Isolated LOC in head trauma associated with significant injury on brain CT scan

Background A report of loss of consciousness (LOC) is frequently considered reason enough to obtain a computed tomography (CT) scan in the evaluation of head trauma. We conducted this study to reduce exposure to radiation from CT, while still not overlooking clinically significant injuries. Objective The objective of the study is to determine the correlation between LOC status and brain CT scan results in patients with blunt head trauma and to determine whether there is a subset of patients for whom CT scan need not be performed, without missing clinically significant intracranial injuries. Methods This is a retrospective study conducted in the emergency department of an inner-city hospital. The patient population included patients ranging between 13 and 35 years of age, with blunt head trauma, who presented to the emergency department (ED) between January 2010 and December 2013. Patients were divided into two groups: “LOC” group and “no LOC” group. The results of brain CT scans from each group were compared with LOC status. For study purposes, “clinically significant” were those that required interventions or ICU hospitalization of at least 24 h or extended hospitalization. The results were analyzed using chi-square calculations. Results During the study period, 494 patients were identified as having suffered head trauma. Of these, 185 (37.5%) reported LOC and 309 (62.5%) did not lose consciousness. In the LOC group, 15 (8.1%) had significant CT findings compared to 1.3% (4/309) of those without LOC (p < .001). Of the 4 who had no LOC and had significant brain CT findings, all 4 patients had positive physical findings of head, neck, or facial trauma. In the LOC group, only 1/15 (6.7%) had significant CT findings with a normal GCS of 15 and no physical signs of the head, neck, or facial trauma. Conclusions A small proportion of patients with LOC had CT finding requiring intervention. Head trauma patients with no physical injuries to the head, neck, or face and a normal GCS had no significant brain CT findings. This raises the question of whether a routine brain CT scan should be obtained in patients with LOC, no physical findings, and a normal GCS after blunt head trauma.


Background
Head trauma is a common reason for emergency department visits. Many patients who suffer blunt head trauma are at risk of traumatic brain injury (TBI). This affects approximately two million people in the USA each year [1]. In addition, there are 1.4 million annual hospitalizations for TBI [2]. A report of loss of consciousness is frequently considered an indication for obtaining a brain CT scan in the evaluation of patients with blunt head trauma [3][4][5]. In the USA, physicians routinely obtain CT scans for patients with abnormal Glasgow Coma Scale (GCS) scores and/or LOC. This study evaluated whether a report of LOC in a patient with blunt trauma, but an otherwise normal physical examination, warrants obtaining a brain CT scan.

Methods
This is a retrospective study conducted in the ED of an urban hospital, which is a state-designated level I trauma center in New York City. It has a very busy emergency department with annual visits of approximately 120,000. This hospital has both emergency medicine and surgery residency training programs. The study population is comprised of individuals primarily from the South Bronx, between 13 and 35 years of age, with blunt head trauma, who presented to the ED between January 2010 and December 2013. CT scans were obtained at the discretion of the treating physicians.

Inclusion criteria
For study purposes, we included only previously healthy patients between the ages of 13 and 35 years of age, who presented following blunt head trauma.

Exclusion criteria
Patients with other medical conditions that may influence CT result, such as stroke, were excluded. Similarly, patients more than 35 years of age were excluded. Patients who were at higher risk for intracranial bleeding and those taking anticoagulant medications were also excluded. Patients who were suspected of substance abuse or intoxication were not included in this study.

LOC definition
For study purposes, LOC was considered to be present if there was any reported period of unconsciousness following the traumatic event documented in the patient's medical record.
Based on the LOC status, patients were divided into two groups: "LOC" group and "no LOC" group. The results of brain CT scans from each group were then compared with LOC status. In addition, we looked at GCS status of the patients exhibiting clinically important CT findings.

Physical findings
For study purposes, physical findings were defined as injuries to the head, neck, and face or the presence of alterations in mental status.

Outcome measure
"Clinically important" findings on CT scan either required surgical intervention or greater than 24-h ICU admission or increased length of hospitalization. These included intracranial hemorrhage (including subdural, epidural, or intra-parenchymal types), skull fracture. The presence of a chronic intracranial lesion or findings that did not require acute intervention or increased length of hospitalization were not considered clinically important in the acute management of trauma.
In addition, we collected the following data: patient age, presentation information, initial GCS score on arrival, mechanism of injury, and disposition status. All data collection, including a chart review of the patients in this study, were performed by the primary authors. All CT scans were interpreted by board certified attending radiologists. The SPSS 22.00 (IBM Corporation, New York) was used to analyze the data. A bivariate analysis was performed, and Pearson chi-square calculations was used.

Statement of declarations, consent, or ethics approval
This study received approval from the Institutional Review Board of the Lincoln Medical and Mental Health Center (IRB reference no. 13-023). Written informed consent was waived in retrospective studies of anonymized patient data under Ethical Guidelines for Medical and Health Research Involving Human Subjects.

