Abstract
BACKGROUND: The transfusion of red blood cell (RBC) concentrates represents an emerging approach in paediatric trauma management, even in the prehospital period. Nevertheless, distinctive parameters for predicting the need for transfusion in children are still lacking. This study aimed to identify indicators for early in-hospital RBC transfusions that are primarily accessible either in the prehospital or in the early in-hospital setting in a large paediatric trauma cohort. METHODS: This study comprised a retrospective analysis of the German TraumaRegister DGU(®). It included children and adolescents aged 1 to 16 years from Germany, Austria, and Switzerland over a 15-year period. Contingency tables were used to identify risk factors, which were then assessed through multivariate regression analysis. The model's predictive capacity was evaluated using the receiver operating characteristic (ROC) curve. RESULTS: A total of 11,849 patients were included, with RBC transfusion performed in 5.9% of cases. Polytraumatised patients (adjusted odds ratio (adj. OR) 4.18 [95% confidence interval 3.26-5.34]) and those with penetrating injuries (adj. OR 4.32 [2.96-6.30]) and abdominal injuries (adj. OR 4.18 [3.34-5.24]) exhibited the highest risk of requiring an RBC transfusion. The need for cardiopulmonary resuscitation (adj. OR 2.46 [1.84-3.28]), endotracheal intubation (adj. OR 2.51 [1.93-3.28]), and Glasgow Coma Scale (GCS) ≤ 8 (adj. OR 2.49 [1.85-3.36]) were also significant, but weaker, predictors. A model based on the mentioned parameters achieved an area under the ROC curve of 0.87 [0.85-0.88], whereas predictive performance was lower but still acceptable when only parameters available in the prehospital setting were included (AUC 0.80 [0.78-0.82]). CONCLUSION: The likelihood of requiring an RBC transfusion is increased in cases of polytrauma, abdominal and penetrating trauma, patients with a GCS ≤ 8, and those requiring tracheal intubation or cardiopulmonary resuscitation. Therefore, the proposed risk factors can help identify patients at risk of severe haemorrhage and subsequent transfusion requirement. CLINICAL TRIAL NUMBER: Not applicable.