Machine learning models for predicting unscheduled return visits of patients with abdominal pain at emergency department and validation during COVID-19 pandemic: A retrospective cohort study.

Machine learning (ML) models for predicting 72-hour unscheduled return visits (URVs) for patients with abdominal pain in the emergency department (ED) were developed in a previous study. This study refined the data to adjust previous prediction models and evaluated the model performance in future data validation during the COVID-19 era. We aimed to evaluate the practicality of the ML models and compare the URVs before and during the COVID-19 pandemic. We used electronic health records from Chang Gung Memorial Hospital from 2018 to 2019 as a training dataset, and various machine learning models, including logistic regression (LR), random forest (RF), extreme gradient boosting (XGB), and voting classifier (VC) were developed and subsequently used to validate against the 2020 to 2021 data. The models highlighted several determinants for 72-hour URVs, including patient age, prior ER visits, specific vital signs, and medical interventions. The LR, XGB, and VC models exhibited the same AUC of 0.71 in the testing set, whereas the VC model displayed a higher F1 score (0.21). The XGB model demonstrated the highest specificity (0.99) and precision (0.64) but the lowest sensitivity (0.01). Among these models, the VC model showed the most favorable, balanced, and comprehensive performance. Despite the promising results, the study illuminated challenges in predictive modeling, such as the unforeseen influences of global events, such as the COVID-19 pandemic. These findings not only highlight the significant potential of machine learning in augmenting emergency care but also underline the importance of iterative refinement in response to changing real-world conditions.

Prognostic value of cardiac troponin in dialysis patients with paroxysmal supraventricular tachycardia.

A rise in cardiac troponin I (cTnI) is common in supraventricular tachycardia (SVT). While troponin elevation in SVT is thought to be a predictor of future adverse events in patients with prior coronary artery disease, the prognostic significance of cTnI in end-stage kidney disease (ESKD) patients with SVT are not known. We aimed to examine the prognostic significance of cTnI in ESKD patients presenting with SVT in the emergency department. This was a retrospective, multiple-center observational study utilizing regularly collected electronic medical records. We screened electronic medical records of all dialysis patients presenting to the emergency departments in 5 hospitals over 12 years with SVT. These patients were divided into whether cTnI was tested, and were further stratified into the cTnI-positive and cTnI-negative groups. The primary outcome of the study was the 3-year risk of major adverse cardiovascular events (MACE). Sixty-two patients were qualified for inclusion. Fifty-seven patients (91.9%) were tested for cTnI, and 5 patients were not. Patients with the cTnI test were older (P = .03) and had a longer length of hospital stay (P < .001). Forty-seven patients (82.5%) had a positive result, and 10 (17.5%) had a negative result. A history of hypertension (P = .013) and decreased left ventricular ejection fraction (P = .048) were the independent predictors of cTnI elevation. After a mean follow-up period of 20.6 ± 14.7 months, there were no differences in 3-year MACE between patients with or without elevated cTnI levels in Kaplan-Meier analysis (P = .34). A history of coronary artery disease was the only independent predictor of 3-year MACE (P = .017). Through the subgroup analysis, a history of coronary artery disease (HR 2.73; CI 1.01-7.41; P = .049) remained an independent risk factor for 3-year MACE in patients with elevated cTnI levels. A large proportion (82.5%) of troponin elevation was observed in ESKD patients with SVT, but it had a poor correlation with MACE.

What is the impact of the COVID-19 pandemic on residency training: a systematic review and analysis.

BACKGROUND: The coronavirus disease 2019 (COVID-19) pandemic has greatly affected medical education in addition to clinical systems. Residency training has probably been the most affected aspect of medical education during the pandemic, and research on this topic is crucial for educators and clinical teachers. The aim of this study was to understand the effect of the COVID-19 pandemic comprehensively through a systematic review and analysis of related published articles. METHODS: A systematic review was conducted based on a predesigned protocol. We searched MEDLINE and EMBASE databases until November 30, 2020, for eligible articles. Two independent reviewers extracted data by using a customized form to record crucial information, and any conflicts between the two reviewers were resolved through discussion with another independent reviewer. The aggregated data were summarized and analyzed. RESULTS: In total, 53 original articles that investigated the effect of the COVID-19 pandemic on residency training were included. Studies from various regions were included in the research, with the largest percentage from the United States (n = 25, 47.2%). Most of these original articles were questionnaire-based studies (n = 44, 83%), and the research target groups included residents (79.55%), program directors (13.64%), or both (6.82%). The majority of the articles (n = 37, 84.0%) were published in countries severely affected by the pandemic. Surgery (n = 36, 67.92%) was the most commonly studied field. CONCLUSIONS: The COVID-19 pandemic has greatly affected residency training globally, particularly surgical and interventional medical fields. Decreased clinical experience, reduced case volume, and disrupted education activities are major concerns. Further studies should be conducted with a focus on the learning outcomes of residency training during the pandemic and the effectiveness of assisted teaching methods.

