpopulation. Implementation of safety procedures increases accuracy. Using multiple teledermatologists increases sensitivity but at the cost of specificity unless a Majority Vote consensus strategy is applied. Future teledermoscopy guidelines should encompass safety procedures and protocols for disagreement between assessors.

The triage and diagnostic accuracy of frontier large language models: an updated comparison to physician performance (Preprint)

The triage and diagnostic accuracy of frontier large language models: an updated comparison to physician performance (Preprint)

A novel triage framework for emergency department based on machine learning paradigm

AbstractEmergency departments, crucial in managing patient emergencies, are often challenged by overcrowding and diagnostic errors during triage—the process that assesses the urgency of patients' conditions. Traditional triage systems, heavily dependent on human judgement, are prone to errors like under‐triage, where severe conditions are missed, delaying treatment, and over‐triage, where less severe conditions are overly prioritized, causing unnecessary resource use and morbidity. This thesis presents a novel multi‐model machine‐learning framework designed to enhance triage accuracy by evaluating multiple medical conditions concurrently, including chronic illnesses like heart disease. The framework employs various machine‐learning techniques—such as logistic regression, support vector machines, random forests, deep neural networks, and decision trees—to analyze different medical conditions in two comprehensive phases. In the first phase, patients' conditions are categorized, and a multi‐tiered analysis using multiple classifiers refines the assessment by considering probabilistic outcomes from each classifier. The second phase synthesizes these insights into a unified and precise triage decision, distinguishing between patients who require critical care and those needing less urgent hospitalization. This integration of diverse machine learning models allows for a fused and precise triage decision, overcoming traditional triage systems' limitations that usually focus on single conditions and rely on isolated model predictions. A hybrid feature selection method is also utilized to identify critical predictors, enhancing the decision‐making process. The framework is validated through a specially curated dataset that simulates multiple triage scenarios, evaluated by medical experts for efficacy. Comparative analysis with traditional triage methods demonstrates significant improvements in decision accuracy, as evidenced by higher area under curve (AUC) values—0.95 for critical care and 0.90 for hospitalization. The implementation of this data‐driven approach substantially reduces human error, boosts operational efficiency, and aids medical staff in making rapid, informed decisions. This research represents a significant advancement in emergency medical care, illustrating the benefits of integrating sophisticated machine learning techniques to improve patient outcomes and resource management.

Accuracy of a Commercial Large Language Model (ChatGPT) to Perform Disaster Triage of Simulated Patients Using the Simple Triage and Rapid Treatment (START) Protocol: Gage Repeatability and Reproducibility Study.

The release of ChatGPT (OpenAI) in November 2022 drastically reduced the barrier to using artificial intelligence by allowing a simple web-based text interface to a large language model (LLM). One use case where ChatGPT could be useful is in triaging patients at the site of a disaster using the Simple Triage and Rapid Treatment (START) protocol. However, LLMs experience several common errors including hallucinations (also called confabulations) and prompt dependency. This study addresses the research problem: "Can ChatGPT adequately triage simulated disaster patients using the START protocol?" by measuring three outcomes: repeatability, reproducibility, and accuracy. Nine prompts were developed by 5 disaster medicine physicians. A Python script queried ChatGPT Version 4 for each prompt combined with 391 validated simulated patient vignettes. Ten repetitions of each combination were performed for a total of 35,190 simulated triages. A reference standard START triage code for each simulated case was assigned by 2 disaster medicine specialists (JMF and MV), with a third specialist (LC) added if the first two did not agree. Results were evaluated using a gage repeatability and reproducibility study (gage R and R). Repeatability was defined as variation due to repeated use of the same prompt. Reproducibility was defined as variation due to the use of different prompts on the same patient vignette. Accuracy was defined as agreement with the reference standard. Although 35,102 (99.7%) queries returned a valid START score, there was considerable variability. Repeatability (use of the same prompt repeatedly) was 14% of the overall variation. Reproducibility (use of different prompts) was 4.1% of the overall variation. The accuracy of ChatGPT for START was 63.9% with a 32.9% overtriage rate and a 3.1% undertriage rate. Accuracy varied by prompt with a maximum of 71.8% and a minimum of 46.7%. This study indicates that ChatGPT version 4 is insufficient to triage simulated disaster patients via the START protocol. It demonstrated suboptimal repeatability and reproducibility. The overall accuracy of triage was only 63.9%. Health care professionals are advised to exercise caution while using commercial LLMs for vital medical determinations, given that these tools may commonly produce inaccurate data, colloquially referred to as hallucinations or confabulations. Artificial intelligence-guided tools should undergo rigorous statistical evaluation-using methods such as gage R and R-before implementation into clinical settings.

