Introduction

Traumatic brain injury (TBI) is a major cause of death, disability and economic cost to Irish society. One of the central concepts emerging from research into traumatic brain injury is that all neurological damage from TBI does not occur at the time of impact, but evolves over the ensuing hours and days [1]. Improved outcomes result when these secondary, delayed insults resulting in reduced cerebral perfusion to the injured brain are prevented or overcome. This is reflected in the progressive reduction in mortality from severe TBI from 50% to under 25% over the last 30 years, even when adjusted for injury severity, age and other prognostic factors [2]. The trend in reduced mortality and improved outcomes is, in part due to evidence-based protocols that emphasize monitoring and maintaining adequate cerebral perfusion [3, 4]. The most comprehensive available guidelines produced by a collaboration between the Brain Trauma Foundation, American Association of Neurological Surgeons (AANS) and Congress of Neurological Surgeons (CNS) have been widely adopted by intensive care units in Europe and the USA. The third edition of the evidence-based guidelines for the Management of Severe Traumatic Brain Injury were published in 2007, following the first and second edition in 1995 and 2000 [5, 6]. The guidelines address key topics useful for the management of severe TBI in patients with a Glasgow Coma Score of 3–8. Studies suggest that the consistent application of ICU-based protocols improves outcomes and reduces mortality and length of stay [7–9]. Our retrospective clinical audit was performed in a University teaching hospital in Ireland. Our aims were to determine the incidence, demographics and outcomes of patients with severe traumatic brain injury admitted to the intensive care unit, examine our compliance with the Brain Trauma Foundation Guidelines and identify areas for improvement.

Methods

A retrospective clinical audit was undertaken examining ICU admissions over the 3-year period, 2005–2007. Patients who were admitted to ICU due to traumatic brain injury were identified from the Intensive Care admission register. Though some patients had ancillary injuries, only patients in whom TBI was the main diagnosis were included for audit purposes. A chart review was undertaken for each admission. Demographic data, clinical information, radiological findings, treatments received, monitoring received and Glasgow Outcome Score were recorded. Treatment given was compared with best international practice, as per the Brain Trauma Foundation guidelines, and compliance with the guidelines recorded.

Results

A total of 46 patients were admitted to the intensive care unit over the 3-year study period with a primary diagnosis of traumatic brain injury; 89% were admitted directly from the hospital’s emergency department and 11% were transferred from other hospitals. Of the patients, 80% were male with a mean age of 43 years (±17). The most common aetiology was assault (37%), with road traffic accidents (33%) and falls (24%) also common. Positive blood alcohol levels were recorded in 26% of patients.

GCS at the time of hospital admission ranged from 3 to 14, with a mean of 7.93 (±4.46). Of the patients, 56% were classified as having severe traumatic brain injury (admission GCS 3–8). All patients underwent CT scanning prior to ICU admission and 98% of the patients were intubated at the time of ICU admission.

The mean length of stay in the ICU was 5.45 days (range 1–30 days). Survivors had a shorter length of stay (4.15 days) than non-survivors (7.25 days). Mean hospital length of stay was 26 days (range 2–287 days). The severe traumatic brain injury group (GCS 3–8) had a much longer hospital stay (mean = 44 days) than the moderate traumatic brain injury group (GCS 8–15), which had a mean length of stay of 10.9 days. In 93% of cases, a neurosurgical opinion was sought. Two patients were transferred to a neurosurgical centre for further treatment.

Good compliance was achieved with the guidelines for medical management as set out by the Brain Trauma Foundation (Table 1). All patients complied with guidelines for oxygenation and blood pressure control. All patients received appropriate thromboprophylaxis and 32% of the patients who fulfilled criteria for seizure prophylaxis received the same. Appropriate ventilation to normocapnia occurred in 91% of the cases. Mannitol was used in 11% of cases, though in the absence of ICP monitoring it was difficult to assess compliance with the guidelines. As much as 64% of the patients fulfilled Brain Trauma Foundation guidelines for ICP monitoring, but due to unavailability of ICP monitoring facilities no patient received ICP monitoring. All patients received appropriate nutritional support. Periprocedural antibiotics at the time of intubation are recommended, but received in only 29% of patients. The role of prophylactic hypothermia is uncertain and was not used in any of our patients.

Table 1 Compliance with medical guidelines
Full size table

A total of 15 (37%) patients fulfilled the Brain Trauma Foundation guidelines for surgical intervention. (Table 2) Indications were due to subdural haematoma or traumatic parenchymal lesions. There was no incidence of epidural haematoma or posterior fossa lesions requiring surgery. Two patients in this group were transferred to a neurosurgical centre for further treatment. The mortality rate in the group who fulfilled criteria for surgical intervention was 67%.

Table 2 Compliance with surgical guidelines
Full size table

The overall mortality was 37%. Of those who survived, 62% made a complete recovery with the remainder suffering varying degrees of neurological impairment. In the severe TBI group, mortality was significantly higher (54%) than in the moderate (15%) group (P < 0.05, Fisher’s test). Additionally complete recovery was less likely in the severe TBI group (23%) than in the moderate TBI group (65%). (P < 0.05, Fisher’s test) (Tables 3, 4).

