Multicentre external validation of the Canadian Syncope Risk Score to predict adverse events and comparison with clinical judgement.

BACKGROUND: The Canadian Syncope Risk Score (CSRS) has been proposed for syncope risk stratification in the emergency department (ED). The aim of this study is to perform an external multicenter validation of the CSRS and to compare it with clinical judgement. METHODS: Using patients previously included in the SyMoNE database, we enrolled subjects older than 18 years who presented reporting syncope at the ED. For each patient, we estimated the CSRS and recorded the physician judgement on the patients' risk of adverse events. We performed a 30-day follow-up. RESULTS: From 1 September 2015 to 28 February 2017, we enrolled 345 patients; the median age was 71 years (IQR 51-81), 174 (50%) were men and 29% were hospitalised. Serious adverse events occurred in 43 (12%) of the patients within 30 days. The area under the curve of the CSRS and clinical judgement was 0.75 (95% CI 0.68 to 0.81) and 0.68 (95% CI 0.61 to 0.74), respectively. The risk of adverse events of patients at low risk according to the CSRS and clinical judgement was 6.7% and 2%, with a sensitivity of 70% (95% CI 54% to 83%) and 95% (95% CI 84% to 99%), respectively. CONCLUSION: This study represents the first validation analysis of CSRS outside Canada. The overall predictive accuracy of the CSRS is similar to the clinical judgement. However, patients at low risk according to clinical judgement had a lower incidence of adverse events as compared with patients at low risk according to the CSRS. Further studies showing that the adoption of the CSRS improve patients' outcomes is warranted before its widespread implementation.

Machine Learning and Syncope Management in the ED: The Future Is Coming.

In recent years, machine learning (ML) has been promisingly applied in many fields of clinical medicine, both for diagnosis and prognosis prediction. Aims of this narrative review were to summarize the basic concepts of ML applied to clinical medicine and explore its main applications in the emergency department (ED) setting, with a particular focus on syncope management. Through an extensive literature search in PubMed and Embase, we found increasing evidence suggesting that the use of ML algorithms can improve ED triage, diagnosis, and risk stratification of many diseases. However, the lacks of external validation and reliable diagnostic standards currently limit their implementation in clinical practice. Syncope represents a challenging problem for the emergency physician both because its diagnosis is not supported by specific tests and the available prognostic tools proved to be inefficient. ML algorithms have the potential to overcome these limitations and, in the future, they could support the clinician in managing syncope patients more efficiently. However, at present only few studies have addressed this issue, albeit with encouraging results.

Syncope Time Frames for Adverse Events after Emergency Department Presentation: An Individual Patient Data Meta-Analysis.

Background and Objectives: Knowledge of the incidence and time frames of the adverse events of patients presenting syncope at the ED is essential for developing effective management strategies. The aim of the present study was to perform a meta-analysis of the incidence and time frames of adverse events of syncope patients. Materials and Methods: We combined individual patients' data from prospective observational studies including adult patients who presented syncope at the ED. We assessed the pooled rate of adverse events at 24 h, 72 h, 7-10 days, 1 month and 1 year after ED evaluation. Results: We included nine studies that enrolled 12,269 patients. The mean age varied between 53 and 73 years, with 42% to 57% females. The pooled rate of adverse events was 5.1% (95% CI 3.4% to 7.7%) at 24 h, 7.0% (95% CI 4.9% to 9.9%) at 72 h, 8.4% (95% CI 6.2% to 11.3%) at 7-10 days, 10.3% (95% CI 7.8% to 13.3%) at 1 month and 21.3% (95% CI 15.8% to 28.0%) at 1 year. The pooled death rate was 0.2% (95% CI 0.1% to 0.5%) at 24 h, 0.3% (95% CI 0.1% to 0.7%) at 72 h, 0.5% (95% CI 0.3% to 0.9%) at 7-10 days, 1% (95% CI 0.6% to 1.7%) at 1 month and 5.9% (95% CI 4.5% to 7.7%) at 1 year. The most common adverse event was arrhythmia, for which its rate was 3.1% (95% CI 2.0% to 4.9%) at 24 h, 4.8% (95% CI 3.5% to 6.7%) at 72 h, 5.8% (95% CI 4.2% to 7.9%) at 7-10 days, 6.9% (95% CI 5.3% to 9.1%) at 1 month and 9.9% (95% CI 5.5% to 17) at 1 year. Ventricular arrhythmia was rare. Conclusions: The risk of death or life-threatening adverse event is rare in patients presenting syncope at the ED. The most common adverse events are brady and supraventricular arrhythmias, which occur during the first 3 days. Prolonged ECG monitoring in the ED in a short stay unit with ECG monitoring facilities may, therefore, be beneficial.