Emergency department triaging using ChatGPT based on emergency severity index principles: a cross-sectional study

Erroneous and delayed triage in an increasingly crowded emergency department (ED). ChatGPT is an artificial intelligence model developed by OpenAI® and is being trained for use in natural language processing tasks. Our study aims to determine the accuracy of patient triage using ChatGPT according to the emergency severity index (ESI) for triage in EDs. In our cross-sectional study, 18 years and over patients who consecutively presented to our ED within 24 h were included. Age, gender, admission method, chief complaint, state of consciousness, and comorbidities were recorded on the case form, and the vital signs were detected at the triage desk. A five-member expert committee (EC) was formed from the fourth-year resident physicians. The investigators converted real-time patient information into a standardized case format. The urgency status of the patients was evaluated simultaneously by EC and ChatGPT according to ESI criteria. The median value of the EC decision was accepted as the gold standard. There was a statistically significant moderate agreement between EC and ChatGPT assessments regarding urgency status (Cohen’s Kappa = 0.659; P < 0.001). The accuracy between these two assessments was calculated as 76.6%. There was a high degree of agreement between EC and ChatGPT for the prediction of ESI-1 and 2, indicating high acuity (Cohen’s Kappa = 0.828). The diagnostic specificity, NPV, and accuracy of ChatGPT were determined as 95.63, 98.17 and 94.90%, respectively, for ESI high acuity categories. Our study shows that ChatGPT can successfully differentiate patients with high urgency. The findings are promising for integrating artificial intelligence-based applications such as ChatGPT into triage processes in EDs.

Assessment of the effectiveness of hospital external disaster functional drills on health care receivers’ performance, using standardized patients and mass cards simulation: a pilot study from Saudi Arabia

BackgroundGiven the increasing frequency of disasters globally, it is critical that healthcare systems are prepared for these mass casualty events. The Saudi health system’s preparedness for mass casualty incidents needs to be more robust, potentially due to limited disaster drills and inadequate standardized patient (SP) simulation training. This study aims to (i) assess the performance of front-line hospital staff in Saudi Arabia through a functional drill and (ii) evaluate the drill’s effectiveness using SP and MAC-SIM cards, providing detailed insights into its design and execution.MethodsA functional drill was conducted at a government hospital in Riyadh, Saudi Arabia, on December 19, 2022, using a cross-sectional approach with two phases. 141 healthcare receivers served as subjects, while 23 volunteers acted as SPs. The scenario simulated a building collapse to assess the emergency department (ED) response, interdepartmental communication, and surge capacity. Data were collected through direct observation of healthcare practitioners’ interactions with the SPs, analysis of SP data, and participant feedback. Quantitative data were analyzed descriptively, while qualitative data were examined for patterns and themes related to simulation performance and effectiveness.ResultsThe hospital receivers’ performances demonstrated accurate triage categories. The ED team assessed most patients (67%) in less than 5 min. For patients requiring definitive care, such as intensive care unit, 95% spent less than 2.5 h in the ED. Most patients (65%) required ‘other treatments’. Communication was efficient in the triage zone and the yellow treatment area. Participants’ feedback on using MAC-SIM cards during the simulation was overwhelmingly positive with 82.61% reporting that MAC-SIM use helped them respond better. Experienced SPs (paramedics) with prior disaster knowledge and experience outperformed inexperienced SPs (nurses) in the functional exercise.ConclusionThis groundbreaking study is the first in the Arabic Gulf region to use SPs with MAC-SIM cards in functional drills. The findings highlight the potential of simulation exercises to improve hospital team knowledge and performance when responding to disasters. Multiple evaluation techniques can effectively identify participant strengths and weaknesses, informing future disaster improvement plans. This information is a valuable resource for Arabic and middle-income countries where disaster medicine is still developing.

