Illustration: A translucent figure tipping backward in a faint, the brain gone dark grey while the heart alone glows copper.
← EMA Cited
EMA Cited · Cardiovascular

Syncope

Updated Oct 29, 2026Adults · ED evaluation27 sources · 23 checked in PubMed2 figuresACC/AHA/HRS 2017 · ESC 2018 · SAEM GRACE reviews

Bottom line

  • Every patient gets a history, an exam and a 12-lead ECG. Broad routine lab testing is not useful, and cardiac imaging is only for a suspected cardiac cause.1,2
  • The ECG carries the most weight. Acute ischemia, Mobitz II or complete heart block, sinus rates or atrial fibrillation under 40, ventricular tachycardia, a type 1 Brugada pattern, or a QTc over 460 ms are major high-risk features.3
  • Risk scores help but do not replace judgment. The Canadian Syncope Risk Score validated well in Canada (AUC 0.91), but pooled evidence for every rule is low quality.4,5
  • Don't test everyone for PE. It was found in 17.3% of a systematically tested inpatient cohort, but in only 0.6–1.4% of ED patients in two large prospective cohorts.6,7,8
  • Low-risk patients can go home. In CSRS low-risk patients, half of serious arrhythmias showed up within 2 hours of arrival. For medium and high risk it took 6 hours.9

Is it syncope?

Transient loss of consciousness (TLOC) is a brief episode of unconsciousness followed by full recovery. Syncope is the subset caused by a fall in blood pressure, with global cerebral hypoperfusion and loss of postural tone. Other causes of TLOC include seizures, psychogenic events and rarer conditions. Syncope accounts for about 1–5% of ED visits.3

About 10% of patients have a serious cause that may not be obvious at the first ED assessment.5 Presyncope deserves the same caution: across 5 studies of 2,741 ED patients with presyncope, 30-day serious outcomes ranged from 4.4% to 26.8%, most often arrhythmia.10

First steps

  1. Look for trauma and shock first. Either one moves the patient off the syncope pathway and onto that problem's own care.3
  2. Check vital signs and get a 12-lead ECG. An experienced clinician should read it within 10 minutes of arrival.3
  3. Take a careful history and examine the patient. The guideline recommends this, with the ECG, for every patient. Do an active standing test for orthostatic hypotension where suitable.1,2,3
  4. Decide whether it was syncope. Seizure, intoxication and psychogenic events can look alike. Keep checking the diagnosis as results come back.3
  5. Treat a cause you find. Syncope may be a symptom of PE, aortic dissection, bleeding or sepsis, or have a clear cardiac cause such as aortic stenosis or complete heart block. Manage that condition, and hospitalize when it is serious.3,1,2
  6. If the cause is still unclear, stratify risk with the ESC low, intermediate and high categories or a validated score, alongside your own judgment.11,3,4
  7. Dispose by risk. Low-risk patients go home with follow-up. Intermediate and high-risk patients need monitoring and hospital-based care.3
FIG 1 · Decision flowCited
  1. Trauma or shock?
    Yes
    Treat the trauma or shock first, on its own pathway.3
  2. No
    Every patient: vital signs, history, exam and a 12-lead ECG read within 10 minutes. Glucose and hemoglobin are a common minimum.1,2,3 Treat presyncope with the same caution.10
  3. Was it really syncope?
    No
    Consider seizure, intoxication or a psychogenic event, and keep re-checking the diagnosis as results return.3
  4. Yes
    Test for PE only when the picture fits. In large ED cohorts, PE caused 0.6–1.4% of syncope.7,8
  5. Serious cause found (PE, dissection, bleeding, sepsis, aortic stenosis, heart block)?
    Yes
    Treat the cause and hospitalize when it is serious.3,1,2
  6. No cause found
    Any major high-risk ECG feature?
    Yes
    Monitor and admit. High-risk patients need hospital-based care.3
  7. No
    Stratify risk with the ESC categories or the CSRS, alongside your judgment.3,4 Minor ECG features count as high risk if the history suggests an arrhythmia.3
  8. By risk category
    Low
    Discharge with outpatient follow-up.3 Half of arrhythmias in CSRS low-risk patients appeared within 2 hours of arrival.9
    Intermediate or high
    Monitor in the ED, an observation unit or the hospital. Don't discharge directly.3 Half of arrhythmias appeared within 6 hours.9
    If discharged later, consider 15 days of outpatient rhythm monitoring for selected medium-risk and all high-risk patients.9
  9. Vasovagal: educate the patient. Orthostatic from dehydration: give fluids and review blood-pressure drugs. Cardiac cause: specialist-directed care.1,2
FIG 1ED syncope, from arrival to disposition. Drawn for this page from the EUSEM consensus pathway, the ACC/AHA/HRS guideline summary and the cited cohorts. Each box carries its own citations.3,1,2,9