Diagnostic accuracy of ICD-9 code 780.2 for the identification of patients with syncope in the emergency department.

PURPOSE: Syncope is a common condition that affects individuals of all ages and is responsible for 1-3% of all emergency department (ED) visits. Prospective studies on syncope are often limited by the exiguous number of subjects enrolled. A possible alternative approach would be to use of hospital discharge diagnoses from administrative databases to identify syncope subjects in epidemiological observational studies. We assessed the accuracy of the International Classification of Diseases, Ninth Revision (ICD-9) code 780.2 "syncope and collapse" to identify patients with syncope. METHODS: Patients in two teaching hospitals in Milan, Italy with a triage assessment for ED access that was possibly related to syncope were recruited in this study. We considered the index test to be the attribution of the ICD-9 code 780.2 at ED discharge and the reference standard to be the diagnosis of syncope by the ED physician. RESULTS: The sensitivity, specificity, positive and negative predictive values of the ICD-9 code 780.2 to identify patients with syncope were 0.63 (95% confidence interval [CI] 0.58-0.67), 0.98 (95% CI 0.98-0.99), 0.83 (95% CI 0.79-0.87) and 0.95 (95% CI 0.94-0.95), respectively. CONCLUSIONS: The moderate sensitivity of ICD-9 code 780.2 should be considered when the code is used to identify patients with syncope through administrative databases.

Implantable loop recorder versus conventional diagnostic workup for unexplained recurrent syncope.

BACKGROUND: The most recent syncope guideline recommends that implantable loop recorders (ILRs) are implanted in the early phase of evaluation of people with recurrent syncope of uncertain origin in the absence of high-risk criteria, and in high-risk patients after a negative evaluation. Observational and case-control studies have shown that loop recorders lead to earlier diagnosis and reduce the rate of unexplained syncopes, justifying their use in clinical practice. However, only randomised clinical trials with an emphasis on a primary outcome of specific ILR-guided diagnosis and therapy, rather than simply electrocardiogram (ECG) diagnosis, might change clinical practice. OBJECTIVES: To assess the incidence of mortality, quality of life, adverse events and costs of ILRs versus conventional diagnostic workup in people with unexplained syncope. SEARCH METHODS: We searched the Cochrane Central Register of Controlled Trials (CENTRAL, Issue 3, 2015), MEDLINE, EMBASE, ClinicalTrials.gov and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) Search Portal in April 2015. No language restriction was applied. SELECTION CRITERIA: We included all randomised controlled trials of adult participants (i.e. ≥ 18 years old) with a diagnosis of unexplained syncope comparing ILR with standard diagnostic workup. DATA COLLECTION AND ANALYSIS: Two independent review authors screened titles and abstracts of all potential studies we identified as a result of the literature search, extracted study characteristics and outcome data from included studies and assessed risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions. We contacted authors of trials for missing data. We analysed dichotomous data (all-cause mortality and aetiologic diagnosis) as risk ratios (RR) with 95% confidence intervals (CI). We used the Chi(2) test to assess statistical heterogeneity (with P < 0.1) and the I² statistic to measure heterogeneity among the trials. We created a 'Summary of findings' table using the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness and publication bias) to assess the quality of a body of evidence as it relates to the studies which contribute data to the meta-analyses for the prespecified outcomes. MAIN RESULTS: We included four trials involving a total of 579 participants. With the limitation that only two studies reported data on mortality and none of them had considered death as a primary endpoint, the meta-analysis showed no evidence of a difference in the risk of long-term mortality between participants who received ILR and those who were managed conventionally at follow-up (RR 0.97, 95% CI 0.41 to 2.30; participants = 255; studies = 2; very low quality evidence) with no evidence of heterogeneity. No data on short term mortality were available. Two studies reported data on adverse events after ILR implant. Due to the lack of data on adverse events in one of the studies' arms, a formal meta-analysis was not performed for this outcome.Data from two trials seemed to show no difference in quality of life, although this finding was not supported by a formal analysis due to the differences in both the scores used and the way the data were reported. Data from two studies seemed to show a trend towards a reduction in syncope relapses after diagnosis in participants implanted with ILR. Cost analyses from two studies showed higher overall mean costs in the ILR group, if the costs incurred by the ILR implant were counted. The mean cost per diagnosis and the mean cost per arrhythmic diagnosis were lower for participants randomised to ILR implant.Participants who underwent ILR implantation experienced higher rates of diagnosis (RR (in favour of ILR) 0.61, 95% CI 0.54 to 0.68; participants = 579; studies = 4; moderate quality evidence), as compared to participants in the standard assessment group, with no evidence of heterogeneity. AUTHORS' CONCLUSIONS: Our systematic review shows that there is no evidence that an ILR-based diagnostic strategy reduces long-term mortality as compared to a standard diagnostic assessment (very low quality evidence). No data were available for short-term all-cause mortality. Moderate quality evidence shows that an ILR-based diagnostic strategy increases the rate of aetiologic diagnosis as compared to a standard diagnostic pathway. No conclusive data were available on the other end-points analysed.Further trials evaluating the effect of ILRs in the diagnostic strategy of people with recurrent unexplained syncope are warranted. Future research should focus on the assessment of the ability of ILRs to change clinically relevant outcomes, such as quality of life, syncope relapse and costs.