The speed and accuracy of prehospital triage for simulated disaster patients: Comparing prehospital emergency triage rapid algorithm (PETRA) and simple triage and rapid treatment (START)

The speed and accuracy of prehospital triage for simulated disaster patients: Comparing prehospital emergency triage rapid algorithm (PETRA) and simple triage and rapid treatment (START)

Impact of Triage Nurse Training on Accuracy and Efficiency: A Systematic Review-Based Study

Background Triage in the emergency department (ED) is a crucial procedure that establishes the urgency of patient care based on symptoms at presentation. Although the efficacy of triage nurse training programs varies, their goal is to increase the precision and efficiency of their decision-making. Aim With a focus on research done between 2019 and 2024, the systematic review attempts to evaluate the effect of training programs on the accuracy and efficiency of triage nurses in emergency departments. Method We conducted a thorough search of the PubMed, EMBASE, and Web of Science databases to find pertinent research. Studies assessing training treatments for triage nurses and reporting accuracy and efficiency-related results were included in the inclusion criteria. Both the quality assessment and data extraction were carried out separately by two reviewers. Result Ten studies that used a range of triage training techniques and covered a variety of geographic areas satisfied the inclusion criteria. Research findings consistently showed that, despite variations in study designs and outcome measures, training programs increased triage accuracy and efficiency. Interventions included clinical decision support system integration, training based on simulation, and conventional classroom-based instruction. Conclusion By increasing decision-making efficiency and accuracy in patient evaluation, training programs greatly improve the performance of triage nurses in emergency departments. To improve comparability and support evidence-based policies, future research must standardize training methodologies and outcome measurements.

Biomarkers of COVID-19 short-term worsening: a multiparameter analysis within the prospective multicenter COVIDeF cohort.

During a pandemic like COVID-19, hospital resources are constrained and accurate severity triage of the patients is required. The objective of this study is to estimate the predictive performances of candidate biomarkers for short-term worsening (STW) of COVID-19. Prospective, multicenter (20 hospitals in Paris) cohort study of consecutive COVID-19 patients with systematic biobanking at admission, during the first waves of COVID-19 in France in 2020 (COVIDeF cohort). Consecutive COVID-19 patients were screened for inclusion. They were excluded in presence of severity criteria defined by either an ICU admission, mechanical ventilation (including noninvasive ventilation), acute respiratory distress, or in-hospital death before sampling. Routine blood tests measured during usual care and centralized systematic measurement of creatine kinase, C-reactive protein (CRP), procalcitonin, soluble urokinase plasminogen activator receptor (suPAR), high-sensitive troponin T (TnT-hs), N terminal pro-B natriuretic peptide (NT-proBNP), calprotectin, platelet factor 4, mid-regional pro-adrenomedullin (MR-proADM), and proendothelin were performed. The primary outcome was STW, defined by a severity criteria within 7 days. A backward stepwise logistic regression model and a 'best subset' approach were used to identify independent association, and the area under the receiving operator characteristics (AUROC) was computed. Five hundred and eleven patients were analyzed, of whom 60 (11.7%) experienced STW. Median time to occurrence of a severity criteria was 3 days. At admission, lower values of eosinophils, lymphocytes, platelets, alanine aminotransferase, and higher values of neutrophils, creatinine, urea, CRP, TnT-hs, suPAR, NT-proBNP, calprotectin, procalcitonin, MR-proADM, and proendothelin were predictive of worsening. Stepwise logistic regression identified three biomarkers significantly associated with worsening: CRP [adjusted odds ratio (aOR): 1.10, 95% confidence interval (95% CI): 1.06-1.15 for a 10-unit increase, AUROC: 0.73 (0.66-0.79)], procalcitonin [aOR: 0.42, 95% CI: 0.22-0.81, AUROC: 0.69 (0.64-0.88)], and MR-proADM [aOR: 2.85, 95% CI: 1.74-4.69, AUROC: 0.75 (0.69-0.81)]. These biomarkers outperformed clinical variables except diabetes and cancer comorbidities. In this multicenter prospective study that assessed a large panel of biomarkers for COVID-19 patients, CRP, procalcitonin, and MR-proADM were independently associated with the risk of STW. ClinicalTrials.gov NCT04352348.

The effects of applying artificial intelligence to triage in the emergency department: A systematic review of prospective studies.

