Résumé
Background Severe grade C trauma patients usually benefit from whole-body computed tomography (WBCT) to search for traumatic lesions, in the absence of clinical signs.The systematic use of WBCT in clinically stable patients with severe trauma remains controversial. Objective The aim of this study was to evaluate the diagnostic value of the blood leukocyte count in predicting the existence of traumatic lesions on WBCT in grade C severe trauma patients. Methods This was an observational, retrospective, monocentric study of severe grade C trauma patients who underwent WBCT and leukocyte blood testing in the emergency department. The diagnosis of post-traumatic injury on WBCT was based on the detection of cranial, thoracic, abdominal, large-vessel, spinal and pelvic injuries. The primary endpoint was blood leukocyte count. Results Eight hundred and six patients were included, 301 (37.3%) had severe traumatic lesions and 505 (62.7%) did not.The leukocyte count was significantly higher in patients with traumatic lesions than in those without (15.8 G/L +/- 5.1 vs. 11.0 G/L +/- 4.1; p<0.01).The AUC of the ROC curve derived from this sample was 0.79 [0.75; 0.82], corresponding to a good diagnostic value. Using the optimal threshold of 13.5 G/L, sensitivity was 66% [60%; 71%], specificity 80% [77%; 84%], PPV 67% [61%; 72%] and NPV 80% [76%; 83%]. Conclusion The leukocyte count on venous blood assay was significantly higher in severe grade C trauma patients with severe traumatic lesions on the WBCT but doesn't seem to be a sufficient criterion to avoid WBCT. Its analysis coupled with other biological or clinical criteria could be studied. Trial registration This study was approved by the local institutional review board (no. 24.02.01).