Prognostic accuracy of syncope clinical prediction rules in older adults in the emergency department

Abstract

Abstract Study objective The objective of this study is to evaluate the prognostic accuracy of existing rules (San Francisco Syncope Rule [SFSR], Canadian Syncope Risk Score [CSRS], and FAINT score) in older adults. Methods This is a cohort study of adults aged ≥60 years presenting to an academic emergency department (ED) with syncope or near syncope. We used original criteria for all rules except for the FAINT score, in which N‐terminal pro–brain natriuretic peptide was largely missing from the extracted data. Patients were deemed positive for each rule if classified as non‐low risk. The primary outcome was the presence of 30‐day serious outcome, as defined by syncope research guidelines. Sensitivity and negative likelihood ratio (NLR) were calculated with 95% confidence intervals (CIs). Results A total of 404 ED visits (mean age of patients, 75.5 years) were included. Of these, 44 (10.9%) had a 30‐day serious outcome, and 24 (5.9%) had incomplete 30‐day follow‐up. SFSR was positive for 280 of 380 visits with complete follow‐up. Its sensitivity and NLR for predicting 30‐day serious outcomes were 86.4% (95% CI, 72.0%–94.3%) and 0.53 (95% CI, 0.25–1.15), respectively. The CSRS was positive for 299 of 380 visits (sensitivity was 88.6% [95% CI, 76.4%–95.7%], and NLR was 0.50 [95% CI, 0.22–1.17]). The modified FAI(N)T score was positive for 318 of 380 visits (sensitivity was 90.9% [95% CI, 77.4%–97.0%], and NLR was 0.53 [95% CI, 0.20–1.38]). Conclusion Existing rules are suboptimal to predict 30‐day serious outcomes in older adults presenting with syncope or near syncope to the ED.

Keywords

• clinical prediction rules
• geriatric emergency medicine
• older adults
• pre‐syncope
• prognosis
• risk stratification