What scores exist for assessing risk in patients with syncope?

Syncope Risk Stratification Scores

Multiple validated risk scores exist for syncope, but the 2017 ACC/AHA/HRS guidelines give them only a Class IIb recommendation because they have not outperformed unstructured clinical judgment and have significant methodological limitations. 1

Primary Validated Scores

The following nine risk scores have been formally studied and validated, though with varying degrees of rigor 1:

Short-Term Risk Prediction (7-30 days)

• San Francisco Syncope Rule (SFSR, 2004): Predicts 7-day serious events with 99% negative predictive value; includes abnormal ECG, dyspnea, hematocrit <30%, systolic BP <90 mm Hg, and heart failure history 1, 2

• Canadian Syncope Risk Score (CSRS, 2016): Most recently developed and extensively validated tool with highest predictive power for 30-day mortality (AUC 0.869); includes 9 variables such as predisposition to vasovagal syncope, heart disease, abnormal vital signs, elevated troponin, and specific ECG abnormalities 3, 2, 4

• ROSE Score (2010): Predicts 30-day serious events with 98% negative predictive value using abnormal ECG, B-natriuretic peptide, hemoglobin, oxygen saturation, and fecal occult blood 1

• Boston Syncope Rule (2007): Achieved 100% negative predictive value for 30-day serious events; includes acute coronary syndrome symptoms, worrisome cardiac history, family history of sudden cardiac death, valvular heart disease, conduction disease signs, volume depletion, persistent abnormal vital signs, and primary CNS event 1, 5

• STePS Score (2008): Predicts 10-day serious events using abnormal ECG, trauma, absence of prodrome, and male sex 1

Intermediate-Term Risk Prediction

• Syncope Risk Score (2009): Evaluates 30-day serious events in 2,584 patients with 97% negative predictive value; includes abnormal ECG, age >90 years, male sex, positive troponin, arrhythmia history, and systolic BP >160 mm Hg 1

Long-Term Risk Prediction (1 year)

• Martin Score (1997): Predicts 1-year death/arrhythmia with 93% negative predictive value using abnormal ECG, age >45 years, ventricular arrhythmias, and heart failure 1

• OESIL Score (2003): Predicts 1-year mortality with 100% negative predictive value; includes abnormal ECG, age >65 years, absence of prodrome, and cardiac history 1, 2

• Sarasin Score (2003): Predicts inpatient arrhythmia with 98% negative predictive value using abnormal ECG, age >65 years, and heart failure 1

Etiology-Focused Score

• Del Rosso Score (2008): Identifies cardiac etiology with 99% negative predictive value; evaluates abnormal ECG/cardiac history, palpitations, exertional syncope, supine position, precipitants, and autonomic prodrome 1

Critical Limitations to Understand

The ACC/AHA/HRS explicitly warns that these scores have major methodological flaws that limit their clinical utility 1:

• Inconsistent definitions of syncope, outcomes, and time frames across studies 1
• Many scores include patients with serious outcomes already identified in the ED, creating bias toward detecting "obvious" events 1
• Composite outcomes combine events with different pathophysiologies, reducing clinical relevance 1
• Small sample sizes limit model reliability 1
• Limited external validation across different populations and settings 1
• Most importantly: risk scores have not demonstrated superiority over unstructured clinical judgment 1, 2

Practical Clinical Application

Rather than relying on formal scores, the ACC/AHA/HRS recommends focusing on individual high-risk features for disposition decisions 6:

High-Risk Features Requiring Immediate Evaluation

Short-term risk factors (<30 days) 1:

• Abnormal ECG (any rhythm other than sinus, conduction delays, Q waves, ST abnormalities, prolonged QT) 1
• Systolic BP <90 mm Hg or >180 mm Hg 1
• Elevated troponin above 99th percentile 1
• Dyspnea or signs of heart failure 1
• Anemia (low hematocrit/hemoglobin) 1

Long-term risk factors (>30 days) 1:

• Male sex 1
• Older age 1
• Absence of nausea/vomiting before syncope 1
• Ventricular arrhythmias 1
• Structural heart disease or heart failure 1
• Cancer 1
• Cerebrovascular disease 1
• Diabetes mellitus 1
• High CHADS-2 score 1
• Lower glomerular filtration rate 1

Mandatory Initial Evaluation

All syncope patients require detailed history, complete physical examination, and 12-lead ECG regardless of risk score 1, 6. This foundational assessment remains more valuable than any risk score for guiding management decisions 6.

Recent Evidence Synthesis

A 2025 systematic review found that only three CDRs (SFSR, CSRS, OESIL) have been validated in more than two studies, with significant overlap in their operating characteristics and generally low-quality evidence for routine clinical use 2. The CSRS demonstrated the highest predictive accuracy in a 2023 comparison study, with sensitivity of 82.6% and specificity of 81.9% at a cut-off value of 0.5 3.