|Thomas A. Gibson,||University of California, Los Angeles, Department of Biostatistics, Los Angeles, California|
|Robert E. Weiss, PhD||University of California, Los Angeles, Department of Biostatistics, Los Angeles, California|
|Benjamin C. Sun, MD, MPP||Oregon Heath & Science University, Center for Policy and Research in Emergency Medicine, Department of Emergency Medicine, Portland, Oregon|
We performed a systematic review and meta-analysis to identify predictors of serious clinical outcomes after an acute-care evaluation for syncope.
We identified studies that assessed for predictors of short-term (≤30 days) serious clinical events after an emergency department (ED) visit for syncope. We performed a MEDLINE search (January 1, 1990 – July 1, 2017) and reviewed reference lists of retrieved articles. The primary outcome was the occurrence of a serious clinical event (composite of mortality, arrhythmia, ischemic or structural heart disease, major bleed, or neurovascular event) within 30 days. We estimated the sensitivity, specificity, and likelihood ratio of findings for the primary outcome. We created summary estimates of association on a variable-by-variable basis using a Bayesian random-effects model.
We reviewed 2,773 unique articles; 17 met inclusion criteria. The clinical findings most predictive of a short-term, serious event were the following: 1) An elevated blood urea nitrogen level (positive likelihood ratio [LR+]: 2.86, 95% confidence interval [CI] [1.15, 5.42]); 2); history of congestive heart failure (LR+: 2.65, 95%CI [1.69, 3.91]); 3) initial low blood pressure in the ED (LR+: 2.62, 95%CI [1.12, 4.9]); 4) history of arrhythmia (LR+: 2.32, 95%CI [1.31, 3.62]); and 5) an abnormal troponin value (LR+: 2.49, 95%CI [1.36, 4.1]). Younger age was associated with lower risk (LR−: 0.44, 95%CI [0.25, 0.68]). An abnormal electrocardiogram was mildly predictive of increased risk (LR+ 1.79, 95%CI [1.14, 2.63]).
We identified specific risk factors that may aid clinical judgment and that should be considered in the development of future risk-prediction tools for serious clinical events after an ED visit for syncope.
There are over 1.3 million annual events of syncope (transient loss of consciousness with rapid and spontaneous recovery1) in the United States that lead to an emergency department (ED) visit, resulting in 440,000 admissions2 and $2.4 billion in yearly hospital costs.3 Syncope may be a harbinger of sudden death, dangerous arrhythmias, or other serious medical conditions (e.g., pulmonary embolism). Evaluation of syncope is challenging as symptoms have resolved by the time patients seek medical evaluation, and fewer than 10% of ED evaluations reveal a serious condition that may explain the episode of syncope.4 To mitigate the risk of sudden death or other dangerous clinical events, up to 85%5 of older adults who present with syncope of unclear cause are hospitalized for a diagnostic evaluation.6,7 However, admission is associated with low diagnostic and therapeutic yield,8,9 and there is no evidence that current practice patterns improve quality of life or long-term survival.10
According to the 2017 American College of Cardiology (ACC) / American Heart Association (AHA) / Heart Rhythm Society (HRS) Syncope Guidelines, accurate risk prediction using the clinical examination (including history taking, physical exam, and a 12-lead electrocardiogram [ECG]) is fundamental to guiding diagnostic and disposition decisions.1 For ED decision-making, the focus is on predicting the risk of short- term (≤30 days) events that may warrant immediate hospitalization and testing.11,12 Unfortunately, unstructured physician judgment appears to have poor reliability and accuracy.13
Multiple research groups have proposed objective risk-stratification scores, each with different combinations of predictors.4, 14-21 However, meta-analyses of published tools suggest equivalent performance compared to unstructured provider judgment.22, 23 The 2017 ACC/ AHA/ HRS Syncope Guidelines1 identified several limitations of the existing literature, including small sample sizes that limit the reliability of prediction models.24 Although the 2017 Guidelines did identify potential predictors of short- and long-term clinical adverse events, diagnostic test characteristics of such factors were not described. Thus, it may not be apparent to front-line clinicians how to weight the presence or absence of high-risk factors identified in the clinical examination. In addition, prior attempts to develop risk scores have not incorporated existing information about potential predictors. A Bayesian approach that includes prior risk-stratification data may potentially result in the development of more reliable and valid risk scores.
