Study Setting and Population
This was a single centre study, conducted retrospectively in Yokohama City University Medical Centre (Yokohama, Japan). Our centre is one of the two Yokohama City Major Trauma Centres (YCMTCs), which were established to serve a population of 3.7 million, including 446,000 children, and to provide 24/7 trauma care by a specialised team, including an interventional radiologist, on-call around the clock [12,13]. Before severe trauma patients arrive at our centre, an in-hospital trauma code is activated and preparations for blood transfusions, urgent surgery, and interventional radiology (IVR) are initiated. Whenever trauma patients are in an unstable condition, we can perform urgent surgical and/or radiological interventions within the first 30–60 minutes of arrival to hospital.
For this study, we used the dataset from our centre to include information between January 1, 2014, and December 31, 2017, which initially yielded the data for 19,207 patients. The inclusion criteria for this study were: blunt torso trauma patients and patients who underwent TAE. Patients who were dead on arrival were excluded from this study. Figure 1 presents a flow chart of the patient population in this study.
Emergency department algorithm in blunt torso patients with haemorrhage at YCMTC
All trauma patients underwent evaluation and trauma care in the acute care phase, according to the Japanese trauma evaluation and care guidelines [14]. The general approach is based on the patients’ haemodynamic stability and response to fluid resuscitation. If systolic blood pressure is less than 90 mmHg in an adult and less than the age-related baseline value in children [14], these patients are considered to be in hypovolemic shock.
Regarding fluid resuscitation, 20 ml of crystalloid solution per body weight (kg) is rapidly administered, followed by whole blood transfusion. First, if the patients do not respond to fluid resuscitation but achieve haemodynamic stability, they are assessed as ‘non-responders’ and treated by surgical intervention with damage control techniques. In case of persistent haemodynamic instability with ongoing arterial bleeding, subsequent TAE is performed. Second, if the patient responds to fluid resuscitation and maintains haemodynamic stability, the patient is assessed to be a ‘responder’ and will undergo contrast-enhanced computed tomography (CT) scan. In case of arterial extravasation, TAE is performed. Finally, if the patient responds to fluid resuscitation and achieves haemodynamic stability at least temporarily, they are assessed to be ‘transient-responder’ and undergo contrast-enhanced CT scan. According to the patient’s condition and the results of the CT scan, the trauma team will perform TAE and/or surgical intervention for the transient-responders.
All blunt trauma patients with haemorrhage in this study underwent trauma care according to our institutional algorithm, as aforementioned. Therefore, the trauma team performed operative and/or non-operative intervention, using the same therapeutic strategy, regardless of age; however, in the case of paediatric patients, all interventions were performed under the management of general anaesthesia and mechanical ventilation. The technique for TAE started with percutaneous femoral artery vascular access with sheath placement. Arterial puncture in younger paediatric patients used an ultrasonically guided puncture needle to ensure safety and reliable manual operation. A 4-, 5-, or 6-Fr sheath was used to maintain arterial access throughout the procedure. Finally, all image diagnoses and IVR were performed by the interventional radiologist. Various materials are used for TAE, broadly categorised as permanent (coils or n-butyl-2-cyanoacrylate) or temporary (gelatin particles). The method of following up paediatric patients who underwent TAE was as follows: (a) an emergency physician confirmed that the patients are in a haemodynamic stability state and removed the arterial sheath after 6 or 12 hours of TAE; (b) before and after TAE, the patients, with a pressure band, rested on the bed for 6 hours; (c) an emergency physician evaluated whether the patients have complications, such as puncture site hematomas or arterial embolism; and (d) an emergency physician also evaluated whether the patients had complications, such as pseudoaneurysm or arteriovenous fistulas on injured organs, using ultrasound or CT scan approximately 10 days after injury.
Data collection and outcome measurements
We collected the following information: age (years), body weight, mechanism of injury, transportation method, vital signs and haemodynamic state on hospital arrival, Injury Severity Score (ISS) [16], Revised Trauma Score (RTS) [17], predicted survival rate (%) calculated by using the trauma and injury severity score (TRISS) [18], provision of an urgent examination and treatment during the acute care phase, duration of mechanical ventilation (days), intensive care unit (ICU) stay (days), and hospital stay (days), in-hospital mortality rate (%), standardised mortality ratio (SMR), TAE data including treatment before IVR, time interval from arrival to the beginning of IVR, embolic agents, target region and artery of embolisation, degree of haemorrhage control and complication rate. The SMR was calculated by dividing the in-hospital mortality rate by the mean predicted mortality rate.
The outcome measures for TAE were the success of haemorrhage control and complications following embolisation, the need for surgical intervention or repeat embolisation, and SMR. The degree of haemorrhage control was classified as follows: (a) effective haemorrhage control; (b) ongoing haemorrhage; and (c) exsanguination and death [8]. Complications were classified as major and minor using the Society of Interventional Radiology classification system [19].
Data analysis
The patients enrolled in this study were categorised into a ‘paediatric patient group’ (younger than 15 years) and an ‘adult patient group’ (older than 15 years). The age threshold of 15 years set by Basis Education Law, at which Japanese children must receive compulsory education was chosen to classify patients as either children or adults. The results of these comparisons are expressed as the medians and interquartile ranges (IQRs) [25th–75th percentile] for continuous variables and as the means and percentages for categorical variables. The Mann–Whitney U test and Kruskal–Wallis tests were used to analyse the continuous variables, whereas Fisher’s exact test was used for the categorical variables. All statistical analyses were carried out by using STATA/SE software, version 16.0 (StataCorp; College Station, Texas, USA). A two-tailed P-value of <0.05 indicated statistical significance.