Table 3 Outcomes as measured by the Glasgow Outcome Score
Full size table
Table 4 The Glasgow Outcome Score
Full size table

Discussion

Traumatic brain injury in Ireland remains mainly an affliction of the young male population. Criminal assault and alcohol are factors in a large number of injuries. Though the majority of patients were admitted from the hospital’s emergency department, the transfer of four patients with a primary diagnosis of TBI from an outside hospital to a non-neurosurgical centre is a cause for concern.

Medical management of TBI patients in the ICU was broadly appropriate and in compliance with international guidelines. All patients underwent timely CT imaging (within 1 h of request). Appropriate blood pressure and oxygenation parameters and ventilator strategy were employed in almost all cases. The low use of seizure prophylaxis (32%) is at odds with best practice guidelines. Prophylactic anticonvulsants can decrease the incidence of early posttraumatic seizures from 14.2 to 3.6% [10]. Low level of mannitol use was acceptable in the absence of ICP monitoring.

None of our patients had intracranial pressure monitoring (ICP). Brain Trauma Foundation guidelines dictate that intracranial pressure should be monitored in all salvageable patients who present with a GCS score of 3–8 after resuscitation, and who have an abnormal CT scan [1]. Additionally, ICP monitoring may be indicated in patients over 40 years, with severe TBI and a normal CT scan, who have other features of severe injury such as unilateral or bilateral motor posturing or systolic blood pressure <90 mmHg [1]. The main objective of intracranial pressure monitoring is to maintain adequate cerebral perfusion and oxygenation and avoid secondary injury. Cerebral perfusion pressure (CPP), an indirect measure of cerebral perfusion, incorporates mean arterial pressure (MAP) and ICP parameters. CPP values below 50 are associated with poor outcome. The only way to reliably determine CPP and cerebral hypoperfusion is to continuously monitor ICP and blood pressure [11–13]. Protocols that incorporate ICP monitoring and other advanced monitoring have demonstrated improved outcomes when compared with earlier time periods without a protocol [14, 15]. Additionally, the frequency of ICP monitoring in trauma centres has been reported to be associated with improved outcomes [16]. Intracranial pressure monitoring is not available in Ireland outside of existing neurosurgical centres, limiting the option of ICP-guided therapy for patients with traumatic brain injury.

Analysis of the radiological and clinical evidence and comparison with current Brain Trauma Foundation guidelines suggested that 34% of our patients may have benefitted from surgical intervention. A decision to undertake operative intervention depends on many other factors, which may not be borne out in a retrospective chart review. As a result, this component of our data is inherently inaccurate, but the figure does raise the concern that without resource constraints more patients might receive surgical intervention.

The finding that only two of our patients were managed in a neurosurgical centre points to significant capacity constraint within the neurosurgical services in Ireland. The practice in Ireland and much of Europe is that trauma patients are transferred to the nearest emergency department, where they are triaged, stabilized and transferred to a specialist hospital if their injuries require such treatment. This contrasts sharply with the USA, where all patients with head injuries are brought initially to a Level 1 trauma centre, which by definition has on-site neurosurgical services. Our practice in Ireland can delay the transport of patients to receive neurosurgery, and transfers depend entirely on the available capacity at our existing centres.

Odds of death for patients with severe TBI managed in a non-neurosurgical centre are twice that of patients managed in a neurosurgical centre [17]. The improved outcomes in a neurosurgical centre relate to both surgical intervention and institution of specialist packages of care such as targeted control of intracranial pressures [15, 18]. Though patients with surgical lesions are typically given preference for transfer to neurosurgical centres, patients with non surgically treatable head injuries have a higher mortality [19], and interventions such as decompressive craniectomy may be beneficial [20, 21]. Additionally, outcome after traumatic brain injury is better in high-volume centres than in those seeing fewer head-injured patients [22]. Currently, 53% of patients with severe TBI in the UK are transferred to receive neurosurgical care [17], which is far greater than in our study. Both NICE (National Institute of Clinical Excellence, UK) and the Royal College of Surgeon guidelines advocate transfer of all patients with serious head injuries (GCS < 9) to neurosurgical centres, irrespective of the need for neurosurgery [23].

The overall mortality of 37% in our study was higher than internationally accepted norms. This figure is largely as a result of the high mortality in the severe TBI group (54%) and compares unfavourably with UK statistics, where the mortality in a group of severely brain injured patients (GCS < 8) with similar demographics was 35% [17]. As with other international studies, admission GCS in our group accurately predicted outcome [24]. Though there is no single identifiable factor that explains the higher than predicted mortality in our group, the lack of ICP monitoring and lack of surgical intervention and management in a non-specialist centre may all help account for the observed differences.

Despite 250 randomised controlled trials, no single therapy has been identified that would improve the outcome in head injury [21]. Despite this lack of a “magic bullet” though, the institution of packages of specialist neurocritical care would significantly improve outcomes [17, 18]. Our study demonstrates that specialist neurocritical care is unavailable to many severely brain injured patients in Ireland. Despite our institution of best medical care in a non-neurosurgical centre, the mortality rate in our study is significantly higher than for similarly injured patients in the UK. To improve our outcomes in this young group of patients and comply with international best practice, major investment will be required in both the infrastructure and personnel in our neurosurgical services. Perhaps the time has come for the implementation of a national brain injury strategy.