Relationship between sympathetic activity and pain intensity in fibromyalgia.

OBJECTIVES: Fibromyalgia (FM) is a syndrome characterised by chronic musculoskeletal pain, hyperalgesia on specific areas of tenderness (tender points) and by an autonomic nervous system dysfunction consistent with sympathetic overactivity. It is not known whether there is any relationship between the amount of cardiovascular sympathetic activity and the magnitude of pain. Our objective was to assess this potential relationship in patients with FM. METHODS: Electrocardiogram, finger blood pressure, respiration and post-ganglionic sympathetic discharge activity (muscle sympathetic nerve activity, MSNA) were continuously recorded at rest in 25 patients with primary FMS. The autonomic profile was assessed by MSNA and spectral indices of cardiac sympathetic (LFRR) and vagal (HFRR) modulation and of sympathetic vasomotor control (LF-SAP) computed by spectrum analysis of RR and systolic arterial pressure (SAP) variability. Cardiac baroreflex function was evaluated by the index α (αLF). Baroreceptor modulation of the sympathetic vasomotor control (sBRS) was assessed by the MSNA/diastolic pressure relationship. RESULTS: Pain intensity was linearly correlated with LFRR/HFRR (r² = 0.21; p=0.03), LFSAP (r² = 0.26; p=0.02) and MSNA (burst rate) (r² = 0.45; p=0.003). Pain intensity was inversely correlated with the αLF index (r² = 0.24; p=0.02) and the sBRS (r² = 0.28; p=0.03). Thus, the higher the sympathetic drive to the heart and vessels, the higher the magnitude of chronic pain. Also, the gains of both the cardiac and MSNA baroreceptor control were inversely related to the pain intensity. CONCLUSIONS: These findings raise the theoretical possibility that in FM patients the use of anti-adrenergic agents might lessen chronic pain intensity by reducing the underlying excessive sympathetic activity.

Syncope recurrence and mortality: a systematic review.