Accurate and rapid triage can reduce undertriage and overtriage, which may improve emergency department flow. This study aimed to identify the effects of a prospective study applying artificial intelligence-based triage in the clinical field. Systematic review of prospective studies. CINAHL, Cochrane, Embase, PubMed, ProQuest, KISS, and RISS were searched from March 9 to April 18, 2023. All the data were screened independently by three researchers. The review included prospective studies that measured outcomes related to AI-based triage. Three researchers extracted data and independently assessed the study's quality using the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) protocol. Of 1633 studies, seven met the inclusion criteria for this review. Most studies applied machine learning to triage, and only one was based on fuzzy logic. All studies, except one, utilized a five-level triage classification system. Regarding model performance, the feed-forward neural network achieved a precision of 33% in the level 1 classification, whereas the fuzzy clip model achieved a specificity and sensitivity of 99%. The accuracy of the model's triage prediction ranged from 80.5% to 99.1%. Other outcomes included time reduction, overtriage and undertriage checks, mistriage factors, and patient care and prognosis outcomes. Triage nurses in the emergency department can use artificial intelligence as a supportive means for triage. Ultimately, we hope to be a resource that can reduce undertriage and positively affect patient health. We have registered our review in PROSPERO (registration number: CRD 42023415232).

Improving Triage Accuracy in Prehospital Emergency Telemedicine: Scoping Review of Machine Learning-Enhanced Approaches.

Prehospital telemedicine triage systems combined with machine learning (ML) methods have the potential to improve triage accuracy and safely redirect low-acuity patients from attending the emergency department. However, research in prehospital settings is limited but needed; emergency department overcrowding and adverse patient outcomes are increasingly common. In this scoping review, we sought to characterize the existing methods for ML-enhanced telemedicine emergency triage. In order to support future research, we aimed to delineate what data sources, predictors, labels, ML models, and performance metrics were used, and in which telemedicine triage systems these methods were applied. A scoping review was conducted, querying multiple databases (MEDLINE, PubMed, Scopus, and IEEE Xplore) through February 24, 2023, to identify potential ML-enhanced methods, and for those eligible, relevant study characteristics were extracted, including prehospital triage setting, types of predictors, ground truth labeling method, ML models used, and performance metrics. Inclusion criteria were restricted to the triage of emergency telemedicine services using ML methods on an undifferentiated (disease nonspecific) population. Only primary research studies in English were considered. Furthermore, only those studies using data collected remotely (as opposed to derived from physical assessments) were included. In order to limit bias, we exclusively included articles identified through our predefined search criteria and had 3 researchers (DR, JS, and KS) independently screen the resulting studies. We conducted a narrative synthesis of findings to establish a knowledge base in this domain and identify potential gaps to be addressed in forthcoming ML-enhanced methods. A total of 165 unique records were screened for eligibility and 15 were included in the review. Most studies applied ML methods during emergency medical dispatch (7/15, 47%) or used chatbot applications (5/15, 33%). Patient demographics and health status variables were the most common predictors, with a notable absence of social variables. Frequently used ML models included support vector machines and tree-based methods. ML-enhanced models typically outperformed conventional triage algorithms, and we found a wide range of methods used to establish ground truth labels. This scoping review observed heterogeneity in dataset size, predictors, clinical setting (triage process), and reported performance metrics. Standard structured predictors, including age, sex, and comorbidities, across articles suggest the importance of these inputs; however, there was a notable absence of other potentially useful data, including medications, social variables, and health system exposure. Ground truth labeling practices should be reported in a standard fashion as the true model performance hinges on these labels. This review calls for future work to form a standardized framework, thereby supporting consistent reporting and performance comparisons across ML-enhanced prehospital triage systems.

Multisource Data Framework for Prehospital Emergency Triage in Real-Time IoMT-Based Telemedicine Systems

Background and ObjectiveThe Internet of Medical Things (IoMT) has revolutionized telemedicine by enabling the remote monitoring and management of patient care. Nevertheless, the process of regeneration presents the difficulty of effectively prioritizing the information of emergency patients in light of the extensive amount of data generated by several integrated health care devices. The main goal of this study is to be improving the procedure of prioritizing emergency patients by implementing the Real-time Triage Optimization Framework (RTOF), an innovative method that utilizes diverse data from the Internet of Medical Things (IoMT). MethodsThe study’s methodology utilized a variety of Internet of Medical Things (IoMT) data, such as sensor data and texts derived from electronic medical records. Tier 1 supplies sensor and textual data, and Tier 3 imports textual data from electronic medical records. We employed our methodologies to handle and examine data from a sample of 100,000 patients afflicted with hypertension and heart disease, employing artificial intelligence algorithms. We utilized five machine-learning algorithms to enhance the accuracy of triage. ResultsThe RTOF approach has remarkable efficacy in a simulated telemedicine environment, with a triage accuracy rate of 98%. The Random Forest algorithm exhibited superior performance compared to the other approaches under scrutiny. The performance characteristics attained were an accuracy rate of 98%, a precision rate of 99%, a sensitivity rate of 98%, and a specificity rate of 100%. The findings show a significant improvement compared to the present triage methods. ConclusionsThe efficiency of RTOF surpasses that of existing triage frameworks, showcasing its significant ability to enhance the quality and efficacy of telemedicine solutions. This work showcases substantial enhancements compared to existing triage approaches, while also providing a scalable approach to tackle hospital congestion and optimize resource allocation in real-time. The results of our study emphasize the capacity of RTOF to mitigate hospital overcrowding, expedite medical intervention, and enable the creation of adaptable telemedicine networks. This study highlights potential avenues for further investigation into the integration of the Internet of Medical Things (IoMT) with machine learning to develop cutting-edge medical technologies.

