Head injury in the elderly – an overview for the physician
ABSTRACT
Head injury is a common cause for hospital admission and additionally 250,000 UK inpatients fall during hospital admissions annually. Head injury most commonly occurs as a result of falls from standing height in older adults. Older adults are frequently frail and multi-morbid; many have indications for anticoagulation and antiplatelet agents. The haemorrhagic complications of head injury occur in up to 16% of anticoagulated patients sustaining a head injury. These patients suffer adverse outcomes from surgery as a result of medical complications. Although geriatric trauma models are evolving to meet the demand of an ageing trauma population, medical support to trauma services is commonly delivered by general physicians, many of whom lack experience and training in this field. Determining the role of surgery and interrupted anticoagulation requires careful personalised risk assessment. Appreciation of the opposing risks can be challenging; it requires an understanding of the evidence base in both surgery and medicine to rationalise decision making and inform communication. This article aims to provide an overview for the physician with clinical responsibility for patients who have sustained head injury.
Introduction
Head injury has historically been the domain of the surgeon. However, populations are ageing rapidly, and the number of persons aged over 85 is anticipated to double in forthcoming decades.1 In England and Wales there are 1.4 million hospital attendances due to head injury per annum, the majority of which occur in older persons due to low-energy falls from standing height.2
Between 2009 and 2015, the rate of patients being newly prescribed oral anticoagulation increased by 58%, with over 500 per 100,000 people receiving oral anticoagulation prescriptions in 2015.3 Where head injury arises from falls in the context of anticoagulants or antiplatelet agents, haemorrhagic complications may be as high as 15.9%.4
Although neurosurgical teams are highly skilled in the management of haemorrhagic complications, age is associated with the accumulation of multi-morbidity.5 Multi-morbid patients challenge traditional surgical services because increasing super-specialisation has attenuated generalist skills. Adverse outcomes in older surgical patients are well-recognised,6 and these typically occur primarily as a consequence of medical complications of surgery. In the modern era, we need to recognise that traditional surgical services cannot deliver high-quality medical care to complex patients without the support of medical teams.
Modern models of collaborative trauma care have evolved in orthopaedics; however these services are not yet ubiquitous. Impending changes to best practice tariffs for major trauma centres across England may soon incentivise diversion of geriatrician resource, but in many centres, medical support to trauma teams is provided by the duty medical registrar, who typically receives little (or no) dedicated training. Physicians are therefore frequently inadequately equipped to handle situations which are often complicated.
Furthermore, many haemorrhagic complications of head injury are sustained during complex medical admissions as a result of the 250,000 inpatient falls occurring annually in the UK. Trauma services do not have capacity to take over the care of all inpatients with complications of head injury, and typically only accept candidates for neurosurgery. Physicians therefore often have clinical responsibility for the care of patients with complications of head injury, yet lack knowledge and expertise to manage this cohort optimally.
There is therefore a need for better knowledge of head injury amongst physicians to provide more effective support to trauma services, and advocate for patients sustaining head injury during medical admissions.
The effects of ageing on the brain
A number of physiological changes occur with ageing that predispose older patients to haemorrhagic complications of head injury (Table 1).
