Injuries in Fatalities of Dismounted Blast: Identification of Four Mechanisms of Head and Spine Injury

Abstract

Blast is the most common injury mechanism in conflicts of this century due to the widespread use of explosives, confirmed by recent conflicts such as in Ukraine. Data from conflicts in the last century such as Northern Ireland, the Falklands, and Vietnam up to the present day show that between 16% and 21% of personnel suffered a traumatic brain injury. Typical features of fatal brain injury to those outside of a vehicle (hereafter referred to as dismounted) due to blast include the presence of hemorrhagic brain injury alongside skull fractures rather than isolated penetrating injuries more typical of traditional ballistic head injuries. The heterogeneity of dismounted blast has meant that analysis from databases is limited and therefore a detailed look at the radiological aspects of injury is needed to understand the mechanism and pathology of dismounted blast brain injury. The aim of this study was to identify the head and spinal injuries in fatalities due to dismounted blast. All UK military fatalities from dismounted blast who suffered a head injury from 2007–2013 in the Iraq and Afghanistan conflicts were identified retrospectively. Postmortem computerized tomography images (CTPMs) were interrogated for injuries to the head, neck, and spine. All injuries were documented and classified using a radiology brain injury classification (BIC) tool. Chi-squared (χ² ) and Fisher’s exact tests were used to investigate correlations between injuries, along with odds ratios for determining the direction of correlation. The correlations were clustered. There were 71 fatalities from dismounted blast with an associated head injury with a CTPM or initial CT available for analysis. The results showed the heterogeneity of injury from dismounted blast but also some potential identifiable injury constellations. These were: intracranial haemorrhage, intracranial deep haemorrhage, spinal injury, and facial injury. These identified injury patterns can now be investigated to consider injury mechanisms and so develop mitigation strategies or clinical treatments.

Level of Evidence: Observational.

Study type: cohort observational.

Introduction

Blast is the most common injury mechanism in conflicts of this century^1

–4 due to the widespread use of explosives and improvised explosive devices (IEDs) and this has continued up to this day in Ukraine.⁵ Blast has multi-modal effects: primary, secondary, tertiary, and quaternary. Primary injury is the result of the blast wave, which affects the interfaces between materials of different densities such as gas or fluid-filled cavities like the lung and bowel. Secondary blast injury is trauma (anatomical disruption) caused by highly energized fragments from the blast, which is thought to be higher in dismounted blast than mounted (in-vehicle) blast due to the lack of protection from the vehicle exterior.⁶ Tertiary blast is where the force of the blast physically moves the body or large materials causing extreme blunt force trauma.⁷ Quaternary blast is an injury that is not encompassed by the other three types and includes those produced by chemicals, blood-born viruses, or heat injury.⁸ It is important to understand the categories of blast as an exposed victim will experience primary blast differently to a mounted victim. Traumatic brain injury (TBI) is commonly suffered in conflict,¹ and historically brain injuries seen in soldiers were predominantly a result of penetrating, high energized fragments, including shells or ammunition,⁹ causing injuries such as occipital penetrating injuries.¹⁰ In dismounted personnel there are fewer impact skull fractures and more penetrating injuries,¹¹ and severe head injury is uncommon (5%).¹² Injuries from recent conflicts have included hemorrhagic injury and skull fractures.^3,13,14 As much of the current research is based on categorical data from spreadsheets and databases (including the Joint Trauma and Theater Registry-JTTR; Smith et al.¹⁵), the lack of imaging precluded the identification of hemorrhagic and fracture patterns. The overall injuries are different between mounted and dismounted cohorts with a higher prevalence of dismounted fatalities (11).

As the pattern of injuries in the mounted cohort from this dataset has been previously published,¹⁶ this article will analyze dismounted injuries only in order to hypothesize the mechanisms that produced these in order to enable future mitigation and protection to be investigated. The aim of this study is to combine JTTR injury scoring with radiological findings to identify the patterns of head injury of fatalities in dismounted blast in order to elucidate the mechanisms causing such injuries.

Methods to obtain relevant information on conflict injuries are facilitated for UK military personnel by the JTTR.¹⁵ The UK JTTR is a database comprising all injuries sustained for both survivors and fatalities from operation Telic (Iraq) and Herrick (Afghanistan). Parameters collected and used here include Glasgow Coma Scale score and anatomical injuries scored using the Abbreviated Injury Scale (AIS version 2005 Military2). In addition, radiological images are available for many of the injuries. This is the first analysis of head injuries in dismounted fatalities.

