Diagnostic performance of the extended focused assessment with sonography for trauma (EFAST) patients in a tertiary care hospital of Nepal

Background: Thoracoabdominal trauma presents a diagnostic challenge for the emergency physician. The introduction of bedside ultrasonography (USG) provides a screening tool to detect hemoperitoneum, hemothorax, pneumothorax and pericardial effusion in torso injuries.

Aim: To evaluate the accuracy of extended focused assessment with sonography for trauma (EFAST) for chest and abdominal injuries performed by first responders in a tertiary care hospital of Nepal.

Methods: This was a prospective study including all trauma patients who obtained either an Injury Severity Score ≥15 or direct trauma to the trunk in 1 year period in the emergency department (ED) of Dhulikhel Hospital-Kathmandu University Hospital. The results of the EFAST were then compared with contrast-enhanced CT (CECT), radiology ultrasound (USG)/chest X-ray, or intraoperative findings when the EFAST was positive. The negative EFAST cases were observed for a minimum of 4 hours in the ED. Descriptive statistics and sensitivity, specificity, positive predictive value, negative predictive value and accuracy were calculated.

Results: Out of 267 cases, 261 patients underwent an EFAST examination. The sensitivity and specificity were 94.8% and 99.5%, respectively. The negative predictive value was 98.53% whereas the positive predictive value was 98.21%. The overall accuracy was 99.4%.

Conclusion: The results of this study suggest that EFAST examination performed by first-line healthcare providers is a useful method for assessment of thoracic and abdominal injuries. EFAST was found to have a high specificity (99.5%) and positive predictive value (98.21%) which indicates that it is an effective technique for detecting intra-abdominal or thoracic injuries. However, the effectiveness of EFAST is limited by its being operator dependent, and thereby human error. For negative EFAST cases, we recommend a monitoring period of at least 4 hours, serial fast scan, or further investigation through other methods such as a CECT.

Level of evidence: Level I.