Investigation
Indications for head imaging
National Institute for Health and Care Excellence guidelines for head computed tomography (CT) are described in Table 2.2 However, evidence indicates that 30% of intracranial injuries do not present with reliable clinical findings.10 Atrophy of the ageing brain can allow older patients to tolerate substantial intracranial haemorrhage better than younger patients with equivalent injury. This can lead to underestimation of the extent or severity of injury in older persons and delayed presentation. Non-contrast CT head imaging may therefore be appropriate in all older patients presenting with significant head injury, especially if imaging outcome will influence medical decision making (eg prescribing of anticoagulants). It should be noted that Trauma Audit and Research Network (TARN) data indicate that older patients wait longer for head imaging than younger patients, despite being at higher risk of bleeding and or adverse outcomes.11
Most of the common haemorrhagic consequences of head injury can be adequately imaged by CT (Table 3), although magnetic resonance imaging (MRI) may have greater sensitivity for evaluating diffuse parenchymal damage, especially where small volume injury is not detected on plain CT. In patients on anticoagulation with initial normal CT, there is a small risk of subsequent scans demonstrating haemorrhage. Anti-coagulated patients may therefore benefit from a period of observation and repeat CT imaging prior to discharge.12
Management
Indications for neurosurgery
The role of neurosurgery is to relieve elevated intracranial pressure (ICP), improve the cerebral perfusion pressure and correct depressed skull fractures. NICE guidelines state that the following signs indicate neurosurgical consultation:
unexplained confusion of >4 hours
persistent Glasgow coma score (GCS) <8
deteriorating GCS after admission
development of focal neurological signs
seizure without full recovery
penetrating head injury
cerebrospinal fluid (CSF) leak.2
North American consensus guidelines indicate that in patients with potential to benefit, surgical evacuation of intracranial haemorrhage should be considered (regardless of GCS) where radiological parameters show:
either subdural haematoma with 10 mm+ clot thickness or 5 mm+ midline shift22
or epidural haematoma with volume >30 cm3.23
Surgery may also be indicated where these criteria are not met, and there remains clinical evidence of brain herniation or raised ICP (eg anisocoria), especially in the context of documented clinical deterioration. Surgical outcomes may be dependent on time between onset of neurological signs and surgical decompression. Where surgery is indicated, it should take place within 4 hours.
In patients not meeting surgical criteria but with GCS <9, ICP monitoring in a critical care setting is advised. Surgery should be considered if ICPs are sustained above 20 mmHg. There exists considerable debate and controversy as to the indication and outcome of decompressive neurosurgery in these patients who do not fit the haematoma criteria above. Various studies suggest that these patients often have improved survival but potentially worse neurological outcomes, and a role for decompressive craniectomy in patients with advanced frailty is generally less accepted.
Non-operative management
Non-operative management of raised ICP is appropriate when there is a small haematoma, or the intracerebral pressure is below 20 mmHg.22 Non-surgical interventions include head elevation, sedation, analgesia, mechanical hyperventilation to maintain normal pCO2, euthermia, antiepileptics, mannitol (or hypertonic saline) and ICP monitoring.24 In patients who are potential surgical candidates, the risk of haematoma expansion indicates serial CT 8–12 hours after injury.25 For ICP monitoring, ventricular catheters are considered the gold standard in the USA. These devices permit therapeutic CSF drainage if required, although they do carry a greater infection risk. In the UK intraparenchymal monitoring devices are more commonly deployed.26
Outcomes from intracranial haemorrhage have been shown to be superior in patients transferred to a specialist neurology or neurosurgical unit, even if surgical intervention is not required immediately.27 In symptomatic patients who would be good surgical candidates should they deteriorate, transfer to a neurosurgical centre may therefore be appropriate for monitoring and medical management (eg for severe hypertension).
Neurosurgery in the older patient
Despite the availability of consensus guidelines, older patients are less likely to undergo neurosurgical intervention than younger patients with equivalent injury. This may be because older patients treated surgically have inferior outcomes.28 In one study, survival following surgical intervention for traumatic head injury was 59.3% for ages 65–74 and only 32.4% for those aged over 75.29 However, compared to age matched peers treated conservatively, older persons undergoing surgery following head trauma had significantly reduced mortality and better functional outcomes.28 Debate persists about the benefits and role of decompressive neurosurgery in the older patient with haemorrhagic complications of head injury. This is complicated by difficulties interpreting the literature when it is known that older patients tend to tolerate more substantial intracranial injury before reflecting the severity of their injury in their GCS. Adverse outcomes in older patients with matched GCS to younger patients may therefore represent more severe injury and unfair comparison.
