Trauma in pregnant women: assessing detection of post-traumatic placental abruption on contrast-enhanced CT versus ultrasound
To evaluate detection of post-traumatic placental abruption with contrast-enhanced CT (CECT) and comparison with Ultrasound (US).
Picture Archive and Date System database at a level-1 trauma center was retrospectively reviewed using keywords pregnancy, trauma, and/or placental abruption over 10 years. CT was compared to US, if performed within 24 h. Two subspecialty-trained radiologists blindly reviewed the studies. Placental features on delivery and pregnancy outcomes were used as reference standard. Lack of adverse pregnancy/fetal outcome was treated as the absence of abruption.
CECT was performed in 36 patients, with 27 US within 24 h. There were three complete and eight partial abruptions. Reader sensitivity for CT was 100% for both reviewers; however, specificity was 54.5% and 56.7%. No sonographic abnormality was noted in both partial and complete abruption. Using kappa statistics, inter-observer agreement was low for both CT (0.169) and US (0.078). False-positive reads were from misinterpretation of normal placental structures like cotyledons, age-related infarcts, and marginal sinus of the placenta.
CECT identifies post-traumatic placental abruption with high sensitivity but low specificity for clinically significant abruptions, and performs better than US. Pitfalls from normal placental structures mimicking abruption should be avoided. US markedly underdiagnoses abruption.