Objective: The objective of the study was to obtain the best estimates of the test performance of abdominal ultrasonography (US) for identifying children with intraabdominal injuries (IAIs).
Methods: We gathered studies on the use of abdominal US in injured children from the following sources: a MEDLINE and Embase search, hand searches of 5 specialty journals and 4 clinical textbooks, the bibliographies of all identified articles, and contact with experts. Both prospective and retrospective studies were included if they used abdominal US for the detection of intraperitoneal fluid or IAIs in blunt trauma patients less than 18 years of age. All authors independently abstracted data from the selected studies. Disagreements between abstractors were resolved by mutual agreement.
Results: Twenty-five articles met the inclusion criteria, and 3838 children evaluated with abdominal US were included. Abdominal US had the following test characteristics for identifying children with hemoperitoneum: sensitivity, 80% (95% confidence interval [CI] 76%-84%); specificity, 96% (95% CI 95%-97%); positive likelihood ratio, 22.9 (95% CI 17.2-30.5); and negative likelihood ratio, 0.2 (95% CI 0.16-0.25). Using the most methodologically rigorous studies, however, yielded the following test characteristics of abdominal US for identifying children with hemoperitoneum: sensitivity, 66% (95% CI 56%-75%); specificity, 95% (95% CI 93%-97%); positive likelihood ratio, 14.5 (95% CI 9.5-22.1); and negative likelihood ratio, 0.36 (95% CI 0.27-0.47).
Conclusions: Abdominal US has a modest sensitivity for the detection of children with hemoperitoneum; however, its test performance characteristics worsen when only the most methodologically rigorous articles are included. A negative US examination has questionable utility as the sole diagnostic test to rule out the presence of IAI. Because of the high risk of IAI, a hemodynamically stable child with a positive US examination should immediately undergo abdominal computed tomographic scanning.