Balancing risk and benefit in older patients with complications of head injury is therefore complex. Careful consideration of likely post-operative outcome is required on an individual basis. Age and frailty are not synonymous, and wide functional heterogeneity can be found in older persons. Emerging evidence indicates that frailty assessment provides a more accurate prediction of surgical risk than age, and frailty seems to intuitively influence the decision to operate in patients undergoing oncological neurosurgery.30 There is some evidence that frailty is associated with poor outcomes in geriatric trauma, but little published data specific to elderly traumatic brain injury. Sarcopaenia defined by CT criteria is emerging as a potential surrogate for frailty assessment that may prove useful in predicting outcome in the future.31
Role of the physician
Importantly, the UK hub and spoke trauma model means that many decisions on whether to operate are currently made without the direct assessment of an experienced neurosurgeon. The decision on whether or not to offer surgery therefore depends on information-gathering and communication in relation to baseline state. Combined with injury severity, this determines potential for recovery and allows neurosurgical prognostication. However, subsequent communication with patient and family is often the responsibility of the referring medical team, especially where surgery is not indicated. This emphasises the need for physicians to understand the basis and rationale for decision-making.
In addition to review of mechanism of injury and optimisation of co-morbidity, the responsibility of the physicians is also to treat complications and promote rehabilitation. Management of key medical issues is discussed in the following paragraphs.
Management of anticoagulation
Many older persons regularly take anticoagulants and antiplatelet agents which contribute to haemorrhagic complications of head injury. Challenging decisions are often required pertaining to the reversal of anticoagulant and antiplatelet effect. Timed reintroduction of these drugs following haemorrhagic complications of head injury is also complex.
In these cases, judicious assessment of risk is required. The risk of thrombosis in each individual must be balanced against the risks of recurrent intracranial bleeding. These risks are often difficult to estimate for general physicians without experience in the management of traumatic head injury. This can lead to poor decision-making if there are delays in obtaining specialist advice.
Mortality from haemorrhagic complications of head injury in the presence of continued anticoagulation is strikingly elevated, and may reach 80.6%.32 The risks associated with cessation of anticoagulation are generally extrapolated from annualised population risks (Table 4) and therefore may be higher in the acute setting. Notwithstanding this observation, the risks of thrombosis are usually substantially lower than that of the acute bleed, even for patients considered at high risk of thrombosis. Suspension of anticoagulation is therefore almost always required; this should be discussed with a haematologist with specialist coagulation experience. A risk–benefit analysis and management plan should be clearly documented and regularly reviewed.
Reversal of anticoagulants
NICE recommends that warfarinised patients presenting with suspected or confirmed intracranial bleeding should undergo immediate reversal with prothrombin complex concentrate (PCC).40 This is supported by the British Committee for Standards in Haematology, which states that following head-injury admission anticoagulated patients are likely to require a CT scan, need their international normalised ratio (INR) measured, and if head injury is strongly suspected reversal of the INR with PCC.41 The use of recombinant factor VIIa or fresh frozen plasma is not acceptable management.42 Further measures should include 5–10mg of intravenous vitamin K, immediate cessation of all anticoagulant therapy and serial measurement of INR at 30 minutes, 4–6 hours and 24 hours in case further doses of PCC are required.
The reversal of direct oral anticoagulants (DOACs) is more complex. The factor II inhibitor dabigatran can be effectively reversed with idarucizumab. Andexenant, a specific reversal agent for factor X inhibitors, is in development. Meanwhile, expert consensus currently advises that suspected intracranial bleeding in patients taking the commonly prescribed factor X inhibiting DOACs, be reversed with PCC, although this is unlikely to be fully effective.43 Tranexamic acid can be used adjunctively and carries limited risk of thrombosis.
Reversal of antiplatelet effect
Some evidence indicates that clopidogrel may predispose to traumatic intracranial haemorrhage to a greater extent than warfarin. Limited evidence is available to determine the risks of modern potent antiplatelet agents. Strategies for reversal of antiplatelet effect include regular tranexamic acid, platelet infusion (typically 2 pools) and in extreme cases recombinant-activated factor VIIa. Of note, whilst two pools of platelets would be expected to largely reverse aspirin effect, the prolonged pharmacodynamics of clopidogrel (5–7 day washout) limits the efficacy of platelet transfusion and multiple pools are often required. However, none of these interventions improves outcome following intracerebral haemorrhage.44 Suspension of antiplatelet agents following acute intracranial haemorrhage is however advised.
