Contribution of Point of Care Ultrasound by the Emergency Physician to Rule Out the Small Bowel Obstruction: a Diagnostic, Multicenter Study

Small Bowel Obstruction (SBO) is a frequent pathology in the emergency department (ED). Diagnosis is currently based on abdominal CT scan (CT). Moreover, CT is warranted to determine the therapeutic strategy in patients with SBO which could include medical treatment; surgical intervention or both. However, CT is associated with drawbacks such as radiation exposure, increased cost and ED length-of-stay. In a prospective observational study, a SBO was excluded by CT in 45% \[95%CI: 37-53\] of patients. There is, thus, a need for improving the appropriateness of CT-scan for suspected SBO. A recent meta-analysis showed that Point of care ultrasound (POCUS) had a good diagnostic accuracy (sensitivity 83% \[95%CI 71.7%-90.4%\]), specificity 93% \[95%CI 55.3%-99.3%\]). Another meta-analysis found rather similar results (sensitivity 83% \[(95% CI 89.0% to 94.7%\], specificity 96,6% \[95% CI 88.4% to 99.1%\]). In order to improve the negative predictive value of POCUS for its implementation as a rule-out strategy, CHU of Nantes emergency unit studied the combination of POCUS with Gestalt pre-test probability of SBO determined by the emergency physician. This SBO probability classified the patients as low, moderate or high risk of SBO. In patients with low or moderate Gestalt probability, CHU of Nantes emergency unit found that this combined strategy had a sensitivity of 100% \[95% CI: 88-100\] and NPV 100% \[92-100%\]. By (i) focusing on patients with a low or moderate Gestalt clinical probability and (ii) increasing the number of patients included, CHU of Nantes emergency unit intends to demonstrate that POCUS is able to exclude SBO in this population. This would avoid unnecessary CT and thus lower costs, ED length-of-stay and hospital radiologists workload. A POCUS will be performed followed by a CT (gold standard). The main objective will be the ability of POCUS to rule-out SBO in patients with low or moderate Gestalt clinical probability.