Reliability of injury grading systems for patients with blunt splenic trauma

Abstract Objectives The most widely used grading system for blunt splenic injury is the American Association for the Surgery of Trauma (AAST) organ injury scale. In 2007 a new grading system was developed. This ‘Baltimore CT grading system’ is superior to the AAST classification system in predicting the need for angiography and embolization or surgery. The objective of this study was to assess inter- and intraobserver reliability between radiologists in classifying splenic injury according to both grading systems. Methods CT scans of 83 patients with blunt splenic injury admitted between 1998 and 2008 to an academic Level 1 trauma centre were retrospectively reviewed. Inter and intrarater reliability were expressed in Cohen's or weighted Kappa values. Results Overall weighted interobserver Kappa coefficients for the AAST and ‘Baltimore CT grading system’ were respectively substantial (kappa = 0.80) and almost perfect (kappa = 0.85). Average weighted intraobserver Kappa's values were in the ‘almost perfect’ range (AAST: kappa = 0.91, ‘Baltimore CT grading system’: kappa = 0.81). Conclusion The present study shows that overall the inter- and intraobserver reliability for grading splenic injury according to the AAST grading system and ‘Baltimore CT grading system’ are equally high. Because of the integration of vascular injury, the ‘Baltimore CT grading system’ supports clinical decision making. We therefore recommend use of this system in the classification of splenic injury.