INTRODUCTION

Trauma is the leading cause of death among persons aged 1 to 44 years worldwide.¹ According to data from the World Health Organization, 5.8 million people die from this cause annually.² The National Safety Council estimated that, in 2007, the economic burden of trauma in the United States³ was approximately USD 684 billion, not including the estimated cost of lost productivity due to sequela. In Brazil, the DATASUS registered around 145,842 deaths due to external causes in 2011.⁴ Thus, the magnitude of the problem and the need to provide effective and safe care for trauma patients are clear.

Thoracic injuries are often involved in trauma-related deaths. However, the majority of trauma patients have no serious injuries and can be managed using simple measures.^5,6 To this end, much attention has been given to the type of complementary examination to be ordered.^7,8 Currently, most services order routine chest radiography for blunt trauma patients as a supplementary scan to the physical examination, based on protocols, such as that of the Advanced Life Trauma Support (ATLS) course.^9,10

In the past decade, a series of studies have focused on determining the best radiologic examination to be ordered in chest trauma.¹¹ Chest computed tomography (CT) has been proven to be superior for diagnosis, although therapeutic intervention was needed in approximately one-third of the diagnosed injuries.¹² Other studies have also shown some benefits of chest CT over simple radiography in severe trauma, but the same correlation was not found for minor trauma.^13,14

Concerns exist over the indiscriminate use of complementary examinations in trauma patients. This practice increases costs, exposure to radiation, and emergency department time. The number of negative examinations is directly associated with poor optimizing of resources. There is also a growing concern regarding the possible collateral effects of indiscriminate use of CT due to exposure to radiation, especially in the pediatric population.^15-18 Thus, better ways of selecting patients to be submitted to complementary examinations are being sought.

In recent years, some authors have questioned the need for requesting chest X-rays (CXRs) in minor chest trauma. Prospective studies have been conducted showing favorable results.^19-22 Optimizing the indication for CXRs can have a number of benefits. Conversely, the failure to identify severe chest injuries has clinically unfavorable consequences. There are arguments advocating systematic use of CXRs, given the examination is low cost and delivers a low dose of radiation. Nevertheless, clinical experience shows that most examinations performed are normal and do little to inform decisions on conduct.

The objective of the present study was to identify, using clinical criteria, a subgroup of blunt trauma patients that do not require CXR for assessment.

MATERIALS AND METHODS

This project was approved by the Research Ethics Committee of the hospital. We used the trauma registry database, collected prospectively for quality control between 2008 and 2010, which included all blunt trauma patients aged >13 years.

The following clinical variables on admission were analyzed: Age, trauma mechanism, physical examination of chest, Glasgow coma scale (GCS), and reliability of physical examination (level of preserved consciousness, no use of analgesics or drugs that change sensory level, absence of distracting injuries, and spinal cord trauma). The severity of injuries was stratified using the Abbreviated Injury Scale (AIS-90).²³ Hemodynamic stability (HS) was defined as concomitant systolic arterial pressure >100 mm Hg and heart rate <100 bpm.

Chest X-rays were performed in front view, after the initial assessment. Both the chief resident and surgery staff on call reviewed the image to state the final result (normal or abnormal). Findings not related to the traumatic event, as well as those due to past traumas, were not considered for this decision.

We selected five clinical criteria that could be easily retrieved from the usual trauma assessment in order to estimate the patient’s chance of sustaining chest injuries: Normal neurologic examination on admission (NNEx), HS, normal physical examination of the chest on admission (NCEx), age ≤ 60 years, and absence of distracting injuries (AIS > 2 in head, abdomen, and extremities).

These clinical criteria were progressively merged to select a group with lowest risk of exhibiting abnormal CXR on admission and absence of internal chest injuries, with AIS > 1.

RESULTS

A total of 5,536 blunt trauma patients were analyzed. Of these patients, 302 (5.5%) had chest injuries (AIS > 1) (Table 1). Of the 4,647 cases submitted to CXRs on admission, 268 (5.7%) had abnormal findings on examination.