Risk assessment and prognostic analysis of patients with splenic infarction in emergency department: a multicenter retrospective study.

Splenic infarction is a thromboembolic disease that is frequently missed in acute settings. Previous reviews were rarely presented from a clinical perspective. We aimed to evaluate the clinical characteristics, risk factors with diagnostic value, and prognostic factors using large cohort data and a matched case-control study method. A retrospective medical record review of six hospitals in Taiwan from January 1, 2005, to August 31, 2020, was conducted. All patients who underwent contrast CT with confirmed the diagnosis of splenic infarction were included. Their characteristics were presented and compared to a matched control group with similar presenting characteristics. Prognostic factors were also analyzed. A total of 130 cases were included, two-thirds of whom presented with abdominal pain. Atrial fibrillation was the most common associated predisposing condition, followed by hematologic disease. A higher proportion of tachycardia, positive qSOFA score, history of hypertension or atrial fibrillation, leukocytosis, and thrombocytopenia were found in splenic infarction patients compared to their counterparts. An underlying etiology of infective endocarditis was associated with a higher proportion of ICU admission. Splenic infarction patients often presented with left upper abdominal pain and tachycardia. A history of hypertension, atrial fibrillation, a laboratory result of leukocytosis or thrombocytopenia may provide a clue for clinicians to include splenic infarction in the differential list. Among the patients diagnosed with splenic infarction, those with an underlying etiology of infectious endocarditis may be prone to deterioration or ICU admission.

A Machine Learning Model for Predicting Unscheduled 72 h Return Visits to the Emergency Department by Patients with Abdominal Pain.

Seventy-two-hour unscheduled return visits (URVs) by emergency department patients are a key clinical index for evaluating the quality of care in emergency departments (EDs). This study aimed to develop a machine learning model to predict 72 h URVs for ED patients with abdominal pain. Electronic health records data were collected from the Chang Gung Research Database (CGRD) for 25,151 ED visits by patients with abdominal pain and a total of 617 features were used for analysis. We used supervised machine learning models, namely logistic regression (LR), support vector machine (SVM), random forest (RF), extreme gradient boosting (XGB), and voting classifier (VC), to predict URVs. The VC model achieved more favorable overall performance than other models (AUROC: 0.74; 95% confidence interval (CI), 0.69-0.76; sensitivity, 0.39; specificity, 0.89; F1 score, 0.25). The reduced VC model achieved comparable performance (AUROC: 0.72; 95% CI, 0.69-0.74) to the full models using all clinical features. The VC model exhibited the most favorable performance in predicting 72 h URVs for patients with abdominal pain, both for all-features and reduced-features models. Application of the VC model in the clinical setting after validation may help physicians to make accurate decisions and decrease URVs.

Prediction of emergency department volume and severity during a novel virus pandemic: Experience from the COVID-19 pandemic.

BACKGROUND: During a novel virus pandemic, predicting emergency department (ED) volume is crucial for arranging the limited medical resources of hospitals for balancing the daily patient- and epidemic-related tasks in EDs. The goal of the current study was to detect specific patterns of change in ED volume and severity during a pandemic which would help to arrange medical staff and utilize facilities and resources in EDs in advance in the event of a future pandemic. METHODS: This was a retrospective study of the patients who visited our ED between November 1, 2019 and April 30, 2020. We evaluated the change in ED patient volume and complexity of patients in our medical record system. Patient volume and severity during various periods were identified and compared with data from the past 3 years and the period that SARS occurred. RESULTS: A reduction in ED volume was evident. The reduction began during the early epidemic period and increased rapidly during the peak period of the epidemic with the reduction continuing during the late epidemic period. No significant difference existed in the percentages of triage levels 1 and 2 between the periods. The admission rate, length of stay in the ED, and average number of patients with out-of-hospital cardiac arrest increased during the epidemic periods. CONCLUSION: A significant reduction in ED volume during the COVID-19 pandemic was noted and a predictable pattern was found. This specific change in pattern in the ED volume may be useful for performing adjustments in EDs in the future during a novel virus pandemic. The severity of patients visiting the ED during epidemic periods was inconclusive.