ECG

ECG findingRisk
Major: high risk on its own
Signs of acute ischemiaMajor
Mobitz II second-degree or third-degree AV blockMajor
Slow atrial fibrillation (rate < 40)Major
Persistent sinus bradycardia < 40, repeated sinoatrial block, or pauses > 3 s while awake and not athletically trainedMajor
Bundle branch block, other conduction delay, ventricular hypertrophy, or signs of ischemic heart disease or cardiomyopathyMajor
Sustained or nonsustained ventricular tachycardiaMajor
Pacemaker or ICD malfunctionMajor
Type 1 Brugada pattern (ST elevation in V1–V3)Major
QTc > 460 ms on repeated 12-leadsMajor
Minor: high risk only if the history suggests an arrhythmia
Mobitz I, or first-degree AV block with a markedly long PRMinor
Asymptomatic bradycardia 40–50Minor
Paroxysmal SVT or atrial fibrillationMinor
Pre-excited QRSMinor
Short QTc (< 340 ms)Minor
Atypical Brugada patternMinor
Negative T waves in the right precordial leads or epsilon waves (suggesting ARVC)Minor

From the EUSEM consensus, which follows the 2018 ESC guideline. Only a normal ECG counts as low risk.3,11 See the LITFL Brugada ECG library for examples.12

Risk scores

Use a score only after the ED workup has found no serious cause. A systematic review of the San Francisco Syncope Rule found it performed best in that group, and that missed arrhythmias caused most of its false negatives.13 The CSRS was validated for serious outcomes that were not evident during the ED evaluation.4

San Francisco Syncope RuleCanadian Syncope Risk Score
ItemsHistory of heart failure; hematocrit < 30%; abnormal ECG (new changes or non-sinus rhythm); shortness of breath; triage systolic BP < 90 mm Hg.14Predisposition to vasovagal syncope; heart disease; any ED systolic BP < 90 or > 180 mm Hg; troponin above the 99th percentile; QRS axis < −30° or > 100°; QRS > 130 ms; QTc > 480 ms; ED diagnosis of cardiac syncope; ED diagnosis of vasovagal syncope.15
Derivation684 visits, 7-day outcomes. Sensitivity 96%, specificity 62%.164,030 patients, 3.6% with a 30-day serious event. C statistic 0.88.15
Validation791 visits at one hospital: sensitivity 98%, specificity 56%.14 Pooled across 12 studies: sensitivity 0.87, specificity 0.52, with a wide prediction interval for sensitivity (0.55–0.98).133,819 patients at 9 Canadian EDs: AUC 0.91. Serious outcomes in 0.3% of very-low-risk and 51.3% of very-high-risk patients. At a threshold of −1, sensitivity 97.8%, specificity 44.3%.4
Outside its home country—2,283 patients aged ≥ 40 in 8 countries: AUC 0.85.17 In 345 Italian patients, 6.7% of CSRS low-risk patients had an event, vs 2% of those judged low risk by clinicians.18
Pooled (SAEM GRACE)Sensitivity 0.86, specificity 0.56.5Sensitivity 0.89, specificity 0.48, from 5 validation studies.5

Items only; this page does not score patients. The GRACE review rated the evidence for all these rules low quality.5

Sources disagree

The CSRS authors recommend using the score to guide ED decisions.4 Others are less sure. An individual-patient meta-analysis found that the OESIL, San Francisco and EGSYS tools did no better than clinical judgment.19 In the international CSRS validation, the clinician's own classification of the syncope explained much of the score's performance, and its authors called its clinical value uncertain.17 The GRACE review concluded that clinical judgment remains a valid way to stratify risk.5 REBEL EM notes that the Italian cohort was small, older, from one country, and had many patients lost to follow-up.20