Methods

With permission from the Surgeon General (UK), Defence Medical Services Caldicott Guardian, a JTTR search was conducted for all patients who had suffered a brain injury. The search criteria were fatalities with a head injury with AIS of 2 or more due to dismounted blast. The inclusion dates were 2007–2013 to include fatalities from Iraq and Afghanistan. CT postmortems (CTPMs) were collected as soon as possible following confirmation of fatality.¹⁷ In some cases, patients survived until transfer to a UK medical facility before subsequently dying of wounds in hospital (n = 3); in these cases, the initial CT upon arrival to the theater of operation was used.

Data collected were as follows: basic demographics, radiological features, and the two body regions with the most severe injuries. In detail, head injury severity and type were scored radiologically using the Society of British Neurological Surgeons Brain Injury Classification (SBNS BIC)¹⁸ which assigns a score of 0–3 for each injury; inclusive of extradural hematoma, acute subdural hematoma, chronic subdural hematoma, subarachnoid hemorrhage, contusions, diffuse axonal injury, hypoxia, brain tightness (raised intracranial pressure), presence of skull fracture, and whether the injury was penetrating (secondary blast) or blunt force (tertiary blast) trauma. Facial injury was described as any injury to the face that had fractured facial bones. The radiology records were investigated and classified by a consultant radiologist, neurosurgical consultant, and a trauma nurse. Additional radiological parameters were collected to ensure complete analysis, these were: ventricular blood and its location, pattern of injury, skull fractures, spinal injury and location, and extracranial injury of AIS 3 or more in the abdomen or thorax.

To test the hypothesis that there exists a discrete set of constellations or patterns of injury, a correlation analysis of injuries was conducted, where Pearson’s chi-squared (χ²) or Fisher’s exact test was used to test the independence of each individual variable against each other. Fisher’s exact was used when there were fewer than five counts of a variable; chi-squared (χ²) was in those variables with five or more counts. An alpha of 0.08 or less was used as the threshold to reject independence, higher than normally used as the objective is to hypothesize injury mechanisms and the numbers available were low considering the heterogeneity of injury constellations. As multiple independence tests were calculated on the data, the risk of false positives (Type I error) was increased. However, the risk of Type I error was deemed to have a lower impact than the risk of a Type II error, so Bonferroni correction was not applied to the data.¹⁹ In this case, the risk of a Type II error would risk missing potential mechanisms and therefore limit future work, particularly where this study might produce new hypotheses that could be subsequently tested. Previous publications have placed emphasis on the importance of exploring potential hypotheses that would otherwise be ignored should Bonferroni correction be applied, thereby not missing important findings.²⁰ As this is a new dataset, which has not been analyzed in this way, multiple comparisons need to be made to ensure correct hypotheses. An odds ratio was performed where possible to determine whether the correlation was positive or negative.

Statistical analysis was performed using Microsoft Excel 2007 (version 12.0, Redmond, WA) and IBM SPSS (version 26, Armonk, NY).

Results

Complete CTPMs were available for 71 UK military fatalities of dismounted blast (2007–2013); 11 fatalities had no images or CTPM available. Radiological brain injury parameters are shown in Figure 1; skull fracture was the most prevalent injury (in 59 of the fatalities, 83.1%) followed by subarachnoid hemorrhage (31, 43.7%). Contusion (18, 13.1%), tightness (6, 8.5%), acute subdural hematoma (5, 7.0%), and extradural hematoma (1, 1.4%) were seen in much smaller numbers (Table 1). Chronic subdural hemorrhage, diffuse axonal injury, and hypoxia were not seen on imaging.

FIG. 1.

Chi-squared (χ²) and Fisher’s exact analysis of independent variables. Those variables where independence could be rejected are highlighted in color by injury constellation. Blue—intracranial hemorrhage, orange—deep hemorrhage, purple—facial, green—spinal. Where the shade of color is lighter it indicates a p value between 0.05 and 0.08. Where a shade of color is darker it indicates p < 0.05. The cells marked black are negative correlations.

Table 1.

Pathophysiology of Dismounted Blast Brain Injury as Categorized by the SBNS Brain Injury Classification (n = 71)

	Extradural	Acutesubdural	Chronicsubdural	Subarachnoid	Contusion	Diffuseaxonal injury	Hypoxia	Tightness	Skull fracture(Y/N)
Number of fatalities (n = 71)	1	5	0	31	13	0	0	6	59
Percentage of total	1.4	7.0	0.0	43.7	18.3	0.0	0.0	8.5	83.1

The most common additional injury was blood located in the ventricles (Table 2), equally in the lateral ventricles (23, 32.4%) and fourth ventricle (22, 31.0%). Blood was also seen in the foramen magnum in 5 patients (7.0%), brainstem in 2 (2.8%), and pneumocephalus in one patient (1.4%).

