Early diagnosis and risk stratification of patients with syncope

Background: Syncope is a common and challenging problem in the Emergency Department (ED), representing about 1-2% of patients visits. Early detection of the underlying cause is critical as it defines treatment and prognosis. Cardiac syncope is associated with the highest mortality of all syncope etiologies and requires specific interventions such as implantation of a pacemaker or defibrillator. ED clinicians struggle to rapidly identify the underlying cause and the threat of a possible serious cardiac origin which leads to numerous diagnostics and high hospitalisation rates. In an attempt to improve diagnosis and risk-stratification in syncope patients in the ED, several rules and scores were derived, mostly of mono-centric and rather small studies. Additionally, their external validity has never been assessed and their complexity represent a problem for a rapid and efficient implementation in the ED. Similarly, the diagnostic and prognostic value of some readily available cardiac biomarkers (such as cardiac troponins or B-type Natriuretic peptides) has been investigated in some pilot studies but the small number of patients assessed and the use of poorly sensitive assays resulted in varying results and did not allow for any definitive conclusions.
Aim and Hypothesis: The aim of this thesis is to assess the accuracy of diagnostic and prognostic accuracy of scores and biomarkers in a large international cohort of syncope patients presenting to the ED. First, the accuracy of existing syncope-specific diagnostic and risk-stratification rules will be compared and their complexity put in perspective through a comparison with the CHADS2 score. Second, the diagnostic and prognostic performance of cardiac troponins, as assessed by three different assays, and BNP will be investigated. We hypothesize that complex syncope-specific scores might not reliably diagnose or risk-stratify syncope patients and that both assessed biomarkers, at least in certain subgroups of patients for which the determination of a precise etiology appears particularly difficult, could be of specific interest to improve the diagnosis and risk stratification of patients presenting with syncope to the ED.
Patients and Methods: BASEL IX is an ongoing prospective international multicenter diagnostic cohort study coordinated by the University Hospital Basel. Patients >40y presenting to the ED with a syncope within the 12 last hours are enrolled and blood is drawn for the blinded analysis of the investigational biomarkers. All patients underwent clinical assessment that included standardized and detailed assessment of predefined details of the medical history and syncopal event. The adjudication of the final diagnosis is performed by two independent cardiologists based on all available information after diagnostic work-up of patients as well as the clinical follow-up at 12 months. Patients are followed up to 5 years. The diagnostic endpoint is the diagnostic accuracy for cardiac syncope, the prognostic endpoints are the accuracy to predict death or major cardiovascular events (MACE).
Results: Syncope diagnostic and risk stratification rules showed a moderate accuracy in patients presenting with syncope to the ED (with Area Under The Receiver Operating Curve (AUC) between 0.67 and 0.75 for diagnostic endpoints and between 0.57 and 0.79 for prognostic endpoints) and most of them were, not superior to a readily calculable CHADS2 score. Both assessed biomarkers performed with a moderate-to-good accuracy for the diagnosis and risk stratification of the overall cohort (AUC for diagnostic endpoints between 0.76 and 0.77, AUC for prognostic endpoints between 0.73 and 0.8 depending on the chosen time-point). When assessed in patients for whom the diagnosis stayed unclear despite initial ED evaluation, these biomarkers could provide guidance to the ED physician regarding his decision for hospitalization and further testing.
Conclusion: Diagnosis and risk-stratification of patients with syncope is a challenging task and currently available structured clinical assessments scores do not sufficiently help with initial ED evaluation. Common and readily available cardiac biomarkers seem to represent a valuable tool, especially in patients for whom a first evaluation did not lead to a satisfactory diagnosis.