2 Research recommendations
The Guideline Development Group has made the following recommendations for research, based on its review of evidence, to improve NICE guidance and patient care in the future. The Guideline Development Group's full set of research recommendations is detailed in the full guideline.
What is the clinical and cost effectiveness of the 2014 NICE guideline recommendation on CT head scanning versus clinical decision rules (including CHALICE, CATCH and PECARN) for selecting children and infants for head CT scanning?
The recommendations in this updated guideline for determining which patients need a CT head scan are based on the CHALICE clinical decision rule. CHALICE was derived in the UK but has yet to be validated, and limited evidence has been identified since the NICE clinical guideline was published in 2007. There is a need for a prospective validation and direct comparison of the 2014 NICE guideline and CHALICE, CATCH and PECARN clinical decision rules in a UK setting to determine diagnostic accuracy (sensitivity, specificity, and predictive values for intracranial injury and the need for neurosurgery) and cost effectiveness within the relevant population to which the NICE guideline is applied.
The study should be a prospective study with economic evaluation and should capture subgroups by age, separating out infants (under 2 years), children and young people (under 16 years) and adolescents (16–18 years). The results of such a study will confirm whether current practice is optimal and, if not, which would be the ideal clinical decision rule to implement in a UK population. To warrant recommendation of a different clinical decision rule and a consequent substantial change in practice, significant improvement in diagnostic accuracy must be demonstrated. This can only be done through such a prospective comparative validation study performed in our population.
In patients with head injury does the use of antiplatelet and anticoagulant drugs increase the risk of intracranial haemorrhage over and above factors included in the current recommendations for CT head scans?
Antiplatelet and anticoagulant drugs are widely and increasingly prescribed, and many patients presenting with a head injury to the emergency department are taking these drugs. While the majority of these drugs are prescribed in older patients they are also used in younger people. This guideline provides recommendations on performing CT head scans in patients on warfarin. However, limited evidence has been identified for patients using other antiplatelet or anticoagulant drugs within studies deriving or validating clinical decision rules for determining which patients need CT head scans. There is a particular paucity of evidence in determining whether they are at increased risk of intracranial haemorrhage.
A study with appropriate economic evaluation is needed to quantify the risk of taking these drugs over and above the risk factors included in an existing clinical decision rule. Antiplatelet and anticoagulant drugs should be studied as a predictor of intracranial injury and analysed within a multivariate analysis with other predictors (including the risk factors used in this guideline to determine when a CT head scan is needed). Univariable analyses of risk of intracranial injury in groups of head injury patients who are taking these agents and those who are not, and who have no other indications for CT head scan under current guidance would also be useful.
The GDG felt that, where possible, each drug should be considered separately, particularly aspirin and clopidogrel, and that the reference standard should include CT head scan and a follow-up period of sufficient duration to capture delayed bleeding, for example, at 7 days and 1 month. Analysis would benefit from subgroup results by age (children, adults and patients over 65 years). The GDG suggested reporting similar data used in the AHEAD study.
In adults with medium risk indications for brain injury under the 2014 NICE CT head injury guidance, what is the clinical and cost effectiveness of using the diagnostic circulating biomarker S100B to rule out significant intracranial injury?
Circulating biomarkers, if validated, could provide a convenient and clinically applicable aid to the diagnosis of mild traumatic brain injury (TBI) – a 'troponin for the brain'. If such biomarkers were sufficiently sensitive as well as specific for injury type (separating patients with traumatic axonal injury (TAI) from those with contusions), panels of biomarkers might not only help to determine which patients need neuroimaging but also allow us to devise rational, cost-effective pathways for neuroimaging – perhaps reserving primary use of advanced MRI for patients who have TAI as these lesions are undetectable on CT head scans. In addition, the availability of quantifiable biomarkers, scaled with the severity of injury, could help clinicians monitor the progression of brain injury in patients with more severe TBI, help stratify patients for trials and therapies, and provide significant prognostic information across all severities of TBI.
There is low-quality clinical effectiveness data for using the biomarker S100B to rule out significant intracranial injury in patients in the emergency department. Current evidence suggests that there is variation in the use of biomarker tests, including in the timing of testing, the concentration of biomarker used as a diagnostic cut-off, protocols used for sample transport and storage, and the equipment used for biomarker assays in laboratories. A diagnostic study (using randomised or consecutively selected patients) is needed to investigate the role of S100B in patients with selected head injury patterns.
The GDG also recognised the potential utility use of near-patient testing for biomarker tests to reduce the time from injury and blood sampling to test results. In addition, the GDG would welcome an additional outcome of 3-month follow-up of functional outcome/post-concussion symptoms alongside this study with appropriate economic evaluation. This research would provide UK-based evidence as to the potential benefit of biomarkers and any associated reduction in CT head scans and hospital admissions.
Research is needed to summarise and identify the optimal predictor variables for long-term sequelae following mild traumatic brain injury (TBI). A systematic review of the literature could be used to derive a clinical decision rule to identify relevant patients at the time of injury. This would in turn lay the foundation for a derivation cohort study.
Although this recommendation was first made in 2007, the GDG felt that this is still an area of high priority for research and the question remains unanswered. The diagnosis of TBI is essentially a clinical one. However, although this approach provides the best current solution it can be imprecise, particularly in mild TBI where conventional imaging may be normal and cognitive abnormalities may be due to confounders such as pre-existing dementia, hypoxia or hypotension from associated injuries, alcohol or recreational drugs, and/or other conditions (such as post-traumatic stress disorder) which result in overlapping phenotypes (and possibly even imaging findings).
The availability of novel, objective methods of detecting brain injury provides an attractive means of better defining the presence of TBI in these contexts, with improvements in epidemiological precision. Perhaps more importantly, there is an increasing recognition that even mild TBI can result in prolonged cognitive and behavioural deficits, and the ability to identify patients at risk of these sequelae would aid clinical management, help determine which patients need novel therapeutic interventions, and refine resource allocation.
The techniques that have been explored in this regard include advanced neuroimaging with MRI, electroencephalographic (EEG) based diagnosis, and circulating biomarkers. The relative effectiveness and cost effectiveness of these techniques, individually and in combination, is not yet completely defined, and their role in contributing to a clinical decision rule that allows triage of patients to specific management pathways needs definition. A systematic review would be the first step in collating the available evidence in this area, followed by a rational application of available evidence, identification of key research questions that need to be addressed, and definition of the data collection needed in a derivation cohort study that allows these questions to be addressed.