Table 2.

Location of Blood Identified Radiologically in Dismounted Blast Fatalities with Head Injury (n = 71)

	4th ventricle	Lateral ventricles	Foramen magnum	Brainstem	Pneumocephalus
Number of fatalities (n = 71)	22	23	5	2	1
Percentage of total	31.0	32.4	7.0	2.8	1.4

Spinal fractures were seen in all regions of the spine (Table 3) but were most prevalent in the sacral region (19, 26.8%) and in C0-C1 (10, 14.1%) and thoracic spine (10, 14.1%) regions. Other C-spine fractures occurred in eight (11.3%) fatalities and lumbar fractures in seven (9.9%) fatalities.

Table 3.

Spinal Fractures in Dismounted Blast Fatalities with Head Injury (n = 71)

	C0-C1	Other C-spine	T	L	S
Number of fatalities (n = 71)	10	8	10	7	19
Percentage of total	14.1	11.3	14.1	9.9	26.8

Penetrating injuries due to secondary blast were prevalent in 21 (29.6%) of fatalities.

Based on AIS scores, thorax injury was described as the secondary injury in 41 (57.8%) fatalities, abdominal in 32 (45.1%). When analyzing skull fractures (Table 4), facial only fractures were seen in 25 (35.2%) of fatalities. Of those with facial fractures, only 16 had no further underlying brain injury (blood in the ventricles excluded). Four (5.6%) fatalities had facial fractures with an additional base of skull fracture, three of whom had no demonstrable underlying brain injury on imaging, which was confirmed through the accompanying radiology reports.

Table 4.

Facial and Skull Injuries in Dismounted Blast Fatalities Who Had Head Injury (n = 71)

	Facial fractures only with brain injury	Facial fractures only with no brain injury	Skull fractures only with brain injury	Both facial and skull fractures	Facial fractures with base of skull fracture only	Decapitation or little brain tissue/skull left	No fractures with brain injury
Number of fatalities (n = 71)	25	16	7	18	4	9	8
Percentage of total	35.2	22.5	9.9	25.4	5.6	12.7	11.3

There were no correlations between pneumocephalus and any other injuries. This is excluded from the results (Fig. 1). Odds ratios less than 1.0 showed a negative correlation between thoracic injury and fourth ventricle blood, abdominal and penetrating injury, and skull fracture with tightness, resulting in the identification of four mostly distinct injury clusters. The first is nonoverlapping with other constellations and contains acute subdural, subarachnoid, and extradural injuries (which we term intracranial hemorrhage). There were 32/71 fatalities with this injury constellation. The second injury constellation, which we term deep hemorrhage, is a proximal set of injuries with blood in the ventricles, around the brainstem, and perimesencephalic blood (total: 37/71 fatalities). The third injury constellation partially overlaps with the second with both exhibiting L- and S-spine injuries, the spinal injury constellation, comprising C-, T-, L-, and S-spine injuries combined with abdomen and thorax injury (total:45/71 fatalities). The fourth injury constellation, facial, includes facial injuries, skull fracture, and other C-spine injuries (total: 61/71 fatalities). Twenty-seven (38%) had a CT scan displaying no brain tissue injury (excluding intraventricular blood).

Discussion

This study is the first to identify injury constellations of fatalities with associated head and neck injuries in the dismounted cohort using data from the JTTR with the available scans and images of postmortems from Operation Herrick in Afghanistan and Telic in Iraq. Four injury constellations have been identified as follows (Table 5).

Table 5.

Coexistence of the Four Identified Injury Constellations

Injury constellations	Number of fatalities (%)
Spinal injury and facial injury	17 (23.9)
Deep hemorrhage, intracranial hemorrhage, and facial injury	10 (14.1)
Deep hemorrhage, spinal injury, and facial injury	9 (12.7)
All four constellations	8 (11.3)
Deep hemorrhage and facial	5 (7.0)
Facial injury alone	5 (7.0)
Deep hemorrhage, spinal injury, and facial injury	4 (5.6)
Deep hemorrhage, intracranial hemorrhage, and spinal injury	3 (4.2)
Intracranial injury and facial injury	3 (4.2)
Intracranial injury and spinal injury	2 (2.8)
Spinal injury alone	2 (2.8)
Deep hemorrhage and Intracranial hemorrhage	1 (1.4)
Deep hemorrhage alone	1 (1.4)
Intracranial injury alone	1 (1.4)
Deep hemorrhage and spinal injury	0 (0.0)

Intracranial hemorrhage

The cluster of correlation between acute subdural, subarachnoid hemorrhage, contusions, and extradural hemorrhage show that there was prevalence of intracranial hemorrhage even in early fatalities. Although these injuries are nonspecific to any direct mechanism, they identify that the PPE used did not prevent intracranial injury.

