Syncope may be relatively benign, or it may be a manifestation of serious underlying pathology. The gestalt of the emergency clinician is highly sensitive but poorly specific for predicting high-risk patients. The creators of the San Francisco Syncope Rule designed it to improve specificity while maintaining high levels of sensitivity. This tool can be utilized to predict low-risk syncope patients who are unlikely to benefit from aggressive workup and hospitalization, thus reducing unnecessary testing, healthcare costs, and potential harms associated with hospitalization.
The most common serious outcome in the studies of the San Francisco Syncope Rule was cardiac dysrhythmia; most of those were bradydysrhythmias or sick sinus syndrome, and the remainder were supraventricular or ventricular dysrhythmias. Other common serious outcomes in patients misclassified as low-risk were stroke and intracranial hemorrhage. Validation studies were not able to replicate the high sensitivity found in the original studies published by the creators of the rule.
A commonly used mnemonic to recall the 5 variables of the San Francisco Syncope Rule is “CHESS”:
Cullen Clark, MD
Marc Probst, MD, MS
Since its inception, the San Francisco Syncope Rule has undergone considerable scrutiny and has been evaluated in validation trials and meta-analyses (key studies include Sun 2007 and Birnbaum 2008). The initial derivation study (Quinn 2004) was a single-center, prospective cohort study of 684 patients with a chief complaint of syncope or near-syncope. Fifty predictor variables selected from literature review and expert consensus were incorporated into a survey completed by the physician evaluating the patient after a syncopal episode. The patients were followed through hospital admission and/or contacted at 7 days after the initial presentation to assess for serious outcomes. The predictor variables associated with serious outcomes were identified on univariate analysis. Recursive partitioning was used to find the best combination of predictor variables that would maximize specificity while preserving high sensitivity. Once the combination of predictors was identified, the rule was applied to the derivation set of patients, resulting in sensitivity of 96% and specificity of 62%.
The rule was validated by Quinn et al 2 years later with another single-center, prospective cohort study of 760 patients. Physicians treated patients with syncope or near-syncope in their usual manner and were then asked to fill out a questionnaire about the patients, which included the 5 San Francisco Syncope Rule syncope predictors. After following the patients for 30 days, the San Francisco Syncope Rule was able to predict serious outcomes in all but 1 patient (98% sensitivity). The rule had higher specificity than physician judgment (56% specificity), with an estimated decrease in overall admission rate by 7%, if applied. The single missed patient was suspected to have had transient ischemic attack related to extensive carotid and vertebral artery disease requiring a stent.
Several external validation studies and meta-analyses have also examined the San Francisco Syncope Rule since the derivation study. Sun et al externally validated the rule with a single-center, prospective cohort study with a sample size of 477 patients. Using definitions, exclusion criteria, and methods similar to the internal validation study, the authors found a sensitivity of 89% and a specificity of 42%. Of the 6 patients misclassified as low-risk, 3 had dysrhythmias, 1 had a cerebrovascular accident, 1 had an intracerebral hemorrhage, and 1 had hypertrophic obstructive cardiomyopathy. This study also compared physician judgment to the clinical decision rule. Physician judgment had a sensitivity of 100% and a specificity of 30% for predicting serious outcomes at 7 days.
Another single-center, prospective, observational cohort study by Birnbaum et al performed at a large urban academic center found that the San Francisco Syncope Rule did not perform as well as previously reported. The study found the rule had a sensitivity of 74% and specificity of 54% for predicting serious outcomes at 7 days. The rule would have decreased the overall admission rate by 41%, but it misclassi-fied 16 of the 61 patients who had serious outcomes as low-risk. Physician judgment in this study was 100% sensitive. In a written response to the study published in the Annals of Emergency Medicine, the derivation study authors criticized the validation study for incorrect application of ECG criteria used to define “high risk” in the study. The original study authors argued that the subjective interpretation of the ECG is enhanced when performed by a treating physician who is familiar with the clinical context. They posited that the sensitivity of the rule in the New York study was reduced because ECGs were evaluated retrospectively by experts rather than by the physician seeing the patient.
Meta-analyses published in the Annals of Emergency Medicine and the Canadian Medical Associa-tion Journal found sensitivity and specificity similar to the external validation study by Sun et al. The Annals of Emergency Medicine meta-analysis found 86% sensitivity and 49% specificity, while the Canadian Medical Association Journal meta-analysis found 87% sensitivity and 52% specificity. The most commonly missed serious outcome in all of the studies assessing the predictive value of the San Francisco Syncope Rule was cardiac dysrhythmia.