AIMS: Data on adverse events and death rates following syncope are heterogeneous among studies, and knowledge of syncope prognosis could help to better define the correct management of patients. METHODS AND RESULTS: We performed a systematic review of literature by searching for prospective observational studies enrolling consecutive patients presenting to the Emergency Department because of syncope. The outcomes considered were syncope recurrence and short- and long-term mortality. Morbidity and a composite of morbidity and mortality were also assessed. Pooled event rates and 95% confidence intervals (CI) were calculated for each outcome using the random effects model. Twenty-five studies (11 158 patients) were included. The incidence of syncope relapse linearly increased from 0.3% at 30 days to 22% at 2 years follow-up. One-year mortality rate varied between 5.7 and 15.5%; the pooled estimate was 8.4% (95% CI: 6.7-10.2%). The incidence of adverse events (morbidity) varied between 6.1 and 25.2% at 10 days and 2 years, respectively. The short-term (10 days) pooled incidence of the composite of morbidity and mortality was 9.1% (95% CI: 6.6-12.5%). We found a high statistical heterogeneity between studies. CONCLUSION: This meta-analysis of prospective observational studies shows that the chance of being asymptomatic linearly progressively decreased over time after the first syncope. Short-term (10-30 days) mortality after syncope was <2% and that the overall 10-day rate of the composite endpoint of death and major events was ∼9%. The knowledge of syncope prognosis could help clinicians to understand syncope patients' prognosis and researchers to design future studies.

Priorities for emergency department syncope research.

STUDY OBJECTIVES: There is limited evidence to guide the emergency department (ED) evaluation and management of syncope. The First International Workshop on Syncope Risk Stratification in the Emergency Department identified key research questions and methodological standards essential to advancing the science of ED-based syncope research. METHODS: We recruited a multinational panel of syncope experts. A preconference survey identified research priorities, which were refined during and after the conference through an iterative review process. RESULTS: There were 31 participants from 7 countries who represented 10 clinical and methodological specialties. High-priority research recommendations were organized around a conceptual model of ED decisionmaking for syncope, and they address definition, cohort selection, risk stratification, and management. CONCLUSION: We convened a multispecialty group of syncope experts to identify the most pressing knowledge gaps and defined a high-priority research agenda to improve the care of patients with syncope in the ED.

A large language model-based clinical decision support system for syncope recognition in the emergency department: A framework for clinical workflow integration.

Differentiation of syncope from transient loss of consciousness can be challenging in the emergency department (ED). Natural Language Processing (NLP) enables the analysis of free text in the electronic medical records (EMR). The present paper aimed to develop a large language models (LLM) for syncope recognition in the ED and proposed a framework for model integration within the clinical workflow. Two models, based on both the Italian and Multilingual Bidirectional Encoder Representations from Transformers (BERT) language model, were developed using consecutive EMRs. The "triage" model was only based on notes contained in the "triage" section of the EMR. The "anamnesis" model added data contained in the "medical history" section. Interpretation and calibration plots were generated. The Italian and Multi BERT models were developed and tested on both 15,098 and 15,222 EMRs, respectively. The triage model had an AUC of 0·95 for the Italian BERT and 0·94 for the Multi BERT. The anamnesis model had an AUC of 0·98 for the Italian BERT and 0·97 for Multi BERT. The LLM identified syncope when not explicitly mentioned in the EMR and also recognized common prodromal symptoms preceding syncope. Both models identified syncope patients in the ED with a high discriminative capability from nurses and doctors' notes, thus potentially acting as a tool helping physicians to differentiate syncope from others transient loss of consciousness.

Will Artificial Intelligence Be "Better" Than Humans in the Management of Syncope?

Clinical decision-making regarding syncope poses challenges, with risk of physician error due to the elusive nature of syncope pathophysiology, diverse presentations, heterogeneity of risk factors, and limited therapeutic options. Artificial intelligence (AI)-based techniques, including machine learning (ML), deep learning (DL), and natural language processing (NLP), can uncover hidden and nonlinear connections among syncope risk factors, disease features, and clinical outcomes. ML, DL, and NLP models can analyze vast amounts of data effectively and assist physicians to help distinguish true syncope from other types of transient loss of consciousness. Additionally, short-term adverse events and length of hospital stay can be predicted by these models. In syncope research, AI-based models shift the focus from causality to correlation analysis between entities. This prompts the search for patterns rather than defining a hypothesis to be tested a priori. Furthermore, education of students, doctors, and health care providers engaged in continuing medical education may benefit from clinical cases of syncope interacting with NLP-based virtual patient simulators. Education may be of benefit to patients. This article explores potential strengths, weaknesses, and proposed solutions associated with utilization of ML and DL in syncope diagnosis and management. Three main topics regarding syncope are addressed: 1) clinical decision-making; 2) clinical research; and 3) education. Within each domain, we question whether "AI will be better than humans," seeking evidence to support our objective inquiry.

