The use of chest computed tomography versus chest X-ray in patients with major blunt trauma

Introduction 

Two-thirds of patients with multiple injuries suffer from blunt chest trauma and severe thoracic trauma is associated with multiple injuries in 70–90% of cases.5, 7 Among blunt injuries to the chest, lung contusion is considered one of the most important factors contributing to the increased morbidity and mortality of patients with multiple injuries.2, 9

The usual diagnostic work-up in the emergency department for blunt injuries to the chest includes a routine chest X-ray taken in the supine position and an ultrasound. Despite this approach, significant injuries, such as pneumothoraces, haemothoraces, and lung contusions can be missed during the initial trauma assessment.1, 8, 11 Another investigation that is relevant to assess blunt trauma to the chest is computed tomography (CT) scanning. Several studies have shown that CT scanning is accurate in visualising intrathoracic injuries, such as pneumothoraces, haemothoraces, and lung contusions.3, 6, 10 In addition, the availability, reliability, and low complication rate of CT scans has led to its widespread use in the evaluation of blunt trauma.

A number of authors have suggested that the CT chest scan should be routinely considered in the initial assessment of chest trauma.6, 12 However, this suggestion remains controversial. For example, some studies have reported clinical changes in management after CT scans in up to 70% of cases,10 whereas others have suggested that routine CT scans do not have a major impact on the management of blunt trauma to the chest.8, 6

With the wide availability of CT scanners and with the technical improvements in image quality and speed over the past decade, overuse and perhaps overdependence on CT results for the management of patients with chest trauma has occurred. Although CT is an excellent diagnostic tool for chest trauma, it is costly, requires radiation exposure, and removes the patient briefly from direct clinical care at a time when close monitoring of the patient is critical. In a busy trauma or emergency facility, overuse of CT scans can lead to inappropriate delays in patient care.

The purpose of the study is to identify the clinical features associated with further diagnostic information obtained on a CT chest scan compared with a routine chest X-ray in patients sustaining blunt trauma to the chest. This will help to guide decisions about the further investigation and management of blunt trauma to the chest; which is important given that two-thirds of patients with multiple injuries sustain blunt chest trauma.5, 7

Methods 

Results 

Of the 141 patients, 75% (n=106) were men with the patients ranging in age from 17 to 89 years (mean=47.2 years). The most common mechanism of blunt trauma to the chest was as a result of a motor vehicle crash (37%, n=52), a motorcycle crash (8.5%, n=12), a fall (23%, n=33) or pedestrian injury (17%, n=24). The mean ISS was 24 (range 3–59).

Almost a third of the patients (n=45) arrived intubated or were intubated in the emergency department. Eighteen percent (n=26) of the patients had a chest tube placed prior to CT scan and an additional 11% (n=15) of the chest tubes were placed after obtaining a CT chest scan. No patient required needle thoracocentesis. Forty-two percent (n=59) of the patients underwent an operation including two (1.4%) thoracotomies, 12 (8.5%) laparotomies, 24 (17%) open reductions and internal fixations and 11 (7.8%) craniotomies. Sixty percent of the patients (n=85) needed admission to ICU, where they were intubated for an average of 3 days and stayed for an average of 5 days. The overall mortality rate of the study participants was 12.8% (n=18).

Ninety percent of patients (n=127) had at least one pathological finding on CT scan compared to 58% of patients (n=82) on plain chest X-ray. Additional investigations and/or interventions following CT scan were performed on 19% of patients (n=27). These included aortic angiography (n=4), transoesophageal echocardiography (n=4), and two patients required insertion of an endoluminal stent for aortic injuries. Thoracotomies were performed on two patients; one for a haemothorax and the other a transsected thoracic aorta. Intercostal catheters were required in 15 patients following the CT scan for a pneumothorax that was not seen on a plain chest X-ray or for an additional chest tube for a persisting pneumothorax. Ten percent of patients (n=14) had no abnormality detected on CT scan. Half of those scans involved intubated patients.

A widened mediastinum was suggested in 14 patients following chest X-ray. A haematoma was confirmed on CT scan in six of these; four prevertebral haematomas with vertebral fractures present and two anterior haematomas in the presence of a sternal fracture. In the eight remaining cases, CT scanning could not confirm the suspicion of a widened mediastinum. CT scan picked up an additional 21 mediastinal haematomas, 3 of which were aortic injuries not seen initially on plain X-ray. There were three elevated haemidiaphragms on plain X-ray raising suspicion of diaphragmatic injury. Two were confirmed on CT scan and were repaired at subsequent laparotomy. The third case had an elevation on CT scan as well, but was not confirmed clinically. CT scan was significantly better than routine chest X-ray in detecting lung contusion, pneumothorax, mediastinal haematoma, as well as rib fracture, fractured scapula's, fractured sternum and fractured vertebra (see Table 1).