Goal of this Investigation
To address this evidence gap, we performed a systematic review and meta-analysis to describe the diagnostic test characteristics of initial history, physical exam, ECG, and selected tests.
We performed a systematic literature review and meta-analysis that complied with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines.25 This study was exempt from institutional review board review.
Search Strategy and Study Selection
We conducted English-language searches in MEDLINE using a combination of the terms “syncope” and “risk” in any field. We considered papers published between January 1, 1990, and July 1, 2017. We also manually reviewed the reference lists of relevant studies. Titles and abstracts for all articles were screened by one of the authors (BCS). The full text of potentially eligible studies was reviewed by the entire study team.
Inclusion criteria included the following: 1) Patients presenting to an ED after a syncope event; 2) serious, short-term (≤30 days) clinical outcomes (such as death, arrhythmia, structural/ ischemic cardiac events, neurovascular events, major bleeding) were reported12; and 3) data sufficient for potential predictors stratified by the presence of adverse outcomes. For univariate data, this meant either (a) counts of the number of patients with a given characteristic (e.g., abnormal ECG) that did have an adverse event and the number of patients with the characteristic that did not have adverse events, along with the total number of patients with and without adverse events (i.e., contingency table counts for adverse outcomes by the presence of the characteristic), or (b) odds ratios and confidence intervals. For multivariate data, we required output from a multiple logistic regression model in (a) coefficient estimates and standard errors, or (b) odd ratios (OR) and confidence intervals (CI). All the papers that we included intended to either identify predictors of adverse events after syncope or validate existing risk-stratification methods.
Population Health Research Capsule
What do we already know about this issue?
The emergency department evaluation relies on careful history, examination, electrocardiogram, and selective testing to evaluate patients presenting with syncope.
What was the research question?
What are the predictors of serious clinical outcomes after an acute-care evaluation for syncope?
What was the major finding of the study?
The strongest predictors included age, cardiac co-morbidities, hypotension, and abnormal biomarkers.
How does this improve population health?
We identified specific risk factors that may aid in clinical evaluation and risk prediction for patients undergoing ED evaluation for syncope.
We excluded studies that included non-ED referral populations (e.g., arrhythmia clinic, falls and syncope unit) or that were conditioned on prior testing (e.g. electrophysiology or tilt-table test) or pre-existing co-morbidities (e.g., prior cardiac arrest, cardiomyopathy). Because the focus of this review was to aid short-term decision-making relevant in the ED, we excluded studies that only had data on long-term (>30 days) outcomes. We excluded studies that had implausible results (defined as OR >20).
For each included article, we evaluated for potential bias with a modified version of the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) Criteria,26 a meta-analytic system devised specifically for diagnostic studies. Two members of the study team (TAG, BCS) independently reviewed the methodological quality of included studies, and disagreements were settled through consensus.
Papers provided varying degrees of covariate information, and the method of extraction varied with the type of information: direct, for papers that provided exact counts for contingency tables; direct after rounding, for papers that provided percentages instead of counts; and extrapolated, for papers that provided ORs and 95% CIs (see Appendix A for detailed description of methodology). We analyzed all variables for which at least two papers provided information.
We calculated sensitivity, specificity, and likelihood ratio positive (LR+) and negative (LR−) for predictor variables. We did not assess diagnostic performance of specific risk scores, as two meta-analyses have previously addressed this issue.22,23 A Bayesian random effects model was used for meta-analyses. The Bayesian approach allows for direct probability statements to be made about quantities of interest; for example, the probability that the positive likelihood ratio for a given variable is above some diagnostic threshold. Additionally, all parameter uncertainty is accounted for automatically in each analysis (see Appendix A for detailed description). Positive and negative likelihood ratios were deemed statistically significant if the 95% posterior interval did not contain 1.0. All analyses were implemented in the R statistical package.27
We numerically assessed the level of heterogeneity of effect sizes via the mean posterior standard deviation of random effects on the log-odds ratio scale. Smaller (larger) values indicate less (more) heterogeneity. Values very close to zero would suggest a fixed-effects model (i.e., no variation in the effect size across papers), while values above 2.5 would signal a problem with the data or the model. We visually assessed heterogeneity with L’Abbé plots.28 The L’Abbé plot shows the event rate for those with a given covariate against those without the covariate for each paper that provided exact count data, where each circle represents a paper. Circle sizes are proportional to the square root of the sample size. Points above the line represent a higher event rate for those with the covariate than those without, and an OR or LR greater than one. Points clustered near the line suggest little or no covariate effect.
