Development and validation of an ECG-based cardiac syncope risk calculator

Abstract

Abstract Background The early diagnosis of cardiac syncope is often challenging. We therefore developed an ECG-based risk calculator as an aid for rapid rule-out or rule-in of cardiac syncope and aimed to validate this decision tool. Methods In a prospective diagnostic international multicenter study (derivation cohort), 2007 patients, 40 years or older, presenting with syncope to the emergency department were recruited. The primary diagnostic outcome, cardiac syncope, was centrally adjudicated by two independent cardiologists using all clinical information obtained during syncope work-up including 12-month follow up. 12-lead ECG was recorded at presentation and read by residents blinded to clinical information. Significant ECG predictors of cardiac syncope were identified using penalized backward selection. Findings were validated in an independent US multicenter cohort with 2'269 syncope patients. Results In the derivation cohort (median age 71 years, 40% women), centrally adjudicated cardiac syncope was present in 267 patients (16%). Seven ECG criteria (rhythm, heart rate, corrected QT-interval, ST-segment depression, atrioventricular-block, bundle-branch-block and ventricular extrasystole/non-sustained ventricular tachycardia) were identified as significant predictors for cardiac syncope and combined into the bAseL Ecg Risk calculaTor for Cardiac Syncope (ALERT-CS). Diagnostic accuracy of ALERT-CS for cardiac syncope, as quantified by the area under the receiver-operating characteristics curve (AUC), was high (0.80, 95%-confidence interval (CI) 0.77–0.83) and significantly higher compared to the EGSYS score (0.73, 95% CI 0.70–0.76, p<0.001). In combination, ALERT-CS significantly increased the AUC of BNP (0.82, 95% CI 0.79–0.85 vs 0.77, 95% CI 0.74–0.81, p=0.003), hs-cTnT (0.84, 95% CI 0.0.81–0.87 vs 0.77, 95% CI 0.74–0.80, p<0.001) and integrated clinical judgment in the ED (0.90, 95% CI 0.89–0.92 vs 0.87, 95% CI 0.84–0.90, p<0.001). A predicted probability for cardiac syncope below 5.5% by ALERT-CS identified 138 patients (8%) eligible for triage towards rapid rule-out of cardiac syncope with a sensitivity of 99%. A predicted probability above 37.5% identified 181 patients (11%) eligible for triage towards rapid rule-in of cardiac syncope with a specificity of 95%. Prognostic verification for 30-day major adverse cardiac events (MACE) showed a high rate of MACE in the rule-in group and a very low rate of MACE in the rule-out group (Figure). External validation (median age 72 years, 48% women) showed similar diagnostic accuracy (AUC 0.76, 95% CI 0.73–0.79) and prognostic results. Conclusion Combining seven ECG criteria within the simple ALERT-CS may aid ED physicians in the early rule-out or rule-in of cardiac syncope. Figure 1 Funding Acknowledgement Type of funding source: Public grant(s) – National budget only. Main funding source(s): Swiss National Science Foundation, Swiss Heart Foundation