Management of isolated sternal fractures using a practical algorithm.

BACKGROUND: The implementation of seat belt legislation has led to an increase in the frequency of isolated sternal fractures (ISFs) in motor vehicle crash. AIMS: We reviewed retrospectively the medical records of our tertiary center in order to find out the frequency of ISFs, review our experience in their management, and define the mean length of hospitalization.
MATERIALS AND METHODS: From January 2008 to April 2012, 64 patients were admitted to the accident and emergency department of the University Hospital of Larissa, Greece, suffering from sternal fractures (SFs). Of these 64 patients, 45 had sustained ISF, while the remaining 19 had SF and additional injuries (intrathoracic and extrathoracic). The files of these 45 patients were further investigating as concerning the mechanism of injury, hospitalization days, morbidity, and mortality.
RESULTS: All the patients had been involved in motor vehicle crashes and most of them were wearing seat belts during the accident (91%). The hospital length of stay (LOS) was 1.85 ± 1.67. All the patients had upon admission chest radiograms, serial electrocardiographs (ECGs), echocardiograms, and cardiac enzyme levels. Two patients had abnormal ECG and abnormal cardiac enzymes which contributed in prolonged hospitalization. However, there was no incidence of cardiac complications or deaths.
CONCLUSIONS: ISFs, with normal electrocardiogram, cardiac enzymes, and chest X-ray in the absence of complications, require no further investigation.

INTRODUCTION

Sternal fractures (SFs) are invariably the result of a considerable force applied directly to the anterior chest wall and their frequency has increased since wearing car seat belts has become compulsory. SFs, as the principles of Advanced Trauma Life Support (ATLS) dictate, can be associated potentially with an increased risk of cardiac and great vessel injury. But not all SFs are the same. According to current literature, isolated sternal fractures (ISFs) can be regarded as benign injuries and treated respectively.[12345678] ISF has been defined as a SF without coexistent thoracic injuries such as hemothorax, pneumothorax, rib fractures, cardiac injury, etc.[5] The introduction of the seat belt law has led to an increase in the frequency of ISF in road traffic accidents due to the concentrated pressure of the belt against the rib cage at the time of injury.[9] Furthermore, there seems to be also a change in the pattern of injury. The evolving mechanism of injury in restrained occupants in vehicles is postulated to be flexion of the sternum across a fulcrum (diagonal seat belt), in contrast to direct impact of the sternum on the steering wheel or other interior parts of the car.[9] It is agreed that the seat belt and the sternum absorb a substantial part of the energy transfer that occurs at the time of collision, thereby providing protection from severe injuries.[34]

The incidence of fractured sternum in motor vehicle crash (MVC) victims attending hospitals has been estimated at 0.45-8%.[410] It is possible that this range reflects the frequency of seat belts use.[9] SFs classically have been associated with major intrathoracic complications such as myocardial contusion, lung, and great vessel or spinal cord injury accounting for a mortality rate ranged from 0.7 to 19.2%.[411] The purpose of our study was to determine whether patients with ISF need admission and extensive work-up or can be discharged home after a short period of observation in the emergency department. Thus, we scrutinized the archives of our department over a 5-year period, in order to identify all patients who sustained an ISF, monitor the hospitalization days and analyze the morbidity and mortality of ISF. We also propose an algorithm for the better assessment and management of ISFs.

MATERIALS AND METHODS

During a 5-year period, from January 2008 to April 2012, 64 patients diagnosed with SFs. The data were gathered from the accident and emergency department of the University Hospital of Larissa, Greece. Approval from the hospital's ethics committee was obtained for the conductance of the study. Of these 64 patients, 45 had ISF and 19 had SF with additional injuries (intra- and extrathoracic). These 19 patients were therefore excluded from the study. Demographic data and clinical findings including the primary complaint, mechanism of injury and length of hospitalization were collected. We also analyzed the results of subsequent investigations obtained during the patient stay, including chest radiograms, serial electrocardiographs (ECGs), echocardiograms, cardiac enzyme levels, and computer tomography (CT) scans. The hospital's protocol required all patients with a radiological diagnosis of SF to be admitted for observation for at least 24 h, serial ECGs, cardiac enzyme measurements such as troponin I (TropI), and creatine kinase-MB (CPK-MB) and cardiac echocardiography examination. We considered the following findings as evidence of potential blunt cardiac injury or possible myocardial contusion:

CPK-MB isoenzyme concentration of > 5% of total CPK enzyme level or abnormal TropI levels;

ECG showing evidence of acute injury pattern or new-onset dysrhythmia requiring therapy; and

Echocardiogram showing pericardial fluid/effusion.

For the comparison of seat belt usage between patients with ISF and patients with SFs and additional injuries a Yates correction for continuity test was used.

RESULTS

The data of the 45 patients are presented in [Table 1]. All the patients had been involved in a MVC and the majority of them was wearing seat belts (41/45, 91%). On the other hand, four patients out of the 19 who sustained additional injuries were wearing seat belts (21%). Statistical analysis between these two groups has showed that seat belt usage was found to be statistically significant in favor for the first group [Table 2]. Prevalent symptoms and signs included sternal or chest pain in all patients. Chest-X rays (posterior/anterior view and lateral), 12-lead electrocardiography, cardiac enzymes measurement, and echocardiography were performed in all patients upon admission, according to our protocol for SF. The sternal radiography (lateral view) was diagnostic for SF in all patients. Further investigation with computed tomography (CT) was done according to clinical indications.