Table 1. Number (%) of positive radiological findings

	Radiological findings	Chest X-ray	CT scan	p-Value
	Rib fractures	47 (33.3%)	68 (48.2%)	<0.001
	Scapula fractures	5 (3.5%)	12 (8.5%)	0.016
	Vertebra fractures	1 (0.7%)	23 (16.3%)	<0.001
	Sternum fractures	0 (0%)	10 (7.1%)	<0.001
	Clavicular fractures	13 (9.2%)	13 (9.2%)	1.00
	Pneumothorax	9 (6.4%)	31 (22.0%)	<0.001
	Haemothorax	10 (7.1%)	16 (11.3%)	0.263
	Haemopneumothorax	1 (0.7%)	16 (11.3%)	<0.001
	Lung contusion	23 (16.3%)	44 (31.2%)	<0.001
	Mediastinal haematoma	10 (7.1%)	25 (17.7%)	0.006
	Aorta dissection	0 (0%)	3 (2.1%)	0.250
	Ruptured diaphragm	2 (1.4%)	2 (1.4%)	1.00

Fifty-nine patients (42%) had additional findings reported by CT scan beyond that found on their chest X-ray, and in a multivariate logistic regression model that included all cases (n=141) and was adjusted for age and gender, a number of clinical features were found to discriminate between findings on a CT chest scan compared to a chest X-ray. These findings were the presence of chest wall tenderness (OR=6.73, 95% CI=2.56, 17.70, p<0.001), reduced air-entry (OR=4.48, 95% CI=1.33, 15.02, p=0.015) and abnormal respiratory effort (OR=4.05, 95% CI=1.28, 12.66, p=0.017).

All patients with a reduced air-entry (n=33) in their clinical assessment had an abnormal CT scan. Furthermore, only 5% (3 of 65) patients with chest wall tenderness and 6% (2 of 35) patients with increased respiratory effort had a normal CT chest scan. Twenty-two percent of patients did not have any of these signs but had an abnormal CT scan.

Discussion 

In this study, CT chest scanning was significantly more effective in detecting pneumothoraces and haemopneumothoraces, lung contusions and mediastinal haematomas compared with a chest X-ray. This is in accordance with several studies that have shown a greater sensitivity for a CT chest scan for detecting intrathoracic injuries.6, 12 Furthermore CT chest scan was significantly better at detecting fractured ribs, scapulas, sternums and vertebrae than a chest X-ray.

CT scanning was more sensitive in the detection of parenchymal abnormalities, such as lung contusions. The majority of patients who require intubation for respiratory failure are usually intubated on clinical grounds and often prior to obtaining a CT scan. Although CT scans may confirm the diagnosis of pulmonary contusion, therapy is based on monitoring physiologic variables, as radiographic findings have not been found to correlate with either mortality or the need for intubation.2

The relatively low rate of thoracotomy in this study group might be correlated to a relatively low ISS and because of the exclusion of patients with penetrating trauma and supports again the fact that most thoracic blunt trauma can be treated conservatively.4

The question of whether the additional information gained with CT scanning changes patient management in this early phase remains controversial. Several studies found changes in patient management of up to 70% in situations where CT chest scans were performed, such as insertion or correction of a chest tube, a change in the mode of ventilation, as well as further investigations and/or interventions.10 In contrast, other studies have not been able to demonstrate significant therapeutic consequences and consequently, they do not recommend CT chest scans as a routine first assessment.8, 6 In this study, only 19% (n=27) of the patients with an abnormal CT scan had a subsequent investigation and/or intervention as a direct result of the findings. Of these patients, the majority (70%, n=19) had chest wall tenderness, reduced air-entry, increased respiratory effort or a combination of these signs.

With the increasing use of CT scanning, the proportion of CT scans showing no abnormality is likely to increase. While an apparently normal CT result might still be of considerable value in the intubated patient, for the remaining patients, factors such as transport out of the resuscitation area, exposure to radiation, possible reaction to the contrast agent, delay in treatment and additional costs need consideration. Although an abnormal CT scan was found in 90% of patients in this study, the clinical management changed in a minority of these patients on the basis of the CT scan results. Of the 10% of patients with normal scans, half were intubated patients, in whom clinical findings are limited and CT scanning is warranted.

Conclusion 

The findings from this research highlight that a CT scan is significantly more likely to yield additional information than a CXR alone under the following circumstances: presence of chest wall tenderness, reduced air-entry and abnormal respiratory effort. Therefore, in alert patients and in the absence of those clinical findings, the results suggest that selective use of CT chest scanning can be considered. This is supported by the fact that most chest injuries can be treated with simple observation. As the clinical information is not obtainable from intubated patients, we suggest that this group receives a routine CT chest scan in their first assessment. Importantly, although the clinical signs we have outlined have some value, clinicians should take into account all of the clinical features and results of other radiological investigations when making a final decision about the use of CT chest scanning. This may be particularly relevant in situations where a cardiac or aortic injury is suspected as these may not present with the signs that we have identified.