Deep hemorrhage

Similar to the mounted cohort¹⁶ perimesencephalic injury was seen and is unusual. The number of fatalities with perimesencephalic blood was much higher in this cohort than typically seen in a civilian population.²¹ In civilians, perimesencephalic blood is usually seen as a result of straining or physical exertion, mimicking the Valsalva maneuver, causing veins and capillaries to swell and a nonaneurysmal bleed to occur. In civilians, this pathology is not fatal and patients recover quickly.²² A hypothesized mechanism, based on the direct connection between the intrathoracic and intracerebral vascular system (the modified Monro–Kellie principle²³), is that blast wave induced rise in intrathoracic pressure is transmitted directly through the vascular system resulting in perimesencephalic vascular rupture in much the same way as seen in spontaneous perimesencephalic bleeds that occur when straining (e.g., valsalva maneuvers).²⁴ Further mechanistic research or using larger datasets will enable this to be explored further.

Spinal injury

The cluster of spinal injuries and injury to the thorax and abdomen suggests evidence of massive through-body injury as a result of blast. This type of injury, although unlikely to be amenable to prehospital treatment due to its severity, could be mitigated through modification of personal protective equipment. We hypothesize that the presence of abdominal or thorax injury may be indicative of an underlying spinal injury in the survivors’ cohort, and so this should be investigated. Spinal injuries were due to blunt force as opposed to penetrating injury.

Facial injury

Fifty-eight percent of fatalities had a facial injury alone with no additional skull fractures (n = 25 with brain injury, n = 16 with no brain injury). Of these, 25.6% had no underlying brain injury other than blood in the ventricles. However, there were four (5.6%) fatalities who had a facial injury with a base of skull fracture only, most of whom had no underlying brain injury other than blood in the ventricles (Table 5). The additional correlation between facial injuries and skull fractures and facial injuries with C0-C1 injury shows a likely mechanism to be an impact force to the face and neck.

As the fatalities were close to the blast, these fractures suggest that the facial bones could be acting as a “crumple” zone thereby preventing underlying brain injury. This has been hypothesized by previous research in civilian trauma which found that there was a link between frontal sinus volume, cranial facial insult, and underlying intracranial trauma supporting a “crumple zone” hypothesis.²⁵ In addition, there are some parallels between this blast injury constellation and temporal-mandible joint fractures in motorcyclists with incomplete helmet protection who have suffered a collision.²⁶

There has been previous research exploring the link between facial fractures and underlying brain injury, suggesting that facial injury is a predictor of central nervous system trauma²⁷ and that concussion was the highest incident injury. As the images in this study were from CTPMs, it is not possible to identify if patients had suffered a concussion. We would hypothesize that this would have been present in survivors. In addition, due to the instant impact of death, microhemorrhages or diffuse axonal injury (DAI) would not necessarily be visualized. The absence of DAI on imaging, in itself, is not evidence of the absence of injury due to the complexity of DAI diagnosis. Further research in this area could be possible in a survivor cohort; however, this is entirely dependent on whether the facial injuries were survivable and did not compromise the airway.

Some injuries had a negative correlation meaning that they did not occur together, such as tightness and skull fractures. This is likely due to the fact that the skull fracture prevented pressure from building up. The negative correlation between abdomen injury and penetrating injury is not understood and highlights the complex environment as does the negative correlation between thoracic injury and fourth ventricle blood.

The four injury patterns are mostly not totally independent with many coexisting. The most common coexisting injuries being spinal with facial injury where 53.5% of fatalities had both injury constellations (including in combination with others). There were just eight (11.3%) casualties who had an overlap of all four injuries suggesting that all four injuries were unlikely (or due to whole body blast exposure causing injury across all four constellations) (Tables 4 and 5).

The theory behind the mechanisms for these injury constellations can only be hypothesized at this point. Those fatalities with the spinal injury constellation would have been located very close to the blast causing catastrophic whole body injury. Those fatalities with facial injury would have most likely been further away and facing the blast therefore causing frontal and facial fractures.

Deep hemorrhage appears to occur in around one-third of fatalities, but the variability of this cannot at this point be explained. However, it is important to note that an isolated deep hemorrhage (perimesencephalic injury) is not usually associated with fatality in civilians.