Testing

  • Blood tests: glucose and hemoglobin are a common minimum. Add tests such as troponin or d-dimer when a specific diagnosis is suspected.3 Broad routine testing is not useful.1,2
  • Troponin: an elevated level is a CSRS item.15 In nearly 5,000 patients, the first troponin was enough for risk stratification. Serial testing did not improve prognosis and lengthened the ED stay (median 5.6 vs 3.8 hours). Repeat it only for ongoing symptoms or ischemic ECG changes.21
  • BNP, echo and outpatient monitoring: a GRACE review of 41 studies found their accuracy and yield too variable to recommend any of them routinely on their own. Echo and monitoring found more in patients with cardiac risk factors.22 The guideline advises cardiac imaging only when a cardiac cause is suspected.1,2
  • Head CT: in ED studies, over half of patients got one, and 3.8% had a finding. Risk factors for a finding were abnormal neurologic signs, age ≥ 65, trauma, warfarin use, and a history of seizure or stroke.23 Carotid imaging is not useful without focal neurologic findings.1,2

Pulmonary embolism

  • PESIT: 560 patients admitted for a first syncope (mean age 76) were all worked up for PE. It was found in 17.3%, including 12.7% of those with another explanation for the syncope.6
  • Administrative data on 1.67 million ED syncope patients in 4 countries: PE in 0.06–0.55% of all patients and 0.15–2.10% of those hospitalized.24
  • North American prospective cohort (9,091 patients, 17 EDs): PE in 0.6%. Four deaths (0.04%) were related to PE.7
  • BASEL IX (1,397 patients): PE in 1.4% at presentation, and 2.3% of those hospitalized.8
Sources disagree

PESIT found PE in nearly one in six admitted patients, in an older inpatient group tested systematically.6 ED cohorts found it in 0.6–1.4%, and their authors warn against indiscriminate PE testing.7,8 PE should be considered in every patient, but not every patient needs testing for it.24

Disposition

  • A serious condition that explains the syncope: hospitalize.1,2
  • ESC categories: discharge low-risk patients to outpatient care. Patients neither low nor high risk need more workup and monitoring, not direct discharge. High-risk patients are monitored and admitted.3
  • CSRS: very-low and low-risk patients can generally be discharged; none died or had a ventricular arrhythmia in the validation cohort. Brief hospitalization can be considered for high risk.4
  • Does admission help? A GRACE review of 18 studies (510,545 patients) found the evidence too uncertain to recommend for or against hospitalization. Discharging low-risk patients carried a low risk of short-term events.25
  • By ED diagnosis: serious outcomes rose from presumed vasovagal to orthostatic, unknown and cardiac syncope. No patient with presumed vasovagal syncope died.26
  • Vasovagal syncope is the most common cause; educate the patient. Give fluids for orthostatic hypotension from dehydration, and consider stopping blood-pressure drugs. Cardiac syncope needs specialist-directed care.1,2

How long to monitor

In 5,581 prospectively enrolled patients, the 30-day risk of a serious arrhythmic outcome was 0.4% for CSRS low risk, 8.7% for medium and 25.3% for high. Half of these outcomes were found within 2 hours of ED arrival in low-risk patients, and within 6 hours in medium and high-risk patients. After those cut points, the remaining risk was 0.2%, 5.0% and 18.1%. In medium and high-risk patients, 91.7% of arrhythmic outcomes, including every ventricular arrhythmia, were found within 15 days. The authors suggest 15 days of outpatient rhythm monitoring for selected medium-risk and all high-risk patients after discharge.9

FIG 2When arrhythmias show up, by CSRS risk. Drawn for this page from a 5,581-patient prospective cohort. Each bar runs to the point by which half of 30-day serious arrhythmic outcomes had been found, measured from ED arrival. The lighter bars continue past the chart: in medium and high-risk patients, 91.7% were found within 15 days.9

Pitfalls

  • Applying a risk score before the ED workup is done. The rules are meant for patients with no cause found.13
  • Missing syncope as the cause of a fall or injury, especially in older patients with dementia.3
  • Treating presyncope as harmless. Its short-term risk may match syncope.10
  • Discharging a patient with a minor ECG finding, such as pre-excitation or a short QT, when the history suggests an arrhythmia.3
  • Ordering serial troponins on everyone.21
  • Reflex head CT. The yield is low without neurologic signs, age ≥ 65, trauma, warfarin use, or a history of seizure or stroke.23
  • Testing everyone for PE, or forgetting PE when the picture fits.7,24

How strong is the evidence?

