Predictive factors of liver injury in blunt multiple trauma.

Introduction: This study was conducted to clarify whether injuries that are likely to be revealed by initial clinical and conventional radiological examination at the trauma bay (e.g., right-side rib fractures) meaningfully contribute to the prior probability of accompanying hepatic lesions in multiple injured patients.
Material and Methods: Fifty-five subjects (sampled from a cohort of 218 patients) with liver injury fulfilling the definition of polytrauma were compared with 55 polytrauma patients without liver injury. Controls were individually matched for age, gender, and Injury Severity Scores. Whole-body, helical, contrast-enhanced computed tomography was applied to all participants. We modeled independent predictors of liver involvement by conditional logistic and random-effects regression analysis.
Results: In the present sample, the prevalence of hepatic injury was 25.2%. Neither the injury mechanism (car crash, pedestrian accident, fall from height) nor certain accompanying injuries (right-side serial rib fractures, lumbar spine fractures) predicted the presence of hepatic injury. Liver injury was particularly unlikely in bikers [odds ratio (OR) 0.78, 95% confidence interval (CI) 0.59-1.03] and patients with left-side rib fractures (OR 0.80, 95% CI 0.66-0.98).
Discussion: There are no index injuries that will reliably indicate the presence of liver involvement in multiple trauma cases. Also, the absence of these injuries cannot rule out liver damage.