Title: Validation of a prediction rule for the identification of children with intra-abdominal injuries after blunt torso trauma
Article Citation: Holmes JF, Mao A, Awasthi S, McGahan JP, et al. Validation of a prediction rule for the identification of children with intra-abdominal injuries after blunt torso trauma. Ann Emerg Med. 2009 Oct;54(4):528-33. PMID: 19250706
What we already know about the topic:
Trauma is a leading cause of morbidity and mortality in children older than a year of age and abdominal trauma is the third leading cause of death. A prediction rule using physical exam findings along with readily available laboratory data exists to guide decision making in obtaining CT imaging for abdominal trauma, but has never been externally validated. The prediction rule consists of the following 6 “high-risk” variables: low age-adjusted systolic blood pressure; abdominal tenderness; femur fracture; aspartate aminotransferase level greater than 200 U/L and alanine aminotransferase level greater than 125 U/L; hematuria level greater than 5 RBCs/high powered field; and an initial hematocrit level less than 30%. Children not meeting any of these variables were at low risk of serious intra-abdominal trauma and unlikely to benefit from an abdominal CT.
Why this study is important:
In this study, the authors validate a previous internally validated decision making rule to identify pediatric patients with abdominal trauma in which abdominal CT would be of no significant benefit.
Brief overview of the study:
This is a prospective observational cohort study to externally validate a clinical prediction tool to identify children at a very low risk for intra-abdominal trauma, as well as those at high risk for intra-abdominal trauma. This clinical prediction tool was derived in an earlier study by the same authors. Children not meeting any of the 6 “high risk” variables were at low risk of serious intra-abdominal trauma and unlikely to benefit from abdominal CT. The study enrolled children younger than 18 years who had blunt torso trauma and underwent a definitive diagnostic test (abdominal CT scan, diagnostic peritoneal lavage, diagnostic laparoscopy, or laparotomy) to evaluate for the presence of an intra-abdominal injury. Of the 1,324 children with blunt torso trauma,1,119 (85%) had the necessary variables in the prediction rule documented for purposes of assessment and made up the validation study sample. 754 patients tested positive based on the prediction tool and 365 patients tested negative for the rule, including 8 patients later found to have intra-abdominal trauma. Based on this, the authors conclude that the clinical prediction rule demonstrated good sensitivity (94.9%), and one third of abdominal CT scans could be obviated if the rule were strictly applied.
My first issue with this study was that the authors do not include newer technologies beneficial in identifying patients at high risk for intra-abdominal trauma, specifically a FAST exam. Since this article was published, a FAST exam has become the standard of care in most emergency physicians’ trauma exam and additional work should be done that takes this piece of data into account. Additionally, the authors point out several additional variables that have a higher relative risk then the variables included in their tool. Inclusion of these variables might allow for a more sensitive predictive tool. Finally, this study was done only at one urban trauma center. Both the type of patient and type of trauma might vary depending on various sites and the tool should be validated in a multi-center study before general use.
Take home points:
While the authors do validate a clinical predictive tool to identify low risk intra-abdominal injuries in pediatric patients in which abdominal CT would be of little benefit, this study is outdated. At the time of the study, ultrasound was not universally available and was, therefore, not incorporated into their methods/criteria. Additionally, this tool needs to be studied in a large multi-center cohort (external validation) prior to widespread adoption.