Feasibility and acceptability of a serious game to study the effects of environmental distractors on emergency room nurse triage accuracy: A pilot study

BackgroundEmergency triage, which involves complex decision-making under stress and time constraints, may suffer from inaccuracies due to workplace distractions. A serious game was developed to simulate the triage process and environment. A pilot study was undertaken to collect preliminary data on the effects of distractors on emergency nurse triage accuracy. MethodA 2 × 2 factorial randomized controlled trial (RCT) was designed for the study. A sample of 70 emergency room nurses was randomly assigned to three experimental groups exposed to different distractors (noise, task interruptions, and both) and one control group. Nurses had two hours to complete a series of 20 clinical vignettes, in which they had to establish a chief complaint and assign an emergency level. ResultsFifty-five nurses completed approximately 15 vignettes each during the allotted time. No intergroup differences emerged in terms of triage performance. Nurses had a very favorable appreciation of the serious game focusing on triage. ConclusionThe results show that both the structure of our study and the serious game can be used to carry out a future RCT on a larger scale. The lack of a distractor effect raises questions about the frequency and intensity required to find a significant impact on triage performance.

A multi-nudge-based behavioural insight into ward nurses' respiratory rate measurement: An observational study.

This study observed changes in respiratory rate measurement (RRM) and identified barriers and challenges in clinical practice that influence healthcare worker behaviour, aiming to improve RRM in a hospital setting. An observational study was conducted. We observed and analysed changes in the behaviour of healthcare workers at a hospital where multi-nudges were introduced to improve RRM. We checked respiration rate using electronic data and discovered that the original measurement rates were low. Measurement rates rapidly increased after posters were added. Barriers such as time constraints and measurement equipment were also noted. RRM was found to be effective in promoting behavioural economics in medical settings. The results show that incorporating behavioural science principles into medical interventions has the potential to positively change behaviour. By increasing nurses' awareness of respiratory rate measurement and addressing barriers to it, the measurement rate of respiratory rate can also increase, leading to more accurate patient evaluations and triage. WHAT PROBLEM DID THE STUDY ADDRESS?: The proportion of respiratory rate measurements leading to rapid response system (RRS) calls was low. WHAT WERE THE MAIN FINDINGS?: The study observed that a multi-nudge approach effectively changes the behaviour of ward nurses, resulting in enhanced quality of medical care. WHERE AND ON WHOM WILL THE RESEARCH HAVE AN IMPACT?: This research can serve as a valuable reference for leaders promoting healthcare quality projects, by offering a method to encourage behavioural change. This study complied with the EQUATOR guidelines and its reporting adheres to the STROBE checklist. No patient or public contribution.

Artificial Intelligence for Pediatric Emergency Medicine

Pediatric Emergency Medicine (PEM) addresses the unique needs of children in emergencies. This subspecialty faces significant challenges, including the need for specialized training, patient crowding, and the demand for timely and accurate management. Artificial Intelligence (AI) presents promising solutions by enhancing diagnostic precision and operational efficiency. This review examines current trends and prospects of AI in PEM, focusing on its applications, benefits, challenges, and transformative potential. The review highlights AI’s role in overcoming PEM challenges and its future opportunities. Key AI applications in PEM include early sepsis detection, improving triage accuracy, predicting injuries, and supporting diagnostics. AI models show significant potential in forecasting clinical outcomes, optimizing resource management, and improving patient care. Despite these benefits, challenges remain, including the need for specialized training for physicians and the integration of AI systems into clinical practice. Yet, AI holds considerable promise for advancing PEM through enhanced diagnostic tools, more efficient patient management, and improved clinical decision support. Continued advancements and collaborations between AI researchers and pediatric emergency practitioners are essential to fully realize AI’s potential in this field.