Non-LOC group
Of the 309 patients who did not report LOC, 225 (72.8%) were male and 84 (27.2%) were female. In this group, significant CT findings were identified in only four patients (1.3% = 4/309). CT was negative in 305  (Table 3).
Based on GCS, patients were further classified into minimal or normal (GCS 15), mild TBI (GCS 13-14), moderate TBI (GCS 9-12), and severe TBI (GCS 8 or below). We then compared GCS score with the CT scan interpretation. The results were as follows: minimal or normal 461 (93.32%), mild TBI 17 (3.44%), moderate TBI 4 (0.81%), and severe TBI 12 (2.43%). In the minimal or normal group, CT scans were normal in 452 (98%) patients, whereas 9 (2%) showed clinically important CT findings. In the moderate TBI group, 2 CT scan (50%) were normal and 2 (50%) were positive. In the severe TBI group, 6 (50%) CT scan were negative and 6 (50%) showed positive findings; P < .001 (Table 4). Table 5 lists the characteristics of patients in this study with positive brain CT scans. There were three trauma causes or mechanisms identified in patients with positive CT scans: assault, motor vehicle crash (MVC), and falls. Of the 19 patients with significant CT findings (CT+), six were assaulted; two had falls subsequent to the assault; and eight were injured in motor vehicle crashes (MVC). The primary mechanism was a fall in five patients.
We categorized the patients into pediatric and adult groups. There were 57 (11.5%) pediatric patients under the age of 19 years and 437 (88.5%) adults. In adults, 164

Discussion
Loss of consciousness is common in patients with blunt head trauma and is an important factor influencing the decision to order a CT in the ED. In a large cohort study of mild TBI patients, CT scan was obtained in 91% [6].   Many physicians have questioned the relationship between a history of LOC and TBI [7,8]. Despite this, the practice of obtaining CT scans in patients with LOC has remained unchanged, and CT has been a mainstay of diagnostic testing in the ED. A brain CT scan is the test most commonly ordered in the evaluation of patients with head trauma [9,10]. Each year, more than 1 million head CTs are obtained for head trauma [11,12]. A report has revealed that the rate of head CT use has increased 35.7% (range = 34.5 to 46.8%) for all ages between the years 2006 and 2011 [13].
This study revisited the question of whether a brain CT scan is always required in patients with blunt head trauma who report LOC on arrival in the ED. Using only clinical evaluation certainly reduces CT scan use, but may not eliminate the diagnosis of intracranial injury. For many physicians, the idea of missing the diagnosis of any intracranial injury is unacceptable. It is possible that the concern of litigation may play a role in medical decision-making, and some would not accept missing the diagnosis of any intracranial injury regardless of its significance. A New Orleans Criteria study has suggested that CT use could be avoided in patients with minor head injury in the absence of any of the following: headache, emesis, age > 60 years, drug or alcohol intoxication, posttraumatic seizure, physical evidence of trauma above the clavicle, and short-term memory deficits [14]. These "New Orleans Criteria" were created to capture all intracranial injuries, including those which did not require neurosurgical intervention. To improve specifically detecting clinically important brain injury, the Canadian CT Head Rule allowed a patient with GCS 13-15 within 2 h of the injury to forego a head CT, but included other criteria such as age < 16 and mechanism of injury that would automatically require a brain CT scan. To identify patients with any clinically important brain injury, the Canadian CT Head Rule also included amnesia as one of the criteria [15]. In our study, LOC was reported in 37.5% of patients who presented with blunt head trauma. Fifteen of these patients (8.1%) had clinically important findings on CT scan. If no physical findings were found in the head, neck, or face, and there was no alteration of consciousness, only one patient had a significant CT scan finding.
In this sample, the majority of patients presenting following blunt head trauma were male (73.1%) and 18 out of 19 (94.7%) of the CT positive patients were also in male patients. This correlates well with the general trend for males being at a higher risk for involvement in MVC and trauma. In order to minimize head injuries, perhaps primary prevention strategies should be directed towards the male population.
We also reviewed the association of GCS and CT scan results. In this study, 7 patients (21.2%) with GCS < 15 had a positive CT scan GCS (n = 33). In a previous study, patients with GCS of 13 had a five times greater risk of having intracranial injury [16]. Another study suggested receiver operating characteristic areas greater than 0.65 with LOC if the GCS score was less than 15 (0.692) [17].
This study had several potential limitations. First, the retrospective design of this study meant that determination of LOC status was dependent on the documentation in the medical record. If it was not recorded, it was assumed that there was no LOC. Second, the study was underpowered to detect differences in physical findings. We acknowledge that sometimes, an accurate report of LOC may be difficult to ascertain because of the lack of a witness to the traumatic event. In addition, there may have been difficulties in determining whether the patient had actually been unconscious. We also do not know the reliability of the LOC reported by the patient or witness in the field. It is possible that the patients who reported no LOC could have thought they were lucid during the trauma, but may have briefly lost consciousness and could not remember the entire event.

Conclusions
In this limited sample, there was only one patient with a significant CT scan finding with isolated LOC, GCS of 15, and otherwise no physical injuries. The practice of routinely ordering brain CT scans should be reconsidered. Until more definite evidence becomes available however, judgment should be used when evaluating patients with blunt head trauma with associated LOC, when deciding whether to obtain a brain CT scan.