Use this score in adult patients presenting with syncope or near-syncope who are back to their neurologic baseline. Do not use it in patients with persistent or new neurologic deficits, alcohol- or drug-related loss of consciousness, definite seizure, or transient loss of consciousness from head trauma.
Ian Stiell, MD, MSc, FRCPC
James V. Quinn, MD, MS
Original/Primary Reference
Validation References
Other References
The Evaluation of Guidelines in SYncope Study (EGSYS) score is for use in patients presenting to the emergency department with syncope. The EGSYS authors suggest the score be used to aid emergency clinicians in screening, not as a substitute for clinical judgment in syncope management.
Cardiac syncope carries high mortality if untreated. A score that identifies these patients can help prevent adverse outcomes from underlying cardiac pathology. Patients with suspected cardiogenic syn-cope should be referred to a cardiologist.
MDCalc recommends the San Francisco Syncope Rule over other syncope risk stratification scores, as it is the most widely validated.
Graham Walker, MD
The EGSYS tool may have limited generalizability. In the original study, a cardiologist specializing in syn-cope was available to review the electrocardiogram in the emergency setting. The score includes specific criteria for electrocardiogram abnormalities. It is less sensitive (95%) than the San Francisco Syncope Rule (96%-98%).
Attilio Del Rosso, MD
Original/Primary Reference
Validation Reference
Other References
Syncope is a common presenting complaint in emergency departments (EDs); it is often benign but sometimes has life-threatening underlying causes. The Canadian Syncope Risk Score (CSRS) identifies patients with syncope who are at higher risk for adverse outcomes within 30 days. Adverse events are defined as any of the following: death, arrhythmia, myocardial infarction, serious structural heart disease, aortic dissection, pulmonary embolism, severe pulmonary hypertension, severe hemorrhage, subarach-noid hemorrhage, or other serious conditions that caused syncope and/or required interventional treatment of syncope. The tool can potentially help avoid unnecessary investigation and/or admission. The CSRS was successfully validated in 2020.
The tool should be used in patients aged ≥16 years who present within 24 hours of syncope. Do not use the CSRS in patients who meet any of the following criteria: prolonged (>5 min) loss of conscious-ness (LOC), change in mental status from baseline, obvious witnessed seizure, major trauma requiring hospital admission, intoxication with alcohol or illicit drugs, language barrier, or head trauma causing LOC. Patients at medium or high risk (CSRS score of ≥1 point) should undergo further investigation for cardiac and noncardiac causes of syncope.
Thomas Yang, MD
Eric Steinberg, DO
The CSRS was derived by Thiruganasambandamoorthy et al in a prospective cohort study based on 6 large Canadian EDs. The study enrolled 4030 patients with syncope, with a mean age of 53.6 years; 55.5% of the patients were women. Forty-three predictors of serious adverse events were studied, with 9 included in the final model. A “serious adverse event” was defined as death, arrhythmia, myocardial in-farction, serious structural heart disease, aortic dissection, pulmonary embolism, severe pulmonary hypertension, severe hemorrhage, subarachnoid hemorrhage, or any other serious condition causing syncope and procedural interventions for the treatment of syncope. The study found that 3.6% of patients sustained a serious adverse event within 30 days of disposition from the ED.
On internal validation, scores ≥-2 were found to have 99% sensitivity and 26% specificity. Scores ≥-1 had 98% sensitivity and 46% specificity. A secondary validation in a group of Canadian EDs found that patients in the low-risk and very low-risk categories appeared to have even lower actual risk percentages (0.7% and 0.3%, respectively).
Venkatesh Thiruganasambandamoorthy, MBBS, CCFP-EM, MSc
Original/Primary Reference
Validation Reference
James Morris, MD, MPH, FACEP
Deborah Diercks, MD, MS, FACEP, FACC; Marc A. Probst, MD, MS, FACEP
June 1, 2021
June 1, 2024 CME Information
4 AMA PRA Category 1 Credits™, 4 ACEP Category I Credits, 4 AAFP Prescribed Credits, 4 AOA Category 2-A or 2-B Credits.
Price: $59
+4 Credits!
Money-back GuaranteeAccredited By
Our Partners
Product Library