Initial Medline and manual screening identified 2,773 potentially eligible studies, and we identified seven additional studies through citation reviews. Seventeen articles met our inclusion criteria and were included in the systematic review (Appendix B [eTable 1]; Appendix C [eFigure 1]).4,15,16,18,20,21,29-39 Definitions for serious clinical outcomes, abnormal ECG findings, and binary thresholds for continuous variables (such as age, vital signs, and biomarker tests) varied across papers; these are summarized in Appendix B [eTables 1–3]. Risk for bias is described in Appendix B [eTable 4]. We identified 32 predictors for which there were sufficient data for analysis. Visual representation of effect size heterogeneity is presented in Appendix C [eFigure 2]. There were 12 predictors with a significant LR+, and 12 predictors with a significant LR−. In this section, we highlight findings of LR+>2.0 and LR−<0.5 as the strongest predictors of short-term outcomes identified in this meta-analysis.
Pretest Probability of Serious Outcomes
Rates of serious outcomes after an ED evaluation syncope ranged from 1.2–36.2% (interquartile range 6.7–16.9%). There was heterogeneity in the definition of serious outcomes across studies (Appendix B [eTable 1]) and whether patients with serious outcomes identified during the index ED visit were excluded from analysis (Appendix B [eTable 4]).
Patient Characteristics and Co-morbidities (Table 1)
|Variable||Papers||Patients||LR+ (95% CI)||LR− (95%CI)||Sensitivity (95%CI)||Specificity (95%CI)||Posterior SD|
|CHF||8||40279||2.65 (1.67, 3.94)||0.73 (0.54, 0.89)||0.23 (0.17, 0.29)||0.85 (0.77, 0.92)||0.740|
|Arrhythmia||5||39773||2.30 (1.29, 3.60)||0.72 (0.48, 0.94)||0.14 (0.10, 0.20)||0.81 (0.70, 0.90)||0.777|
|Heart disease||8||44074||1.82 (1.26, 2.51)||0.73 (0.54, 0.91)||0.14 (0.10, 0.19)||0.75 (0.64, 0.84)||0.716|
|Older age||8||28624||1.80 (1.39, 2.36)||0.44 (0.25, 0.68)||0.11 (0.08, 0.15)||0.58 (0.43, 0.72)||0.778|
|Pacemaker||3||38207||1.58 (0.62, 2.96)||0.89 (0.62, 1.09)||0.11 (0.06, 0.17)||0.83 (0.70, 0.92)||0.840|
|Diabetes||5||39773||1.38 (0.80, 2.12)||0.88 (0.64, 1.07)||0.09 (0.06, 0.13)||0.75 (0.62, 0.86)||0.736|
|Male gender||9||44481||1.35 (1.07, 1.70)||0.72 (0.51, 0.94)||0.09 (0.07, 0.12)||0.55 (0.40, 0.69)||0.666|
|Cerebrovascular||2||36000||1.34 (0.55, 2.43)||0.87 (0.47, 1.20)||0.06 (0.03, 0.10)||0.70 (0.52, 0.86)||0.843|
|Arrhythmic medication||2||977||1.30 (0.37, 2.84)||0.94 (0.56, 1.25)||0.16 (0.08, 0.27)||0.77 (0.60, 0.90)||0.940|
|Hypertension||4||39103||1.28 (0.83, 1.76)||0.77 (0.44, 1.15)||0.08 (0.05, 0.12)||0.54 (0.37, 0.69)||0.801|
|Seizure||2||36014||1.01 (0.27, 2.29)||1.00 (0.68, 1.25)||0.05 (0.03, 0.09)||0.79 (0.62, 0.91)||0.897|
|Hispanic||2||3061||0.97 (0.35, 1.94)||1.02 (0.64, 1.36)||0.08 (0.04, 0.13)||0.70 (0.51, 0.85)||0.853|
|Stroke||2||3268||0.81 (0.24, 1.78)||1.08 (0.73, 1.40)||0.06 (0.03, 0.10)||0.73 (0.54, 0.87)||0.886|
|Previous syncope||4||3908||0.73 (0.30, 1.39)||1.09 (0.88, 1.30)||0.08 (0.05, 0.12)||0.76 (0.60, 0.88)||0.807|
|Nonwhite Race||3||38391||0.67 (0.30, 1.18)||1.24 (0.87, 1.64)||0.04 (0.02, 0.06)||0.59 (0.41, 0.76)||0.791|
CHF, congestive heart failure; CI, confidence interval, SD, standard deviation; Posterior SD is the SD of random effects.