A Hybrid Model for 30-Day Syncope Prognosis Prediction in the Emergency Department.

Syncope is a challenging problem in the emergency department (ED) as the available risk prediction tools have suboptimal predictive performances. Predictive models based on machine learning (ML) are promising tools whose application in the context of syncope remains underexplored. The aim of the present study was to develop and compare the performance of ML-based models in predicting the risk of clinically significant outcomes in patients presenting to the ED for syncope. We enrolled 266 consecutive patients (age 73, IQR 58-83; 52% males) admitted for syncope at three tertiary centers. We collected demographic and clinical information as well as the occurrence of clinically significant outcomes at a 30-day telephone follow-up. We implemented an XGBoost model based on the best-performing candidate predictors. Subsequently, we integrated the XGboost predictors with knowledge-based rules. The obtained hybrid model outperformed the XGboost model (AUC = 0.81 vs. 0.73, p < 0.001) with acceptable calibration. In conclusion, we developed an ML-based model characterized by a commendable capability to predict adverse events within 30 days post-syncope evaluation in the ED. This model relies solely on clinical data routinely collected during a patient's initial syncope evaluation, thus obviating the need for laboratory tests or syncope experienced clinical judgment.

Multimodal deep learning for COVID-19 prognosis prediction in the emergency department: a bi-centric study.

Predicting clinical deterioration in COVID-19 patients remains a challenging task in the Emergency Department (ED). To address this aim, we developed an artificial neural network using textual (e.g. patient history) and tabular (e.g. laboratory values) data from ED electronic medical reports. The predicted outcomes were 30-day mortality and ICU admission. We included consecutive patients from Humanitas Research Hospital and San Raffaele Hospital in the Milan area between February 20 and May 5, 2020. We included 1296 COVID-19 patients. Textual predictors consisted of patient history, physical exam, and radiological reports. Tabular predictors included age, creatinine, C-reactive protein, hemoglobin, and platelet count. TensorFlow tabular-textual model performance indices were compared to those of models implementing only tabular data. For 30-day mortality, the combined model yielded slightly better performances than the tabular fastai and XGBoost models, with AUC 0.87 ± 0.02, F1 score 0.62 ± 0.10 and an MCC 0.52 ± 0.04 (p < 0.32). As for ICU admission, the combined model MCC was superior (p < 0.024) to the tabular models. Our results suggest that a combined textual and tabular model can effectively predict COVID-19 prognosis which may assist ED physicians in their decision-making process.

Can Artificial Intelligence Enhance Syncope Management?: A JACC: Advances Multidisciplinary Collaborative Statement.

Syncope, a form of transient loss of consciousness, remains a complex medical condition for which adverse cardiovascular outcomes, including death, are of major concern but rarely occur. Current risk stratification algorithms have not completely delineated which patients benefit from hospitalization and specific interventions. Patients are often admitted unnecessarily and at high cost. Artificial intelligence (AI) and machine learning may help define the transient loss of consciousness event, diagnose the cause, assess short- and long-term risks, predict recurrence, and determine need for hospitalization and therapeutic intervention; however, several challenges remain, including medicolegal and ethical concerns. This collaborative statement, from a multidisciplinary group of clinicians, investigators, and scientists, focuses on the potential role of AI in syncope management with a goal to inspire creation of AI-derived clinical decision support tools that may improve patient outcomes, streamline diagnostics, and reduce health-care costs.

Learning Analytics Applied to Clinical Diagnostic Reasoning Using a Natural Language Processing-Based Virtual Patient Simulator: Case Study.

