Ultrasound assessment of gastric volume in critically ill patients

Abstract

To assess the feasibility and validity of ultrasonographic measurement of gastric antral cross-sectional area (usCSA) in critically ill patients to predict gastric volume and the use of computed tomography (CT) as a reference to measure gastric volume. This single-center, prospective, cross-sectional study included 55 critically ill patients who had an abdominal CT scan. usCSA measurements were performed within the hour preceding the CT scan. Gastric volumes were measured on the CT scan using semiautomatic software. The feasibility rate, performing conditions (% "good" and "poor"), internal and external validity of antral usCSA measurements, performed by an ICU physician, were assessed to predict gastric volume. Antral usCSA measurements were feasible in 95% of cases and were positively correlated with gastric volume measured by the CT scan when performed in "good" conditions (65%) (r = 0.43). There was good reproducibility of measurements (intraclass correlation coefficient of 0.97, CI 95% 0.96-0.99) and there was clinically acceptable agreement between measurements performed by radiologists and intensivists (bias -0.12 cm(2)). The receiver operating characteristic curve identified a cutoff value of 3.6 cm(2) that discriminated an "at-risk stomach" (volume >0.8 mL/kg) at a sensitivity of 76% and a specificity of 78%. Ultrasonographic measurement of antral CSA is feasible and reliable in the majority of critically ill patients. This technique could be useful to manage critically ill patients at risk of aspiration or with enteral feeding.