No randomized trial has compared a syncope risk tool with clinical judgment, and the evidence for every validated rule is low quality.5 Whether admission helps is uncertain.25 The SAEM GRACE reviews were done to build an ED syncope guideline.5 The 2017 ACC/AHA/HRS and 2018 ESC guidelines were not openly accessible, so their points come from a published summary and from the EUSEM consensus, which builds on the ESC guideline.1,2,11,3

On-shift teaching

Questions to ask a learner at the bedside. Tap a question to see the answer. Every answer uses facts cited above.

Q1Name five ECG findings that make a syncope patient high risk on their own.
Any of: acute ischemia, Mobitz II or complete heart block, atrial fibrillation or sinus rates under 40, ventricular tachycardia, bundle branch block, pacemaker malfunction, a type 1 Brugada pattern, or a QTc over 460 ms on repeated ECGs. Only a normal ECG counts as low risk.3
Q2What are the five San Francisco Syncope Rule items, and how sensitive is it?
History of heart failure, hematocrit < 30%, abnormal ECG, shortness of breath, triage systolic BP < 90. It was 98% sensitive in its validation, but pooled sensitivity across 12 studies was 0.87.14,13
Q3Does the Canadian Syncope Risk Score beat clinical judgment?
Not clearly. In 8 countries, the clinician's own classification of the syncope explained much of its performance.17 In an Italian cohort, CSRS low-risk patients had more events than clinician-judged low-risk patients (6.7% vs 2%).18 The GRACE review rated all rules' evidence low quality.5
Q4PESIT found PE in one in six syncope patients. Why not test everyone?
PESIT tested older, admitted patients systematically.6 In unselected ED patients, prospective cohorts found PE in 0.6–1.4%, and their authors warn against indiscriminate testing.7,8
Q5A CSRS medium-risk patient has been on the monitor for 2 hours. Is that long enough?
Probably not. Half of arrhythmic outcomes in medium and high-risk patients were found within 6 hours, and 91.7% within 15 days. Low-risk patients reached that halfway point by 2 hours.9
Q6When is a head CT worth ordering after syncope?
When there are abnormal neurologic signs, age ≥ 65, trauma, warfarin use, or a history of seizure or stroke. Over half of ED syncope patients got one, and only 3.8% had a finding.23

References

Evidence base

What this page is built on

27 sources, cited 117 times. Dots are sources, filled by how far each was read; stronger color means a larger share of the citations. Tap a tier to see its sources.

Read in full · 4Abstract read · 19Record only · 2Link checked · 2

Checked by the writer, not a physician. The dots show how each source was read, not that anyone has confirmed the claims.

How this page was made
  • Written from scratch. No source text was copied; facts are restated and cited.
  • The 2017 ACC/AHA/HRS and 2018 ESC guideline full texts could not be retrieved (the publisher sites refused automated access). ACC/AHA/HRS points come from a published summary in The Hospitalist and are cited to both. ESC risk categories and ECG criteria come from the EUSEM consensus, read in full, which builds on the ESC guideline. No recommendation class or strength is stated, because none was read.
  • No published SAEM GRACE syncope guideline was found in PubMed; its three evidence reviews are cited instead. The SAEM GRACE risk-tool review was read in full; other PubMed citations were checked against their abstracts.
  • Adults only. The risk scores enrolled patients aged 16 and older; pediatric syncope evaluation differs and is not covered.
  • Scores are described by their items and published performance only; this page does not calculate a score.
  • Figures 1 and 2 were drawn for this page from cited data. The LITFL syncope page and a First10EM page could not be fetched, so they are not cited.
  • Not for clinical use.
Not for clinical use

AI-generated draft. Not reviewed by a physician. This page is for education only and may contain errors, including in doses. Do not use it to diagnose, treat or make decisions about any patient. Rely on your own clinical judgment, your institution's protocols and current drug labeling. Reading this page does not create a physician-patient relationship. Full terms