BACKGROUND: Virtual patient simulators (VPSs) log all users' actions, thereby enabling the creation of a multidimensional representation of students' medical knowledge. This representation can be used to create metrics providing teachers with valuable learning information. OBJECTIVE: The aim of this study is to describe the metrics we developed to analyze the clinical diagnostic reasoning of medical students, provide examples of their application, and preliminarily validate these metrics on a class of undergraduate medical students. The metrics are computed from the data obtained through a novel VPS embedding natural language processing techniques. METHODS: A total of 2 clinical case simulations (tests) were created to test our metrics. During each simulation, the students' step-by-step actions were logged into the program database for offline analysis. The students' performance was divided into seven dimensions: the identification of relevant information in the given clinical scenario, history taking, physical examination, medical test ordering, diagnostic hypothesis setting, binary analysis fulfillment, and final diagnosis setting. Sensitivity (percentage of relevant information found) and precision (percentage of correct actions performed) metrics were computed for each issue and combined into a harmonic mean (F1), thereby obtaining a single score evaluating the students' performance. The 7 metrics were further grouped to reflect the students' capability to collect and to analyze information to obtain an overall performance score. A methodological score was computed based on the discordance between the diagnostic pathway followed by students and the reference one previously defined by the teacher. In total, 25 students attending the fifth year of the School of Medicine at Humanitas University underwent test 1, which simulated a patient with dyspnea. Test 2 dealt with abdominal pain and was attended by 36 students on a different day. For validation, we assessed the Spearman rank correlation between the performance on these scores and the score obtained by each student in the hematology curricular examination. RESULTS: The mean overall scores were consistent between test 1 (mean 0.59, SD 0.05) and test 2 (mean 0.54, SD 0.12). For each student, the overall performance was achieved through a different contribution in collecting and analyzing information. Methodological scores highlighted discordances between the reference diagnostic pattern previously set by the teacher and the one pursued by the student. No significant correlation was found between the VPS scores and hematology examination scores. CONCLUSIONS: Different components of the students' diagnostic process may be disentangled and quantified by appropriate metrics applied to students' actions recorded while addressing a virtual case. Such an approach may help teachers provide students with individualized feedback aimed at filling competence drawbacks and methodological inconsistencies. There was no correlation between the hematology curricular examination score and any of the proposed scores as these scores address different aspects of students' medical knowledge.

Comparison between standard and ultrasound-integrated approach for risk stratification of syncope in the emergency department.

This prospective cohort enrolled all patients above 16 years of age presenting to the in the emergency department (ED) for a reported syncope was designed to test the accuracy of a point-of-care ultrasound (POCUS) integrated approach in risk stratification. The emergency physician responsible for the patient care was asked to classify the syncope risk after the initial clinical assessment and after performing POCUS evaluation. All risk group definitions were based on the 2018 European Society of Cardiology guidelines. Thirty days after the index event, all participants were followed up to assess the frequency of short-term serious outcomes as defined in the San Francisco Syncope Rule (SFSR) cohorts. We estimated the accuracy of clinical and POCUS-integrated evaluation in predicting SFSR outcomes. Between February 2016 and January 2018, 196 patients were enrolled [109 women (55.6%)]. Median age was 64 years (interquartile range 31 years). After a follow-up of 30 days, 19 patients experienced 20 SFSR outcomes. Positive and negative likelihood ratios were 1.73 (95% CI 0.87-3.44) and 0.84 (95% CI 0.62-1.12) for the clinical evaluation, and 5.93 (95% CI 2.83-12.5) and 0.63 (95% CI 0.45-0.9) for the POCUS-integrated evaluation. The POCUS-integrated approach would reduce the diagnostic error of the clinical evaluation by 4.5 cases/100 patients. This cohort study suggested that the integration of the clinical assessment with POCUS results in patients presenting to the ED for non-high-risk syncope may increase the accuracy of predicting the risk of SFSR outcomes and the usefulness of the clinical assessment alone.

A Natural Language Processing-Based Virtual Patient Simulator and Intelligent Tutoring System for the Clinical Diagnostic Process: Simulator Development and Case Study.

