Acute aortic syndrome rule-out criteria: development and external validation

Acute aortic syndrome (AAS) is a rare but catastrophic event, presenting with non-specific symptoms that mimic common conditions, and carries a mortality approaching 80% when missed. Current clinical tools, including the Aortic Dissection Detection Risk Score (ADD-RS) and its refinement, the simplified AORTA score, achieve high sensitivity when combined with D-dimer but fail to identify a population at sufficiently low risk to safely avoid further testing. To address this gap, the Acute Aortic Syndrome Rule-Out Criteria (ARC) were developed and validated, a pragmatic bedside tool designed to safely exclude AAS in low-risk patients without additional testing.

A retrospective multistep derivation and validation study was conducted. The derivation cohort included patients with confirmed AAS from the International Registry of Acute Aortic Dissection (IRAD). Starting from existing scores, candidate predictors were identified from the IRAD dataset and sequentially added to achieve sensitivity ≥ 99%. External validation was performed in patients with suspected AAS across three prospective multicentre cohorts (ADvISED, PROFUNDUS, and DAShED). Diagnostic accuracy was assessed using sensitivity and specificity with 95% confidence intervals (CIs).

The final seven-variable ARC included age >60 years, abrupt onset pain, severe pain, hypotension/shock, neurological deficit, pulse deficit, and known aortic disease. Patients were classified as positive if any ARC variable was present and negative only if all variables were absent. In the derivation cohort, ARC achieved a sensitivity of 99.2% (95% CI 99.1-99.3). In the ADvISED/PROFUNDUS cohort (4598 patients, 10.2% AAS prevalence), sensitivity was 100% (95% CI 99.2-100) with specificity 23.7% (95% CI 22.4-25.0). In the DAShED cohort (5346 patients, 0.3% prevalence), sensitivity was 100% (95% CI 76.8-100) and specificity 36.4% (95% CI 35.1-37.7), ruling out 36.3% of patients without missing any AAS cases.

ARC is an externally validated, pragmatic clinical rule demonstrating high diagnostic sensitivity in a large external validation cohort. When incorporated into existing diagnostic pathways, ARC has the potential to reduce unnecessary testing while maintaining diagnostic performance. Further prospective validation is required to confirm its clinical utility.