Table 1.

Computed tomography angiography (CTA) signs, involved mechanism, association with short-term mortality, and level of evidence.

CTA sign	Pathophysiology	Proportion of patients with PE and positive sign	30-day mortality OR (95% CI)	Data based on	Interobserver variability∗	Level of evidence†
Main pulmonary artery size	Extension of arterial obstruction and pulmonary hypertension/right ventricular afterload	Variable	Not statistically significant	4 small retrospective studies and two meta-analyses [17, 23]	Fair	Low
Emboli burden			Not statistically significant	Meta-analysis of 9 studies [23]	Fair	Good
Emboli position			2.2 (1.3–3.9) for main or lobar arteries localisation	Meta-analysis of 3 studies [23]	Excellent	Good
Blood flow on dual-energy CTA			3.8 (1.0–14.6)‡ for a defect >5%	2 small retrospectives studies	Unknown	Low
Right-to-left ventricular ratio	Right-ventricular dysfunction	>50%	2.1 (1.6–2.8) for all-comers with pulmonary embolism	One meta-analysis (>5000 patients) [17]	Excellent	Good
			1.7 (1.1–2.7) for normotensive patients	Two meta-analyses (>2000 patients each) [4, 17]	Excellent	Good
Interventricular septal bowing		20%	1.8 (1.2–2.7)	One meta-analysis (1422 patients) [17]	Poor	Low
Retrograde reflux of contrast	Tricuspid regurgitation, increased atrial pressure/right- ventricular preload	20%	3.1 (1.2–7.7)§	>6 small and 1 intermediate-size retrospective study	Fair-excellent	Low
Azygos vein size		Variable	1.5 (1.1–2.0)||	1 small retrospective study	Fair	Low

*Based on kappa statistic: <0.4 poor; 0.4–0.75 fair; >0.75 excellent; ^†global appreciation of scientific evidence based on the number, size, quality of the studies, and availability of a meta-analysis; ^‡calculated from Bauer et al. [24]; ^§calculated from Aviram et al. [25]; ^||14-day mortality [26].

CTA: computed tomography angiography; OR: odds ratio; 95% CI: 95% confidence interval.