BACKGROUND: Shortage of human resources, increasing educational costs, and the need to keep social distances in response to the COVID-19 worldwide outbreak have prompted the necessity of clinical training methods designed for distance learning. Virtual patient simulators (VPSs) may partially meet these needs. Natural language processing (NLP) and intelligent tutoring systems (ITSs) may further enhance the educational impact of these simulators. OBJECTIVE: The goal of this study was to develop a VPS for clinical diagnostic reasoning that integrates interaction in natural language and an ITS. We also aimed to provide preliminary results of a short-term learning test administered on undergraduate students after use of the simulator. METHODS: We trained a Siamese long short-term memory network for anamnesis and NLP algorithms combined with Systematized Nomenclature of Medicine (SNOMED) ontology for diagnostic hypothesis generation. The ITS was structured on the concepts of knowledge, assessment, and learner models. To assess short-term learning changes, 15 undergraduate medical students underwent two identical tests, composed of multiple-choice questions, before and after performing a simulation by the virtual simulator. The test was made up of 22 questions; 11 of these were core questions that were specifically designed to evaluate clinical knowledge related to the simulated case. RESULTS: We developed a VPS called Hepius that allows students to gather clinical information from the patient's medical history, physical exam, and investigations and allows them to formulate a differential diagnosis by using natural language. Hepius is also an ITS that provides real-time step-by-step feedback to the student and suggests specific topics the student has to review to fill in potential knowledge gaps. Results from the short-term learning test showed an increase in both mean test score (P<.001) and mean score for core questions (P<.001) when comparing presimulation and postsimulation performance. CONCLUSIONS: By combining ITS and NLP technologies, Hepius may provide medical undergraduate students with a learning tool for training them in diagnostic reasoning. This may be particularly useful in a setting where students have restricted access to clinical wards, as is happening during the COVID-19 pandemic in many countries worldwide.

Correction: Personalized risk stratification through attribute matching for clinical decision making in clinical conditions with aspecific symptoms: The example of syncope.

[This corrects the article DOI: 10.1371/journal.pone.0228725.].

San Francisco Syncope Rule, Osservatorio Epidemiologico sulla Sincope nel Lazio risk score, and clinical judgment in the assessment of short-term outcome of syncope.

OBJECTIVE: The study aimed to compare the efficacy of the Osservatorio Epidemiologico sulla Sincope nel Lazio (OESIL) risk score, San Francisco Syncope Rule, and clinical judgment in assessing the short-term prognosis of syncope. METHODS: We studied 488 patients consecutively seen for syncope at the emergency department of 2 general hospitals between January and July 2004. Sensitivity, specificity, predictive values, and likelihood ratios for short-term (within 10 days) severe outcomes were computed for each decision rule and clinical judgment. Severe outcomes comprised death, major therapeutic procedures, and early readmission to hospital. RESULTS: Clinical judgment had a sensitivity of 77%, a specificity of 69%, and would have admitted less patients (34%, P < .05 vs decision rules). The OESIL risk score was characterized by a sensitivity of 88% and a specificity of 60% (admission 43%). San Francisco Syncope Rule sensitivity was 81% and specificity was 63% (admission 40%). According to both clinical rules, no discharged patient would have died. With combined OESIL risk score and clinical judgment, the probability of adverse events was 0.7% for patients with both low risk scores, whereas that for both high risk scores was roughly 16%. CONCLUSION: Because of a relatively low sensitivity, both risk scores were partially lacking in recognizing patients with short-term high-risk syncope. However, the application of the decision rules would have identified all patients who subsequently died, and OESIL risk score and clinical judgment combined seem to improve the decision-making process concerning the identification of high-risk patients who deserve admission.

Psychogenic Pseudosyncope: Clinical Features, Diagnosis and Management.

Psychogenic Pseudosyncope (PPS) is the appearance of Transient Loss of Consciousness (TLOC) in which movements are absent, but there are no hemodynamic and electroencephalographic modifications as are induced by gravitational challenges which characterize syncope and true loss of consciousness. For younger and adult populations, a detailed history is crucial for the diagnosis. Clinical clues that should raise the suspicion for PPS include prolonged duration of the LOC, eye closure during the episode, unusual triggers, no recognizable prodromes and the high frequency of attacks. The presence of an established diagnosis of syncope should not deter from the concomitant diagnosis of PPS. The gold standard for a proper diagnosis of PPS is the documentation by a tilt test of normal hemodynamic and electroencephalographic parameters, when recorded during an attack. Treatment of PPS, based on the clear and empathetic communication of the diagnosis, can lead to an immediate reduction of attack frequency and lower the need to call on emergency services. Pharmacological treatment of associated psychiatric disorders and psychological interventions may be beneficial in patients with PPS. Cognitive-behavioural therapy holds the most reliable evidence of efficacy. In the present review, we aimed to address PPS with historical aspects, main clinical features and diagnostic tests, current diagnostic classification, underlying neurobiological abnormalities, management and therapy.