Objective: There is a large literature concerning the estimation of gastric content in third-trimester pregnant women but their conclusions remain contradictory. Our objective was to compare gastric content in pregnant women and in non-pregnant ones using gastric ultrasound.

Design: An observational two centres study was designed.

Methods: The antral cross-sectional area (CSA) in the semi-recumbent position (SRP), primary outcome, and in the right lateral position (RLD) were measured in pregnant women scheduled for a caesarean section (CS group) or non-pregnant women who underwent hysteroscopy (HS group) just before surgery. Perlas' score was also evaluated.

Results: One hundred and twenty-six patients were analysed, 61 patients in the CS group and 65 in the HS group. Antral CSA, measured in the SRP, was greater in the CS group than in the HS group (350 mm2 [236 - 415] vs 247 mm2 [180 - 318]; (P=0.001). This difference remained significant when the analysis was adjusted for fasting duration (P<0.001) or anxiety (P=0.001) or both (P<0.001). Among secondary outcomes, Perlas scores did not differ between groups (P=0.860). Concordance between antral CSA, categorised as corresponding to an empty or full stomach, and Perlas score was very poor.

Conclusions: Our results bring arguments to consider that individuals among pregnant women could have a delayed gastric emptying and support performing a gastric ultrasound in term pregnant patients prior to general anaesthesia.

eCRF