What is the impact of the COVID-19 pandemic on emergency medicine residency training: an observational study.

BACKGROUND: The coronavirus disease 2019 (COVID-19) pandemic has engendered difficulties for health systems globally; however, the effect of the pandemic on emergency medicine (EM) residency training programs is unknown. The pandemic has caused reduced volumes of emergency department (ED) patients, except for those with COVID-19 infections, and this may reduce the case exposure of EM residents. The primary objective of this study was to compare the clinical exposure of EM residents between the prepandemic and pandemic periods. METHODS: This was a retrospective study of EM resident physicians' training in a tertiary teaching hospital with two branch regional hospitals in Taiwan. We retrieved data regarding patients seen by EM residents in the ED between September 1, 2019, and April 30, 2020. The first confirmed COVID-19 case in Taiwan was reported on January 11, so the pandemic period in our study was defined as spanning from February 1, 2020, to April 30, 2020. The number and characteristics of patients seen by residents were recorded. We compared the data between the prepandemic and pandemic periods. RESULTS: The mean number of patients per hour (PPH) seen by EM residents in the adult ED decreased in all three hospitals during the pandemic. The average PPH of critical area of medical ED was 1.68 in the pre-epidemic period and decreased to 1.33 in the epidemic period (p value < 0.001). The average number of patients managed by residents decreased from 1.24 to 0.82 in the trauma ED (p value = 0.01) and 1.56 to 0.51 in the pediatric ED (p value = 0.003) during the pandemic, respectively. The severity of patient illness did not change significantly between the periods. CONCLUSIONS: The COVID-19 pandemic engendered a reduced ED volume and decreased EM residents' clinical exposure. All portion of EM residency training were affected by the pandemic, with pediatric EM being the most affected. The patient volume reduction may persist and in turn reduce patients' case exposure until the pandemic subsides. Adjustment of the training programs may be necessary and ancillary methods of learning should be used to ensure adequate EM residency training.

Three-Minute Step Test for Predicting Acute Mountain Sickness: A Post Hoc Analysis of Rhodiola Crenulata Extract for Prevention of Acute Mountain Sickness, a Randomized, Double-Blind, Placebo-Controlled, Crossover Trial.

BACKGROUND: Acute mountain sickness (AMS) often occurs in individuals who rapidly travel above 2,500 m. As the convenience of traveling and the development of mountain sports increase, AMS will become an increasingly important public health problem. However, no method to effectively predict AMS before it occurs is currently available. METHODS: This post hoc study investigated whether the 3-Minute Step Test (3MST), which evaluates physical fitness, is predictive of AMS development. The data collected in "Rhodiola crenulata extract for prevention of AMS: a randomized, double-blind, placebo-controlled, crossover trial" was used in the analysis. This study collected 204 observations of 102 participants who made two ascents of Hehuan Mountain (3,100 m) by bus within a 3-month period. Participants completed the 3MST at 250 m (before ascent) and 3,100 m (on Hehuan Mountain). The presence of AMS was accessed using the Lake Louise scoring system. RESULTS: AMS was identified in 124 observations (60.78%). In the univariate analysis, the pre-departure 3MST score (at 250 m) was not significantly associated with AMS (p = 0.498), but the 3MST score measured at 3,100 m, ascent number, pulse rate at 3,100 m, and saturation of peripheral oxygen (SpO2) measured at 3,100 m were significantly correlated with the occurrence of AMS (p = 0.002, 0.039, 0.005, < 0.001, respectively). In a further multivariate analysis, only SpO2 measured at 3,100 m had a significant association with AMS (p = 0.016 and 0.006, respectively). The trend analysis showed that for every 1-point increase in the 3MST score at 3,100 m, the AMS decreased by 4% (adjusted odds ratio [AOR] = 0.96, 95% confidence interval [CI] = 0.92-1.01). CONCLUSION: The 3MST score cannot be a predictor of AMS, but it may have a potential role in predicting ascent safety in high-altitude areas.

Rhodiola crenulata extract for prevention of acute mountain sickness: a randomized, double-blind, placebo-controlled, crossover trial.

