Publication:

Assessment of the San Francisco syncope rule in detecting high-risk cardiac and neurological causes of syncope

Syncope is a common clinical presentation in emergency departments (ED), often posing significant diagnostic challenges due to its broad differential and the potential for life-threatening underlying causes. Accurate risk stratification is essential to differentiate patients requiring urgent intervention from those who can be safely discharged. The San Francisco syncope rule (SFSR) was developed to aid clinicians in identifying patients at risk of short-term serious outcomes. Despite its widespread adoption, evidence regarding its reliability and predictive accuracy remains mixed. External validation studies have reported variable sensitivity and specificity, with some highlighting its failure to detect neurologic or subtle cardiac causes of syncope. Comparative analysis with other stratification tools such as the OESIL score, ROSE rule, and EGSYS score reveals key differences in design and clinical utility. Each model offers unique strengths but also exhibits important limitations when applied across heterogeneous patient populations. Inconsistent definitions of serious outcomes and variable study methodologies have contributed to difficulty in standardizing syncope assessment across settings. Additionally, neurologic causes are frequently underrepresented in many tools, reducing their diagnostic reach. Biomarkers and imaging have been proposed as adjuncts but are limited by access, cost, and timing. Recent interest has turned to machine learning models capable of integrating broader clinical variables to generate personalized risk profiles. Although early results are encouraging, such approaches require rigorous external validation before widespread clinical use. Overall, existing models offer useful guidance but are not definitive. Risk stratification in syncope should remain a dynamic process informed by evidence-based tools, clinician experience, and ongoing research into more adaptable, data-rich strategies capable of addressing the complexity of syncope presentations in modern emergency care.