Table 1

Demographic, clinical characteristics, and prognosis of 45 patients with isolated sternal fractures

Table 2

Comparison of seat belt usage between patients with ISF and patients with SF and additional injuries

Two patients had abnormal ECG and abnormal cardiac enzymes and in one of them the ECG revealed signs of previous NSTEMI, without associated rise in cardiac enzymes. The echocardiography examination showed dyskinesia of the inferior wall which was attributed to the old myocardial infarction. There was no episode of arrhythmia or cardiac instability during the 3 days of hospitalization. As far as the other patient is concerned, ECG showed nonspecific changes (slight ST elevations) with a minor rise in CPK-MB and TropI levels. However, the echocardiography was normal and the patient did not develop new ECG abnormalities or further rise in cardiac enzymes during his admission. None of these patients met criteria for a diagnosis of myocardial contusion. However, these abnormalities were the cause of prolonged hospitalization. Forty patients (89%) were discharged home within 48 h of admission. Five patients had a history of comorbidities (one chronic obstructive pulmonary disease and four ischemic heart disease) and due to this history they had a relative prolonged hospitalization (3-4 days), although none of them developed any new episode of cardiac complication during hospitalization. There was no death between these 45 patients.

DISCUSSION

It has been shown that SFs can lead to a wide range of injuries.[456] This has led to the wide spread policy of admitting all patients with SF for cardiac monitoring, serial ECGs, and further work-up as needed. However, ISFs may be viewed differently. Many studies have shown that significant factors in the prognosis after SF are not only the mechanism of injury, but also the presence of coexisting injuries.[67111213] In case of ISF, blunt cardiac injury seems to be a rare complication especially in patients wearing seat belt where little force is transmitted on the heart.[114] On the contrary, direct impact with the steering wheel has a result a greater force applied to the heart as it is crushed between the sternum and the thoracic spine, and it is more likely to result in concomitant injuries. This latter was revealed in our study as well by the comparison of seat belt usage between the two groups [Table 2]. Of course this is drawn arbitrarily because there are a lot of other factors contributing to an injury for example the mechanism of injury per se, the speed of collision, etc. In our study 9% of the patients with ISF was not wearing seat belt. However, we found no cardiac or vascular complications. The incidence of myocardial contusion in SF patients has been reported to have a wide range. In one study, based only in an abnormal ECG, it was found to be 62%.[15] While another study which was based on the increase of cardiac enzyme levels, ECG changes and abnormal echocardiography results or autopsy findings, showed that it was only 1.3%.[16] This wide discrepancy can be attributed to the vague definition for myocardial contusion. The literature contains also equivocal evidence of the usefulness of serial ECG, cardiac enzyme measurements, and echocardiography in detection of blunt cardiac injury. Heyes and Vincet noted that ECG, though not specific, is sensitive in identifying potential problems and that CPK-MB measurements are noncontributory.[17] Similarly, Hills et al., stated that myocardial contusion should be routinely investigated by serial ECG, while CPK-MB measurements and echocardiography may contribute not much.[18] On the other hand, Fabian et al., concluded that the CPK-MB level is the most reliable indicator of myocardial injury.[10] Velmahos et al., stated in patients with blunt thoracic trauma, including patients with SF, the combination of normal ECG and TropI at admission and 8 h later rules out the diagnosis of clinically significant blunt cardiac injury.[19] Similarly, Salim et al., proposed that patients with a normal admission ECG and TropI can be safely discharged in the absence of other injuries and that the usefulness of echocardiography is limited if the test is used to confirm the presence rather than predict the development of cardiac complications.[13]

The authors would also like to propose a potential algorithm that could be of significant assistance for the better assessment and management of ISFs [Figure 1]. Our primary goal is to reduce the number of ISF admissions and the length of hospitalization, but not overlooked patients with possible severe accompanying injuries. According to our algorithm, a patient who has sustained a SF with no concomitant injuries, can be safely discharged on oral analgesics after 6-12 h of observation if he meets the following criteria:

Figure 1

A patient with an isolated fracture of the sternum, who has no concomitant injuries, no significant past medical history; with the ECG, chest X-ray and echo being normal and without a cardiac enzyme elevation can be discharged home on regular analgesics after a short period of observation. Algorithm for management of sternal fractures

Normal ECG, echo, and chest X-ray;

cardiac enzymes within normal range; and

No significant past medical history and being otherwise fit.

Pain management can usually be achieved with codeine and nonsteroidal anti-inflammatory drugs. The authors believe that exceptions should be made for patients with advanced age or comorbidities such as preexisting history of heart disease since there are reports that these patients are at higher risk to develop cardiac arrhythmias requiring therapy and therefore a short stay admission is advised.[419] Severely displaced fractures, inadequate pain control, or even insufficient domestic support may also justify hospitalization.[20] On the other hand, a patient with either ECG changes or elevation of cardiac enzyme should be admitted for echocardiography and observation. We believe that this algorithm could assist clinicians in an effort to make an evidence-based judgment in regards to the need for admission.

Our study shows that ISF can be managed as a benign injury and does not justify an inpatient management. Nevertheless, it has some limitations. Firstly, it is based on a relative small sample of patients (45 patients). Probably a larger study would offer more rigorous evidence to sustain our findings. Additionally, we did not endorse follow-up of the discharged patients more than 1 month. A longer follow-up period might have provided us with more comparable data.

CONCLUSION

Conclusively, patients with ISF can be safely discharged after observation in a short stay unit. The presence of additional injuries or history of cardiac disease must be also highlighted and it should act as an alerting factor to the medical staff. These patients should be admitted for close monitoring. Application of these recommendations can be safe and could obviate extensive work-up, especially in the contemporary era of increased financial hospitalization costs and austerity.