Restarting anticoagulants and antiplatelets
Although haematoma expansion and re-bleeding are significant risks following intracerebral haemorrhage, there is also an increased risk of thromboembolism precipitated by immobility, the inflammatory response to trauma and pro-coagulant effects of reversal agents. Non-pharmacological VTE prophylaxis should be instigated and understanding this evolving risk profile permits timely reinstatement of anti-coagulation therapy where thrombosis risk is deemed high.45
In patients with high thrombosis risk, current evidence indicates that anticoagulant therapy can be reinstated after 10 days with modest risk of re-bleeding. However, the risks of reinstating anticoagulation will vary depending on the extent and site of intracranial bleeding. There is no current consensus on what constitutes an ideal interval, and prospective studies to answer this question are currently underway. Many surgeons prefer to delay reinstating anticoagulation until CT brain indicates complete resolution of haemorrhage. Inevitably, this requires judicious appreciation of the risks of continued suspension of anticoagulation. However, it should also be noted that the risks of interrupted anticoagulant therapy are modest even in the context of modern metallic heart valves, where valve thrombosis rates of 4% per annum are seen in patients not taking anticoagulants.37 Outcomes for patients with metallic heart valves whose anticoagulation was suspended report a variable thrombosis risk 0–4% at 30 days.46
Guidelines recommend 7–10 days of suspended anti-coagulation following haemorrhagic complications of head injury.47 Evidence pertaining to the reintroduction of DOACs following haemorrhagic head injury is limited, but re-bleeding following intracerebral haemorrhage is rarer on DOACs than warfarin and likely to be safer.48 Expert consensus therefore also recommends restarting DOACs following 7–10 days of suspension where thrombosis risk is high. The evidence surrounding reinstatement of anti-platelet agents is limited. However, anti-platelet agents are safer than anti-coagulants after intracerebral haemorrhage. Expert consensus indicates a 7–10 day cessation period may therefore be appropriate.49
For patients at extremely high risk of thrombosis in whom it is thought early re-anticoagulation is indicated <7 days, or where re-bleeding risk is considered higher, graduated re-introduction of escalating doses of twice per day (bd) low molecular weight heparin (LMWH) is standard practice. This is generally considered safer and more effective than unfractionated heparin. Of note, in the case of re-bleeding, up to 60% of LMWH effect can be antagonised by protamine. Embolisation from recent lower limb venous thrombosis may be mitigated by IVC filter insertion.
Management of thrombocytopaenia
Thrombocytopaenia carries a >12-fold risk of haematoma expansion.49 Patients presenting with a platelet count <135 x 109/L are 31.5 times more likely to require neurosurgical intervention. Thrombocytopenic patients with intracranial haemorrhage should receive platelet transfusion to maintain platelets >100 x 109/L.49
Seizures
Prophylaxis
Early seizures occur in 30% of patients following traumatic brain injury. Most seizures occur within the first 24 hours of injury. Seizures are believed to increase the risks of secondary brain injury through functional hypoxia and raised ICP. Prophylaxis with anticonvulsants has been shown to reduce the frequency of early seizures, although anticonvulsants have no effect on the development of long-term epilepsy after brain injury. A recent survey of neurosurgical centres in the UK suggests that 50% of neurosurgical centres routinely use anticonvulsants for seizure prophylaxis after brain injury.50 Current evidence indicates that levetiracetam has a better safety profile and equal efficacy to phenytoin. Levetiracetam may also be associated with better functional outcomes. Prophylaxis with levetiracetam 500 – 1000 mg bd for 7 days is therefore recommended.51 In patients who have not suffered a seizure during this period, there is no advantage to long-term anticonvulsant therapy. In elderly patients, lower dosages should be considered and it is important that duration of therapy is confirmed prior to discharge.