Pre-existing co-morbidities predictive of short-term outcomes included a history of congestive heart failure (LR+ 2.65, 95%CI [1.69, 3.91]) and prior arrhythmias (LR+ 2.32, 95%CI [1.31.–3.62]). Younger age was variably defined (e.g., less than 58–75; see Appendix B [eTable 2]) and was associated with lower risk of outcomes (LR− 0.44, 95%CI [0.25, 0.68]). Other patient characteristics, including gender, race and prior history of syncope were weakly predictive of short-term outcomes.
Symptoms (Table 2)
|Variable||Papers||Patients||LR+ (95% CI)||LR− (95%CI)||Sensitivity (95%CI)||Specificity (95%CI)||Posterior SD|
|Dyspnea||6||4772||2.29 (1.31, 3.65)||0.78 (0.57, 0.94)||0.25 (0.18, 0.33)||0.85 (0.75, 0.92)||0.765|
|Trauma||2||3254||1.68 (0.70, 3.00)||0.76 (0.38, 1.12)||0.11 (0.06, 0.18)||0.72 (0.54, 0.87)||0.850|
|Palpitations||4||1713||1.51 (0.59, 3.02)||0.91 (0.66, 1.09)||0.18 (0.11, 0.27)||0.84 (0.72, 0.92)||0.900|
|Supine||2||915||1.42 (0.55, 2.66)||0.82 (0.36, 1.29)||0.17 (0.09, 0.28)||0.66 (0.45, 0.83)||0.987|
|Effort||3||1420||1.36 (0.51, 2.64)||0.91 (0.59, 1.17)||0.14 (0.08, 0.22)||0.78 (0.62, 0.89)||0.887|
|Chest Pain||3||3561||1.18 (0.45, 2.30)||0.96 (0.66, 1.20)||0.12 (0.07, 0.19)||0.77 (0.62, 0.89)||0.886|
|No Prodromes||7||8980||1.10 (0.72, 1.51)||0.93 (0.65, 1.21)||0.09 (0.06, 0.13)||0.58 (0.43, 0.72)||0.810|
CI, confidence interval, SD, standard deviation; Posterior SD is the SD of random effects.
A complaint of dyspnea was predictive (LR+ 2.29, 95%CI [1.31, 3.65]). Other symptoms, including traumatic injury after syncope, palpitations, position, effort, chest pain, and absence of prodromes, had non-significant LR+ and LR−.
Physical Exam (Table 3)
|Variable||Papers||Patients||LR+ (95% CI)||LR− (95%CI)||Sensitivity (95%CI)||Specificity (95%CI)||Posterior SD|
|Hypotension||5||4540||2.62 (1.12, 4.90)||0.82 (0.59, 0.98)||0.25 (0.17, 0.35)||0.89 (0.82, 0.95)||0.936|
|Respiratory Rate||2||4714||2.26 (0.83, 4.44)||0.74 (0.38, 1.04)||0.15 (0.08, 0.24)||0.81 (0.66, 0.92)||0.869|
|Murmur||4||3792||1.86 (0.84, 3.47)||0.83 (0.55, 1.04)||0.20 (0.13, 0.30)||0.82 (0.70, 0.91)||0.873|
|Oxygen||3||1660||1.49 (0.77, 2.46)||0.79 (0.45, 1.12)||0.15 (0.09, 0.22)||0.68 (0.49, 0.84)||0.807|
CI, confidence interval; SD, standard deviation; Posterior SD is the SD of random effects.
Hypotension was variably defined (e.g., systolic blood pressure less than 80–90; see Appendix B [eTable 2]) but identified as a strong predictor (LR+2.62, 95%CI [1.12, 4.90]). Presence of cardiac murmur, rapid respiratory rate, and low oxygen-saturation level were not predictive.