Among the 2,897 trauma patients scoring 15 on the GCS at admission and testing normal on NNEx, 116 (4.0%) had abnormalities on CXRs.

Of the 2,426 patients with NNEx and HS, 74 (3.0%) had positive CXRs. Among the 1,698 trauma patients that, besides the previous variables (NNEx and HS), also had no abnormalities on NCEx, only 24 (1.4%) had abnormalities on CXR. From this group, a subgroup of individuals aged ≤ 60 years was derived. Of the resultant 1,347 patients, 19 (1.4%) had abnormal CXR. Out of 19 patients, 12 had confirmed thoracic injuries with AIS > 1 (0.9% of all individuals that received CXR). Injuries were rib fracture (9), pneumothorax (3), and hemothorax (1). A total of 4 patients needed chest drainage.

Merging previous variables (NNEx, HS, NCEx, and age < 60) with absence of distracting injuries (AIS > 2 in head, abdomen, and extremities) yielded 1,140 cases. In this group, there were 9 (0.8%) abnormal findings on CXR and 8 patients with confirmed thoracic injuries (Table 2): 2 pneumothorax and 6 rib fractures. Only 2 (0.2%) cases required chest drainage.

DISCUSSION

Most studies confirm the low incidence of thoracic injuries in blunt trauma patients in general.^19-22 In the present study, of the 4,647 cases that received CXR on admission, only 5.7% exhibited abnormalities on CXR. It is important to comment that the small percentage of chest injuries in our scenario may be related to the sample of the study. Due to the characteristics of local health system, our hospital receives many patients without serious injuries. So, optimizing the indication of complementary examinations is a key point in our reality.

The need to perform routine chest radiography in trauma, established in some trauma care protocols, such as the ATLS, has been questioned due to the high number of normal examinations and the costs associated with this practice in the emergency service.^15,24,25 Sears et al¹⁹ stated that, using the team’s clinical judgment alone, around 50% of CXRs may not have been performed, saving approximately USD 100,000 in 12 months. In the present study, if we had opted not to perform the X-rays in the group of patients meeting all exclusion variables, 1,140 examinations would have been saved (47 examinations/month). The high rate of patients with blunt trauma for whom CRX was ordered and no injury disclosed reveals the need for reassessing the prevailing indication criteria.

Myint et al²¹ showed that the group of blunt chest trauma patients assessed as stable required no routine CXR, a result corroborated by the findings of the present study. Also a prospective study involving 386 cases, conducted by Nejati et al,²⁰ compared the data collected on admission from patients (trauma mechanism, vital signs, oxygen saturation, auscultation, pain on palpation) with radiographic findings, showing that the combination of tachypnea and pain on palpation can identify significant chest trauma with 100% sensitivity. Our data revealed the association between normal findings on physical examination of the chest and absence of internal injuries.

Recently, the Nexus Chest study proposed a clinical decision instrument for determining the need for chest imaging, based on the absence of the following criteria: Chest pain, age < 14 years and >65 years, intoxication, rapid deceleration mechanism, distracting injury, change in level of consciousness, and pain on chest wall palpation. This model allows the exclusion of thoracic injuries with 98.8% sensitivity, 98.5% negative predictive value, and 13.3% specificity. Of the 17 false negatives found, only 1 case required intervention (1 in 1,478 patients with injury seen on chest imaging).²² In the present casuistic, when all variables proposed were present, including GCS = 15, NNEx, HS, NCEx, age < 60 years, and absence of distracting injuries (AIS > 2 head, abdomen, and extremities), there were 9 abnormal findings on CXR out of 1,140 cases (0.9%). Of these 1,140 cases, only 2 required chest drainage (0.2%).

Therefore, it proved possible to establish a subgroup of patients with blunt chest trauma and low risk of thoracic injury. The need for performing CXR in this group should be reassessed.

CONCLUSION

A subgroup of blunt trauma patients with low probability of exhibiting abnormalities on CXR at admission was identified. The need for routine CXR in this subgroup should be reviewed.