Treatment
In patients sustaining seizures following traumatic brain injury, the risk of long-term seizures depends upon injury severity. Approximately 8–16% will suffer from post-traumatic epilepsy at 2 years. Additionally, patients with persistently reduced consciousness considered disproportionate to the extent of neurological injury, 15–20% will display signs of non-convulsive seizures on continuous electroencephalography.52 Anticonvulsants are typically continued throughout the admission for patients who have sustained a seizure, and weaned gradually after discharge under neurological supervision.
Hyponatraemia
Hyponatraemia is a common complication of intracranial injury. Careful clinical evaluation is required to differentiate between the syndromes of inappropriate antidiuretic hormone secretion (SIADH), cerebral salt wasting (CSW) and other causes such as hypoadrenalism or drugs. SIADH is the most common electrolyte disturbance following head injury, but may be mistaken for CSW. The pathophysiology of CSW is poorly understood although it is known to result in excess renal excretion of sodium and dehydration. Evaluation of hyponatraemia is complicated by the fact that SIADH and CSW often display similar laboratory parameters. However, treatment is diametrically opposed and incorrect treatment risks deterioration in hyponatraemia. Differentiation therefore mandates careful evaluation of fluid status. Patients suffering from CSW are intravascularly deplete and require intravenous volume replacement; patients with SIADH are euvolaemic and require fluid restriction. Resolution of CSW typically occurs within 2–4 weeks of head injury.53
Prognosis
Elderly patients suffer high mortality and reduced functional outcome following head injury when compared to younger cohorts; a recent UK cohort study reported 22.9% inpatient mortality, 10.8% long-term moderate disability and 5.3% severe disability requiring permanent help with activities of daily living.54 Outcomes are considerably worse for severe traumatic brain injury with inpatient mortality as high as 70–80%.
Conclusions
Physicians are increasingly called upon to provide medical support for trauma services, whose population is increasingly old, frail and comorbid. Frail older patients often sustain head injury when they fall, and are predisposed to haemorrhagic complications because of anticoagulant use and the effects of ageing. Determining the role of surgery and interrupted anticoagulation requires careful, personalised risk assessment. Appreciation of the opposing risks can be challenging; it requires an understanding of the evidence base in both surgery and medicine to rationalise decision making and inform bilateral communication.
- © Royal College of Physicians 2019. All rights reserved.
References
- ↵
- Office for National Statistics
- ↵
- National Institute for Health and Care Excellence
- ↵
- Loo S
- ↵
- Alrajhi K
- ↵
- Department of Health
- ↵
- Adhiyaman V
- Teunissen L
- Tobe EH
- ↵
- ↵
- The Trauma Audit and Research Network (TARN)
- ↵
- Zare M
- Agrawal M
- Jauch E
- Talbott JF
- Demetriades D
- Vieira R
- ↵
- ↵
- Ragland J
- ↵
- Sumritpradit P
- ↵
- Tavakoli S
- ↵
- ↵
- Shimoda K
- ↵
- Leung G
- ↵
- Youngerman BE
- ↵
- Hu P
- ↵
- ↵
- Bogseth M
- National Institute for Health and Care Excellence
- Kim J
- Cannegieter S
- ↵
- Aviña-Zubieta J
- Kreher S
- National Institute for Health and Care Excellence
- ↵
- ↵
- Frontera J
- ↵
- Steiner T
- ↵
- Jaben E
- ↵
- Murthy S
- ↵
- ↵
- Broderick J
- ↵
- Milling T
- ↵
- Goldstein J
- ↵
- Bogseth M
- ↵
- Mee H
- ↵
- Xu JC
- ↵
- Zimmermann LL
- ↵
- Verbalis JG
- ↵
- Hawley C
Article Tools
Citation Manager Formats
Jump to section
Related Articles
- No related articles found.