Prospective Validation of Syncope Risk Scores in the ED

Key Takeaways

• FAINT and CSRS scores showed moderate discrimination and adequate calibration in this prospective multicenter cohort. CSRS below 0 had 91.9% sensitivity and 97.5% NPV for serious adverse outcomes, and FAINT of 0 had 96.7% sensitivity and 98.8% NPV for serious cardiac outcomes.
• FAINT was more sensitive than physician-estimated risk for serious cardiac outcomes, while CSRS was similar to physician estimates for serious adverse outcomes.

In a JAMA Network Open cohort study of 1263 adults aged 40 years or older, the FAINT score had an AUROC of 0.76 for 30-day serious cardiac outcomes, and across 6 urban U.S. emergency departments, the Canadian Syncope Risk Score had an AUROC of 0.72 for 30-day serious adverse outcomes.

The prospective, observational, multicenter study ran from September 2020 to September 2024 across 6 urban U.S. emergency departments. Adults aged 40 years or older were enrolled after syncope or presyncope only if no serious diagnosis was identified during the initial emergency department evaluation. Among 1263 analyzed patients, 74 (5.9%) had a serious adverse outcome and 62 (4.9%) had a serious cardiac outcome within 30 days.

The Canadian Syncope Risk Score and FAINT score targeted different 30-day endpoints, with CSRS covering cardiac and noncardiac serious adverse outcomes and FAINT focusing on serious cardiac outcomes. Unstructured physician-estimated risk was available for 1076 patients (85.1%), and outcomes were ascertained by record review and telephone follow-up.

At low-risk thresholds, CSRS below 0 had 91.9% sensitivity and 97.5% NPV for serious adverse outcomes, while FAINT of 0 had 96.7% sensitivity and 98.8% NPV for serious cardiac outcomes. In the reduced sample, physician-estimated risk below 2% had 87.5% sensitivity for serious adverse outcomes and 79.2% for serious cardiac outcomes. CSRS did not significantly differ from physician-estimated risk for serious adverse outcomes. FAINT had higher sensitivity for serious cardiac outcomes by 18.7 percentage points (95% CI, 6.8% to 30.6%; P=.002). A secondary CSRS threshold of less than 1 classified 664 patients (52.6%) as low risk, with 77.7% sensitivity and 97.3% NPV.

The authors wrote that these scores may complement physician judgment, but validation alone did not establish better outcomes, lower emergency department use, or smoother workflow.