BACKGROUND: Rhodiola crenulata (R. crenulata) is widely used to prevent acute mountain sickness in the Himalayan areas and in Tibet, but no scientific studies have previously examined its effectiveness. We conducted a randomized, double-blind, placebo-controlled crossover study to investigate its efficacy in acute mountain sickness prevention. METHODS: Healthy adult volunteers were randomized to 2 treatment sequences, receiving either 800 mg R. crenulata extract or placebo daily for 7 days before ascent and 2 days during mountaineering, before crossing over to the alternate treatment after a 3-month wash-out period. Participants ascended rapidly from 250 m to 3421 m on two separate occasions: December 2010 and April 2011. The primary outcome measure was the incidence of acute mountain sickness, as defined by a Lake Louise score ≥ 3, with headache and at least one of the symptoms of nausea or vomiting, fatigue, dizziness, or difficulty sleeping. RESULTS: One hundred and two participants completed the trial. There were no demographic differences between individuals taking Rhodiola-placebo and those taking placebo-Rhodiola. No significant differences in the incidence of acute mountain sickness were found between R. crenulata extract and placebo groups (all 60.8%; adjusted odds ratio (AOR) = 1.02, 95% confidence interval (CI) = 0.69-1.52). The incidence of severe acute mountain sickness in Rhodiola extract vs. placebo groups was 35.3% vs. 29.4% (AOR = 1.42, 95% CI = 0.90-2.25). CONCLUSIONS: R. crenulata extract was not effective in reducing the incidence or severity of acute mountain sickness as compared to placebo. TRIAL REGISTRATION: ClinicalTrials.gov NCT01536288.

The association of leptin and C-reactive protein with the cardiovascular risk factors and metabolic syndrome score in Taiwanese adults.

BACKGROUND: Serum C-reactive protein (CRP) and leptin levels have been independently associated with the cardiovascular risk factors. The aim of the present study was to determine if their serum levels were associated with cardiovascular risk factors or metabolic syndrome as well as their correlation in the Taiwanese population. METHODS: This retrospective study included 999 subjects (> 18 y), who underwent a physical examination in Chang-Gung Memorial Hospital-Linkou and Chiayi in Taiwan. The associations between CRP and/or leptin levels and cardiovascular risk factors and metabolic syndrome were determined using independent two sample t-tests to detect gender differences and chi-square tests to evaluate differences in frequencies. To compare the means of the variables measured among the four groups (high and low leptin and high and low CRP), analysis of variance (ANOVA) was used. RESULTS: Both CRP and leptin levels were independently associated with several cardiovascular risk factors, including diabetes, hypercholesterolemia and metabolic syndrome in both men and women (P < 0.05). In addition, a positive correlation between leptin and CRP levels was observed in both genders. Both high-CRP and high-leptin were associated with high blood glucose, waist circumference and serum triglyceride. Whereas increased metabolic syndrome incidence was observed in males with elevated leptin regardless of CRP levels, females with elevated CRP or leptin had increased incidence of metabolic syndrome. CONCLUSION: Both leptin and CRP levels were associated with cardiovascular risk factors as well as metabolic syndrome score in both men and women although gender-specific differences were observed. Thus, CRP and leptin may represent useful biomarkers for predicting the onset of cardiovascular disease or metabolic syndrome in Taiwanese adults. TRIAL REGISTRATION: IRB/CGMH 100-3514B.

Utility of impedance cardiography for dyspneic patients in the ED.

BACKGROUND: Dyspnea is one of the most common emergency department (ED) symptoms, but early diagnosis and treatment are challenging because of multiple potential causes. Hemodynamic parameters may aid in the evaluation of dyspnea, but are difficult to assess. Impedance cardiography is a noninvasive hemodynamic measurement method that may assist in early ED decision making. METHODS: This study is intended to determine the accuracy in differentiating cardiac from noncardiac causes of dyspnea using impedance cardiography-derived hemodynamic parameters compared to ED physician opinion in light of initial history, and physical and laboratory tests. The final diagnosis, made after patient hospital record review, was compared with ED physician and impedance cardiography diagnoses. RESULTS: A total of 52 patients were included: 14 women and 38 men, aged 68.5 +/- 14.2 years. There were significant differences in values of stroke index (25.7 vs 32.9, P < .05), cardiac index (2.3 vs 3.1, P < .0001), velocity index (35.1 vs 53.2, P < .01), and systolic time ratio (0.55 vs 0.44, P < .05) between the cardiac and noncardiac groups, respectively. Impedance cardiography measurements demonstrated better sensitivity (75% vs 60%), specificity (88% vs 66%), and positive and negative predictive values (79% vs 52% and 85% vs 72%, respectively) compared with those of the ED physician in distinguishing cardiac from noncardiac causes of dyspnea. CONCLUSION: Impedance cardiography data result in improvement in ED physician differentiation of cardiac from noncardiac causes of dyspnea.