Tests (Table 4)
|Variable||Papers||Patients||LR+ (95% CI)||LR− (95%CI)||Sensitivity (95%CI)||Specificity (95%CI)||Posterior SD|
|Urea||2||4535||2.86 (1.15, 5.42)||0.57 (0.23, 0.96)||0.20 (0.10, 0.31)||0.79 (0.63, 0.91)||0.946|
|Troponin||3||6952||2.49 (1.36, 4.10)||0.54 (0.26, 0.87)||0.19 (0.11, 0.27)||0.75 (0.59, 0.87)||0.857|
|Creatinine||2||4535||2.43 (0.96, 4.59)||0.65 (0.29, 1.01)||0.16 (0.08, 0.26)||0.78 (0.62, 0.90)||0.913|
|BNP||2||663||2.19 (1.14, 4.00)||0.45 (0.15, 0.91)||0.35 (0.20, 0.50)||0.65 (0.41, 0.84)||0.877|
|Hematocrit||8||9354||2.14 (1.21, 3.43)||0.85 (0.67, 0.97)||0.19 (0.14, 0.26)||0.88 (0.80, 0.93)||0.799|
|ECG||7||5114||1.79 (1.14, 2.63)||0.60 (0.33, 0.92)||0.20 (0.13, 0.28)||0.65 (0.51, 0.78)||1.083|
BNP, b-type natriuretic peptide; ECG, electrocardiogram; CI, confidence interval, SD, standard deviation; Posterior SD is the SD of random effects.
We identified multiple biomarkers predictive of short-term risk; thresholds used to dichotomize test values are presented in Appendix B [eTable 2]. Predictors of increased risk include elevated blood urea nitrogen (LR+ 2.86, 95%CI [1.15, 5.42]), troponin (LR+ 2.49, 95%CI [1.36, 4.10]), B-type natriuretic peptide (LR+ 2.19, 95%CI [1.14, 4.00]), and hematocrit (LR+ 2.14, 95%CI [1.21, 3.43]). A low B-type natriuretic peptide level was associated with reduced risk (LR− 0.45, 95%CI [0.15, 0.91]). The presence or absence of ECG abnormalities (Appendix B [eTable 3]) was modestly predictive (LR+ 1.79, 95%CI [1.14, 2.63]; LR− 0.6, 95%CI [0.33, 0.92]).
In this systematic review of ED patients presenting with syncope, we identified 17 articles that contained short-term prognostic information about patient characteristics, symptoms, physical exam findings, and objective testing. We identified several predictors of short-term serious outcomes. The strongest predictors of increased risk were elevated blood urea nitrogen level, a prior history of congestive heart failure, hypotension in the ED, a history of arrhythmia, and an abnormal troponin value. Younger age was the strongest predictor of lower risk. Prognostic factors had modest specificity and poor sensitivity, and likelihood ratios suggest that no single factor is sufficient to classify patients at high risk for short-term serious events. In the absence of effective risk scores, our findings can be used to supplement clinical judgment to guide management of ED patients without an obvious cause of syncope. Our results can also inform the development of more accurate and robust risk-prediction tools. For example, these data can inform “prior” estimates of association in novel risk score development using a Bayesian framework.
Higher risk patient characteristics are mostly concordant with the 2017 AHA Guidelines, including older age, male gender, and prior cardiac co-morbidities such as heart failure, arrhythmias, and coronary artery disease. We did not find an increased event rate associated with previous episodes of syncope. A prior population-based study did suggest increased 30-day mortality associated with prior syncope40; however, this report was excluded from our results due to inability to extract data amenable to meta-analysis.
Despite conventional clinical teaching,41 we found that most symptoms were poor predictors. Dyspnea was the only factor that had significant positive and negative likelihood ratios. The absence or presence of other symptoms, including palpitations, syncope while supine, chest pain, and lack of prodromes, did not significantly alter risk. It is possible that patients had poor or inaccurate recall of the circumstances associated with syncope, which would limit the discriminating value of symptoms.
