Predicting Echocardiography Findings in Adults Presenting to the Emergency Department With Syncope: An External Validation of the ROMEO Score
John DeAngelis, Gymar Vargas, Robert E. Weiss, Carolyn Winskill, Edward H. Suh, Dana L. Sacco, Nancy Wood, Daniel K. Nishijima, Alan B. Storrow, Jonathan Schimmel, Timothy J. Poterucha, Pierre A. Elias, Nataly Beltre, Marc A. ProbstABSTRACT
Background
Syncope is common in the Emergency Department (ED) and can be associated with structural heart disease (SHD). Transthoracic echocardiography (TTE) is commonly ordered to assess for SHD.
Objective
To externally validate the previously developed ROMEO score, which identifies patients with syncope who are at very low risk of significant findings on TTE.
Design
Secondary analysis of a multicenter, prospective, observational cohort study.
Setting
One community and five academic EDs in the United States.
Participants
Adults (≥ 40 years old) presenting to the ED with syncope or presyncope, without a serious ED diagnosis.
Interventions
Receipt of TTE within 30 days of the index visit.
Measurements
Our primary outcome was the rate of significant findings on TTE within 30 days of the index ED visit. We calculated the sensitivity, specificity, negative and positive predictive values (NPV, PPV), and the area under the curve (AUC) of the ROMEO score.
Results
We enrolled 1287 patients, of whom 427 underwent TTE. 88 (20.6%) had a significant finding. A ROMEO score of zero had a sensitivity of 98.9% and a NPV of 98.6%. The specificity and PPV were 20.2% and 24.3%, respectively. The AUC was 0.83 (95% CI: 0.79 to 0.87).
Limitations
Given the rate of non‐enrollment of screened patients, there is potential for selection bias. Our study sample was biased toward urban, academic centers; the results may not apply to community settings.
Conclusions
The ROMEO score demonstrates strong predictive performance and may be useful to help clinicians identify patients who are unlikely to benefit from TTE.
Trial Registration