CT pulmonary angiography in pregnancy and the postpartum period: Diagnostic yield, cardiopulmonary morphometry, and alternative findings

Background Pulmonary embolism (PE) is difficult to diagnose during pregnancy and the postpartum period. Computed tomography pulmonary angiography (CTPA) also provides cardiopulmonary measurements and alternative findings, but non-pregnant reference thresholds may not translate directly to this population. Objective To determine CTPA diagnostic yield, describe cardiopulmonary morphometry and alternative thoracic findings, and explore variables associated with PE. Methods We retrospectively included 101 consecutive pregnant or postpartum women who underwent CTPA for clinically suspected PE — defined as new or worsening dyspnoea, pleuritic chest pain, unexplained tachycardia or hypoxaemia, or unilateral leg swelling prompting the treating clinician to request imaging — at a tertiary maternity centre in 2021–2022. The primary outcome was CTPA-confirmed PE. Great-vessel and ventricular diameters were measured; pulmonary artery-to-ascending aorta (PA/AA) and right-to-left ventricular (RV/LV) ratios were derived. Continuous and categorical variables were compared using Mann-Whitney U and Fisher exact tests. The nine CT morphometric comparisons were adjusted with the Benjamini-Hochberg false-discovery-rate procedure. Results PE was present in 7.9% of examinations (8/101; exact 95% CI, 3.5–15.0%), with a similar yield during pregnancy and postpartum (7.7% [6/78] vs 8.7% [2/23]; P = 1.000). Among the 93 women without PE, median pulmonary trunk and ascending aortic diameters were both 26 mm, and the median PA/AA ratio was 0.96 (IQR, 0.88–1.04) — higher than the 0.81 reported in a non-pregnant reference cohort measured by a different technique. No obstetric, echocardiographic or CT morphometric variable, including the RV/LV ratio, differed significantly by PE status after multiplicity correction. Alternative thoracic findings occurred in 72.3% of examinations (73/101), most commonly linear atelectatic change (54.5%, 55/101) and pleural effusion (53.5%, 54/101). Conclusions CTPA had a low diagnostic yield but frequently identified alternative abnormalities. Median PA/AA measurements in this cohort ran higher than a published non-pregnant, non-contrast reference cohort; because the comparison used a different technique in a different population without a contemporaneous control group, this descriptive gap may reflect technique rather than pregnancy physiology, and isolated pulmonary artery enlargement on CTPA should be interpreted cautiously in this population. Prospective, same-technique controlled studies are needed to establish pregnancy-specific reference values and evaluate structured diagnostic pathways.