There is currently limited research on the underlying radiology and pathophysiology of the dismounted cohort beyond the injuries listed on the JTTR, particularly in the survivor population. Previously, in survivors, research has focused on minor TBI and the crossover with post-traumatic stress disorder, whereas for the basis of this research, the moderate to severe TBI cohort would be comparable. Moderate to severe blast TBI in survivors is yet to be explored.

This study is limited by the use of imaging and the JTTR dataset alone. Information on vehicle data, proximity to blast, and personal protective equipment would provide further information on the injuries seen. However, these data are sensitive and can expose military vulnerabilities. It is also important to note that there are some instances where the dataset is incomplete.

There may be biases in the data where some data were missing initially, in particular in those cases where patients were not taken initially to a UK facility where imaging could be recorded. In these specific regions, there could be a difference in the IEDs used and therefore a different injury was seen. The “died of wounds” cohort is underrepresented and therefore it is hard to stipulate conclusions based on those patients who had longer survival than those who had instant fatality as a result of blast.

Computerized tomography postmortems in themselves can be limiting. There is some evidence of body degradation in some images due to the length of time between death and imaging. In addition, the act of resuscitation can cause some disruption to the body and leave artifacts.

Injuries due to blast are unique in pathophysiology and although some can be compared to nonblast trauma, the mechanisms behind such injuries are hypothetical until further research is undertaken. The injury constellations presented here can be used to identify mitigation strategies and clinical care for future patients.

Transparency, rigor, and reproducibility summary

Due to being the first study of its kind in this cohort, as many cases as possible were analyzed, therefore this is a sample of convenience. Seventy-one patients had scans amenable to analysis. These were analyzed using a validated radiological classification tool. Bonferroni correction was not applied to the data in order to identify potential correlations so that hypotheses on injury mechanisms could be produced. Chi-squared (χ² ) and Fisher’s exact were used to reject independence, and odds ratios determined the direction of correlation. Correlated injuries were clustered to hypothesize mechanisms. Future work should include a power calculation so hypotheses can be tested in an appropriately powered dataset. Given the sensitive nature of the study, the data are not widely available and requests for data will be considered on a case-by-case basis and subject to UK Ministry of Defence clearance.

Footnotes

Authors’ Contributions

E.R.A.: Conceptualization, Methodology, Formal analysis, Investigation, Data Curation, Writing—Original Draft, Writing—Review & Editing, Visualization, and Project administration. D.B.: Investigation and Writing—Review & Editing. I.G.: Investigation, Writing—Review & Editing, and Supervision. M.W.: Writing—Review & Editing and Supervision. A.B.: Conceptualization, Methodology, Writing—Review & Editing, Visualization, Supervision, Project administration, and Funding acquisition.

Funding Information

Funding was received by the Royal British Legion for the support of the lead author’s PhD. Professor Mark Wilson is additionally funded for research by the NIHR.

Author Disclosure Statement

No potential conflicts of interest are reported by the authors. Lt Emily Ashworth, Lt Col David Baxter, and Colonel Iain Gibb are serving members of the British Armed Forces.

Disclaimer

The opinions expressed in this article are those of the authors alone and do not necessarily represent the views or official policy of the Royal Navy, the British Army, the Ministry of Defence, or His Majesty’s Government.

References

Champion

, Bellamy

, Roberts

, et al. A profile of combat injury. J Trauma, 2003; 8.

Bhattacharjee

. Neuroscience. shell shock revisited: Solving the puzzle of blast trauma. Science, 2008; 319(5862):406–408; doi: 10.1126/science.319.5862.406

Wade

, Dye

, Mohrle

, et al. Head, face, and neck injuries during operation Iraqi freedom II: Results From the US Navy and Marine Corps Combat Trauma Registry:. Defense Technical Information Center: Fort Belvoir, VA; 2006; doi: 10.21236/ADA445195

Belmont

PJJ

, McCriskin

, Sieg

, et al. Combat wounds in Iraq and Afghanistan from 2005 to 2009. J Trauma Acute Care Surg, 2012; 73(1):3–12; doi: 10.1097/TA.0b013e318250bfb4

Zhang

, Botterbush

, Bagdady

, et al. Blast-related traumatic brain injuries secondary to thermobaric explosives: Implications for the war in Ukraine. World Neurosurg, 2022; 167:176–183.e4; doi: 10.1016/j.wneu.2022.08.073

Bryden

, Tilghman

, Hinds

, 2nd. Blast-related traumatic brain injury: Current concepts and research considerations. J Exp Neurosci, 2019; 13:1179069519872213–1179069519872213; doi: 10.1177/1179069519872213

Clasper

, