The present study demonstrates that S100B has good diagnostic ability to rule out significant intracranial pathology in patients that present to the ED within 6-hours of injury. Theoretically, a quarter of CT-head scans could be avoided and one in five patients could be discharged earlier if S100B was used as a rule out test in addition to the NICE CT-head guidelines, with < 1% risk of missing a significant injury. Within 24 hours, S100B still had a high NPV but was much less sensitive and the risk of missed injury increased to 2%. However, theoretical CT-head reduction increases to over a third and a third of patients could potentially be discharged faster.
Regarding the optimal timing of S100B measurement, this study demonstrates that within 6-hours of head injury is superior to within 24-hours of injury. The 6-hour data in this study is consistent with international data12 demonstrating the diagnostic ability of S100B to exclude intracranial pathology within 6-hours was between 83–99% sensitivity with a NPV of 87.5–100%. While a 3-hour time limit was noted to be superior with sensitivity of 98–100% and NPV of 97–100%, we did not view a 3-hour window to be long enough for clinical practice. In our setting patients with TBI on average present to ED 2.8 hours following head injury24 and clinicians work to targets of 6 hours length of stay23. On this basis, we view 6-hours as a more realistic and clinically relevant time point for analysis.
Regarding the false negative results or rather the ‘missed injuries,’ one case was identified within the 6-hour time frame. The case was an 86-year-old female who was on dabigatran and had suffered a ground level fall. Her CT scan showed a pinpoint haemorrhage in the left frontal lobe reported to be a contusion. Her case was discussed with the neurosurgical team who advised no intervention was required but observation by the medical team for 24 hours was appropriate. No further deterioration occurred in hospital, and she was discharged after the observation period. Within the 24-hour time frame, six cases had false negative results including the case previously described. The other five included: Two males aged 97 and 82-years-old on aspirin who sustained small subdural bleeds; a 67-year-old female with a small subarachnoid bleed; a 34-year-old male with a small contusion; and a 47-year-old female with a tiny contusion. All cases were discussed with neurosurgery and apart from one were not clinically reviewed and no further input was required. In the case of the 47-year-old, there was a question as to whether the lesion was a glioma, so she was admitted for further investigation and monitoring. Subsequently this was confirmed to be a small traumatic contusion that required no further intervention.
These findings open an interesting conversation with regards to firstly, what makes an injury clinically significant? And secondly, what level of risk are clinicians willing to accept clinically? There is no consensus regarding the definition of a clinically significant injury from an ED perspective. Currently international guidelines, including NICE CT-head guidelines recommend admitting all patients with traumatic CT abnormalities.28, 29 One can presume clinicians would agree that those requiring neurosurgical intervention are clinically significant. On the other end of the spectrum, radiologically significant injuries, as listed in our methods section, include the small contusions and bleeds that contributed to this studies ‘miss rate’ and reduced sensitivity. Yet apart from one, that notably was not necessarily because of a traumatic injury, no neurosurgical review, monitoring or management plan was required. One could argue these were not clinically significant and had S100B theoretically being used as a rule out test, it is unlikely any harm would have come to these patients. Internationally, clinical decision rules for early discharge of patients with traumatic CT abnormalities are being developed and require further prospective validation.30 This interesting area regarding clinically significant injuries is ripe for further investigation; in the meantime S100B appears to be safest within the 6-hour window in order to minimise clinical risk.
Knowing what clinicians and patients will tolerate with regards to risk of missed injury is an interesting conundrum. Not only does individual patient exposure to CT matter,31, 32 but the wider risk to ED attendees who present to overcrowded departments is important.33, 34 In the most part, CT-heads are considered safe but there are risks of radiation exposure that are not yet fully understood.33 ED overcrowding increases mortality for all conditions, 33, 34 and waits for CT head scan can delay intervention for those with intracranial injuries.7 Often those with negative CTs are suffering from concussion,35, 36 being present in the stimulating ED environment likely exacerbates concussive symptoms and therefore earlier discharge from ED may reduce symptom burden. Reduced CT head rates would potentially benefit departments in terms of reduced crowding, reduced resource load and reduced health care costs. From a patient’s perspective, the potential benefits include shorter waits, earlier intervention and earlier discharge.
It is unlikely that any biomarker incorporated into clinical guidelines will be a perfect test. Tests in clinical use rarely are, for example, d-dimer for PE and troponin for myocardial infarction are not 100% sensitive and do not have 100% NPV.37 Therefore, it is important that when looking to incorporate biomarkers clinically, it is targeted to specific clinical settings that minimises the risk to patients and effectively supports clinical decision making. The usefulness of a clinical pathway for TBI that incorporates biomarkers depends on the tolerance of risk for missed injury or deterioration in those discharged without a CT-head scan. The NICE guidelines are well validated and utilised internationally to aid clinicians with CT-head use in ED settings. They are not a completely perfect tool; in fact, a prospective multi-centre study noted that NICE criteria missed 2.4% of all injuries.38 Therefore, incorporating biomarkers into existing clinical guidelines translates the limitations of the original guideline to the cohort of patients that are selected for biomarker testing. This limitation needs to be considered when developing new guidelines. It is worth noting that a guideline is just that, and in certain settings it is reasonable for clinicians to deviate from guidelines based on their clinical judgement.
The findings from this study are similar to other international data investigating S100B, particularly with regards to sensitivity, negative predictive value and theoretical CT head reduction rates.12, 13, 39, 40 It is worth noting that these studies vary somewhat clinically in terms of included population and CT decision threshold. However, given international results are similar, with caution it is likely other international S100B data used in conjunction with NICE CT-head guidelines could be generalised to an Australasian setting. This study has attempted to add to existing literature by focusing the use of S100B to a cohort of patients where S100B could add the most clinical and economic value. All patients in this study were GCS 15, had no diagnosed post traumatic amnesia and were able to consent for themselves at the time of recruitment. Theoretically this equates to a real time clinical cohort of patients that could be potentially discharged from ED after a negative S100B result. This population have a low pre-test probability for significant traumatic CT injury reflecting a group with low clinical risk where CT reduction can be targeted, benefiting patients in terms of shorter stays and faster discharge. This study has highlighted some of the pressures currently facing EDs that could be eased with the incorporation of biomarkers that reduce CT demand and enable early discharge. Although overcrowding was not directly measured, patients on average waited 2.5 hours for CT scans in ED and only 50% were discharged within the recommended 6-hour length of stay target. Theoretically use of S100B in our clinical setting could reduce CT-head request rates and enable early discharge in this low-risk group.
There are several limitations to report from the present study. Firstly, it is worth noting that the consent process is not necessarily reflective of how the biomarker would be used during real time clinical application. It was common for patients to be consented following their CT scan rather than at the decision point for CT-head, which may affect the applicability of the results. In this study only 50% of the target population had S100B tested within 6 hours of injury as there was not always a trained investigator present to obtain consent. If biomarker testing were part of routine care, this proportion would most certainly increase. Furthermore, ideally samples need to be processed within 2–4 hours of blood collection, something that most certainly would occur if used in a real-time clinical model. As mentioned in the methods section, stability has been demonstrated up to 72 hours post collection for research or pre-hospital purposes. In this study 20% of samples were processed within 2 hours, 25% within 4, 63% within 24 hours, 77% within 48 hours. This may affect the accuracy of the biomarker results.
This study only included patients who were GCS 15, had no post traumatic amnesia and were able to consent to the study. This does bias the mild head injury cohort somewhat and therefore these results are only generalisable to this population. For the purpose of this study, this population was deliberately selected to represent the cohort of patients that potentially could be discharged as previously discussed. Notable groups this would miss are those with reduced GCS because of intoxication, where biomarkers could help distinguish between patients that require CT-head scans versus observation whilst sobering up, elderly patients with dementia, post-ictal phases of generalised seizures and those with mental health conditions. Previous literature has reported that alcohol does not affect S100B levels41 so our results could still be potentially relevant for this group. The other groups likely require further investigation.
There is also a low pre-test probability of finding an intracranial injury which means that the overall number of positive CT heads in the study is low. We are reassured these results are still of value given our sample size is above the median for studies of this type and our key results have narrow confidence intervals that are within similar ranges of other published studies.12 However, positive CT heads are the critical factor when determining the diagnostic performance of the test, so ideally a larger sample with higher positive CT rates is required to rigorously test S100B in a variety of clinical settings. It should also be noted that the NICE guidelines were not developed to exclude non-depressed fractures. Our sample size is insufficient to examine the performance of S100B if we excluded these from the CT findings considered “positive” within this study.
No outcome data beyond hospital discharge were recorded in terms of re-presentation, morbidity or concussion outcomes. Given this study was predominantly testing how S100B could potentially aid CT-head request decisions, it is unlikely this affected the main study outcomes. In order to maximise recruitment and have a representative sample from ED, all ED clinicians were able to consent patients to this study. To minimise interruption to clinical duties some clinical data such as presenting features and ED outcomes were collected retrospectively by the lead study investigators. We do not expect this to have significantly affected the main study outcome.
An interesting demographic observation noted in the present study is the loss of typical bimodal age distribution. Classically there is a peak of TBI in the 18–35-year cohort and a further peak in the elderly.25,26 In our study, the peak in the younger cohort is not seen, although on review of those excluded, it was notably the younger cohorts who did not want to have a blood test performed or were unable to consent due to intoxication. S100B is unaffected by alcohol27 and therefore this is a group of patients that could potentially be targeted with blood biomarkers.
The other notable clinical observation is the absence of any positive CT heads with an extradural bleed (EDH). EDH more often affect younger cohorts and mortality can occur suddenly after a period of lucidness.28 Pathologically, initially this injury may not be associated with brain injury in the early stages but as the haematoma expands, brain injury will occur. Theoretically, if S100B was tested too early in this process, it could miss significant injuries. However, studies investigating the use of S100B in Scandinavia, where S100B is used clinically in local neurotrauma guidelines,29 noted that all EDH reported in the literature at the time had an S100B level over the current validated threshold.30
Within the current study we did not exclude patients presenting with polytrauma. As previously mentioned, adipose, muscle and cardiac tissue can all release S100B in response to trauma. This is likely to have increased the rate of false positives observed in our cohort, and clinically this could limit the reduction in head CT rates. Currently the sample size is too small to generate meaningful results if patients presenting with polytrauma were excluded. But this is a future area for investigation. S100B levels can also be affected by poor renal function and increasing age and potentially a higher S100B threshold could be more appropriate in the over 65 age group.42 Furthermore, a high proportion of patients in this study were on anticoagulation and although there is no evidence to suggest this affects S100B levels,43 further evidence is required.