ACEP released a game-changing clinical policy on nontraumatic thoracic aortic dissection (AoD) that shifts the approach to risk stratification for patients with suspected AoD: clinical decision tools combined with a D-dimer are officially in.¹ A decade after an ACEP clinical policy declared we could rely on neither clinical decision tools nor D-dimer alone, the tides have changed.²

What happened? AoD is taught as a “can’t miss” diagnosis in emergency medicine. Yet AoD is rare, with an estimated incidence of 2 to 4 per 100,000 individuals.^3,4 Two seemingly contradictory truths exist regarding the contemporary evaluation of AoD in the emergency department (ED) — we miss aortic dissections although we simultaneously overtest patients for this disease process.⁵ This paradox is driven by the variable clinical presentations of nontraumatic AoD and the historical lack of strategy to reasonably exclude AoD outside of advanced imaging such as computed tomographic angiogram (CTA).

Frustratingly, few history and physical examination features meaningfully change the pre-test probability of AoD. A matched case-control study evaluating 194 cases of AoD and 776 controls with an ED visit for truncal pain of fewer than 14 days duration and no clear diagnosis found abysmal diagnostic characteristics for clinical features. The single negative likelihood ratio (LR) that was able to decrease the probability of AoD was abrupt onset of pain (-LR 0.07; 95% CI 0.03-0.14). The study found essentially useless negative likelihood ratios for the absence of classic features such as pulse deficit, murmur, bilateral blood pressure differential, subjective neurological deficit, and widened mediastinum on chest radiograph (CXR).⁶ Further, bedside testing such as point-of-care ultrasound has been found to be specific for AoD but insufficiently sensitive.⁷

Click to enlarge.

Given these limitations, it is unsurprising that attempts to validate bedside risk stratification tools for AoD have fallen short. Although several scores exist, the Aortic Dissection Detection Risk Score (ADD-RS) is the most widely studied. The tool is comprised of three questions: (1) high-risk condition for AoD (2) high-risk pain features (3) high-risk examination features. Alone, the risk score is insufficiently sensitive, with a pooled sensitivities of 43.4 percent to 94.6 percent, depending on the scoring threshold used to define low risk.⁸

Enter new evidence. Owing to prior limitations, investigators have evaluated a framework for AoD analogous to the evaluation of pulmonary embolism, in which D-dimer testing is used in combination with clinical risk stratification. In a prospective study of 1,850 ED patients with a differential diagnosis including acute aortic syndromes, the failure rate of an ADD-RS less than or equal to 1 combined with a D-dimer less than 500 ng/mL was 0.3 percent (95 percent CI 0.1-1).⁹ As a result, the new ACEP clinical policy states: In adult patients who have an AoD detection risk score (ADD-RS) of 1 or less (i.e., low risk) and a highly sensitive D-dimer of less than 500 ng/mL, emergency physicians can exclude acute nontraumatic AoD without obtaining advanced imaging (ie, CTA, MRI, or TEE) (Level B).