Low blood pressure was the only physical exam factor predictive of risk. Although the auscultation of cardiac murmur is suggestive of structural heart disease,42 this finding was not predictive of short-term serious events. A potential explanation may be poor inter-rater reliability of cardiac auscultation.16,43
Initial ECG testing is a Class I recommendation in the 2017 AHA Guidelines.1 We found that the ECG has modest discriminating value, consistent with a prior study that reported a <3% diagnostic yield.44 Troponin and B-type natriuretic peptide may be objective indicators of cardiac dysfunction, and they appear to be predictive of serious short-term risk after syncope. Low hematocrit may be associated with anemia and other chronic diseases that confer risk after syncope. Finally, our meta-analysis suggests that blood urea nitrogen is associated with short-term serious events. The biological mechanism for this observation is not clear, although patients with renal dysfunction may be at higher risks for arrhythmias, bleeding, or death.45,46
Our study builds on a prior meta-analysis of predictors of adverse outcomes in syncope.47 D’Ascenzo identified eight variables associated with serious outcomes (OR>5) including palpitations, exertional syncope, heart disease, bleeding, supine syncope, absence of prodromes, increasing age, and trauma related to syncope. Our meta-analysis builds on this prior effort by doing the following: focusing on studies of short-term outcomes that would be relevant to emergency physicians; including recently published studies; excluding improbable results (e.g., D’Ascenzo reported an OR=65 for palpitations); analyzing a broad set of potential predictors; and reporting test characteristics that are relevant to clinicians, such as likelihood ratio, sensitivity, and specificity. Discrepancies in the reported findings between the two studies are likely due to these methodological differences.
First, risk prediction studies of ED patient with syncope are characterized by heterogeneity in outcomes.24 The majority of investigations used composite outcomes of a broad range of serious clinical conditions, which reflects clinical reality facing the ED provider – syncope can be related to a wide range of dangerous conditions that may not be apparent on initial evaluation. Although the precise outcomes definitions varied by study, there was substantial overlap in included conditions. For example, all studies included death and arrhythmia, and most studies included such conditions including acute myocardial infarction, pulmonary embolus, and significant hemorrhage.
Second, there was heterogeneity in predictor measurement. For example, binary cutoffs for continuous variables and the definition for “abnormal” ECG results differed across studies. However, we found strong consistency in direction of effect for ECG abnormalities and most continuous variables (e.g., blood pressure, serum tests). Measurement heterogeneity would likely result in a conservative bias toward a null finding.
Third, most of the studies did not exclude patients with outcomes identified during the ED evaluation. The majority of serious clinical conditions are identified during the initial ED evaluation for syncope,4 and failure to exclude such patients biases risk prediction toward “obvious” events. There is also the potential for incorporation bias, where a test result is used to define the outcome (e.g., an ECG that documents a clinically significant arrhythmia). We attempted to mitigate incorporation bias by excluding studies with implausibly high associations.
Finally, we performed a meta-analysis on univariate predictors of outcomes, and it is likely that these factors are not independent of each other. This limitation underscores the importance of developing risk scores that account for the correlations among potential predictors.
We identified specific risk predictors of short-term clinical events after an ED evaluation for syncope. Individual risk factors, symptoms, exam findings, and test results in isolation are modestly predictive of risk. These findings should be used to supplement clinical judgment and inform the development of novel risk scores.
Section Editor: Michael Kurz, MD, MS
Full text available through open access at http://escholarship.org/uc/uciem_westjem
Address for Correspondence: Benjamin C. Sun, MD, MPP, Oregon Heath & Science University, Center for Policy and Research in Emergency Medicine, Department of Emergency Medicine, 3181 SW Sam Jackson Park Road, CR114, Portland, OR 97239-3098. Email: firstname.lastname@example.org. 5 / 2018; 19:517 – 523
Submission history: Revision received November 20, 2017; Accepted February 20, 2018
Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. No author has professional or financial relationships with any companies that are relevant to this study. There are no conflicts of interest or sources of funding to declare.
1. Shen WK, Sheldon RS, Benditt DG, et al. 2017 ACC/AHA/HRS Guideline for the Evaluation and Management of Patients With Syncope: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines, and the Heart Rhythm Society. Circulation. 2017;136(5):e60-e122.
2. Probst MA, Kanzaria HK, Gbedemah M, et al. National trends in resource utilization associated with ED visits for syncope. Am J Emerg Med. 2015;38(8):998-1001.
3. Sun BC, Emond JA, Camargo CA. Direct medical costs of syncope-related hospitalizations in the United States. Am J Cardiol. 2005;95(5):668-71.
4. Sun BC, Derose SF, Liang LJ, et al. Predictors of 30-day serious events in older patients with syncope. Ann Emerg Med. 2009;54:769-78.e1–5.
5. Birnbaum A, Esses D, Bijur P, et al. Failure to validate the San Francisco syncope rule in an independent emergency department population. Ann Emerg Med. 2008;52(2):151-9.
6. Sun BC, Emond JA, Camargo CA. Characteristics and admission patterns of patients presenting with syncope to U.S. Emergency departments, 1992–2000. Acad Emerg Med. 2004;11(10):1029-34.
7. Huff JS, Decker WW, Quinn JV, et al. Clinical policy: critical issues in the evaluation and management of adult patients presenting to the emergency department with syncope. Ann Emerg Med. 2007;49(4):431-44.
8. Linzer M, Yang EH, Estes NA, et al. Diagnosing syncope. Part 1: Value of history, physical examination, and electrocardiography. Clinical Efficacy Assessment Project of the American College of Physicians. Ann Intern Med. 1997;126(12):989-96.
9. Mozes B, Confino-Cohen R, Halkin H. Cost-effectiveness of in-hospital evaluation of patients with syncope. Isr J Med Sci. 1988;24:302-6.
10. Crane SD. Risk stratification of patients with syncope in an accident and emergency department. Emerg Med J. 2002;19:23-7.
11. Sun BC, Costantino G, Barbic F, et al. Priorities for emergency department syncope research. Ann Emerg Med. 2014;64(6):649-55.e2.
12. Sun BC, Thiruganasambandamoorthy V, Cruz JD. Standardized reporting guidelines for emergency department syncope risk-stratification research. Acad Emerg Med. 2012;19(6):694-702.
13. Thiruganasambandamoorthy V, Taljaard M, Stiell IG, et al. Emergency department management of syncope: need for standardization and improved risk stratification. Intern Emerg Med. 2015;10(5):619-27.
14. Martin TP, Hanusa BH, Kapoor WN. Risk stratification of patients with syncope. Ann Emerg Med. 1997;29(4):459-66.
15. Sarasin FP, Hanusa BH, Perneger T, et al. A risk score to predict arrhythmias in patients with unexplained syncope. Acad Emerg Med. 2003;10(12):1312-7.
16. Quinn JV, Stiell IG, McDermott DA, et al. Derivation of the San Francisco Syncope Rule to predict patients with short-term serious outcomes. Ann Emerg Med. 2004;43(2):224-32.
17. Colivicchi F, Ammirati F, Melina D, et al. Development and prospective validation of a risk stratification system for patients with syncope in the emergency department: the OESIL risk score. Eur Heart J. 2003;24:811-9.
18. Grossman SA, Fischer C, Lipsitz LA, et al. Predicting adverse outcomes in syncope. J Emerg Med. 2007;33(3):233-9.
19. Del Rosso A, Ungar A, Maggi R, et al. Clinical predictors of cardiac syncope at initial evaluation in patients referred urgently to general hospital: the EGSYS score. Heart. 2008;94(12):1620-6.
20. Costantino G, Perego F, Dipaola F, et al. Short- and long-term prognosis of syncope, risk factors, and role of hospital admission: results from the STePS (Short-Term Prognosis of Syncope) study. J Am Coll Cardiol. 2008;51(3):276-83.
21. Reed MJ, Newby DE, Coull AJ, et al. The ROSE (risk stratification of syncope in the emergency department) study. J Am Coll Cardiol. 2010;55(8):713-21.
22. Costantino G, Casazza G, Reed M, et al. Syncope Risk Stratification Tools Vs Clinical Judgment: An Individual Patient Data Meta-Analysis. Am J Med. 2014;127(11):1126.e13-25.
23. Serrano LA, Hess EP, Bellolio MF, et al. Accuracy and quality of clinical decision rules for syncope in the emergency department: a systematic review and meta-analysis. Ann Emerg Med. 2010;56(4):362-73.e1.
24. Sun B, Costantino G. Syncope risk stratification in the ED: directions for future research. Acad Emerg Med. 2013;20(5):503-6.
25. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ. 2009;339:b2700.
26. Whiting P, Rutjes AW, Reitsma JB, et al. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol. 2003;3:25.
27. The R Project for Statistical Computing. 2017. Available at: https://www.r-project.org/. Accessed August 18, 2017.
28. Song F. Exploring heterogeneity in meta-analysis: is the L’Abbe plot useful?. J Clin Epidemiol. 1999;52(8):725-30.
29. Derose SF, Gabayan GZ, Chiu VY, et al. Patterns and preexisting risk factors of 30-day mortality after a primary discharge diagnosis of syncope or near syncope. Acad Emerg Med. 2012;19(5):488-96.
30. Gabayan GZ, Derose SF, Asch SM, et al. Predictors of short-term (seven-day) cardiac outcomes after emergency department visit for syncope. Am J Cardiol. 2010;105:82-6.
31. Isbitan A, Hawatmeh A, Elnahar Y, et al. Utility of brain natriuretic peptide assay as a predictor of short term outcomes in patients presenting with syncope to the emergency department. Cardiovasc Diagn Ther. 2016;6(3):234-40.
32. Kayayurt K, Akoglu H, Limon O, et al. Comparison of existing syncope rules and newly proposed Anatolian Syncope Rule to predict short-term serious outcomes after syncope in the Turkish population. Int J Emerg Med. 2012;5:17.
33. Numeroso F, Mossini G, Giovanelli M, et al. Short-term prognosis and current management of syncopal patients at intermediate risk: results from the IRiS (Intermediate-Risk Syncope) Study. Acad Emerg Med. 2016;23(8):941-8.
34. Numeroso F, Mossini G, Lippi G, et al. Evaluation of the current prognostic role of heart diseases in the history of patients with syncope. Europace. 2014;16:1379-83.
35. Quinn J, McDermott D. Electrocardiogram findings in emergency department patients with syncope. Acad Emerg Med. 2011;18(7):714-8.
36. Reed MJ, Mills NL, Weir CJ. Sensitive troponin assay predicts outcome in syncope. Emerg Med J. 2012;29(12):1001-3.
37. Sun BC, Hoffman JR, Mangione CM, et al. Older age predicts short-term, serious events after syncope. J Am Geriatr Soc. 2007;55:907-12.
38. Thiruganasambandamoorthy V, Kwong K, Wells GA, et al. Development of the Canadian Syncope Risk Score to predict serious adverse events after emergency department assessment of syncope. CMAJ. 2016;188(12):E289-E298.
39. Thiruganasambandamoorthy V, Stiell IG, Sivilotti ML, et al. Risk stratification of adult emergency department syncope patients to predict short-term serious outcomes after discharge (RiSEDS) study. BMC Emerg Med. 2014;14:8.
40. Ruwald MH, Nume AK, Lamberts M, et al. Incidence and influence of hospitalization for recurrent syncope and its effect on short- and long-term all-cause and cardiovascular mortality. Am J Cardiol. 2014;113:1744-50.
41. Kapoor WN. Syncope. N Engl J Med. 2000;343(25):1856-62.
42. Sarasin FP, Junod AF, Carballo D, et al. Role of echocardiography in the evaluation of syncope: a prospective study. Heart. 2002;88(4):363-7.
43. Nishijima DK, Laurie AL, Weiss RE, et al. Reliability of clinical assessments in older adults with syncope or near syncope. Acad Emerg Med. 2016;23(9):1014-21.
44. Mendu ML, McAvay G, Lampert R, et al. Yield of diagnostic tests in evaluating syncopal episodes in older patients. Arch Intern Med. 2009;169(14):1299-305.
45. Thiruganasambandamoorthy V, Wells GA, Hess EP, et al. Derivation of a risk scale and quantification of risk factors for serious adverse events in adult emergency department syncope patients. CJEM. 2014;16(2):120-30.
46. Marti-Almor J, Cladellas M, Bazan V, et al. Long-term mortality predictors in patients with chronic bifascicular block. Europace. 2009;11:1201-7.
47. D’Ascenzo F, Biondi-Zoccai G, Reed MJ, et al. Incidence, etiology and predictors of adverse outcomes in 43,315 patients presenting to the emergency department with syncope: an international meta-analysis. Int J Cardiol. 2013;167(1):57-62.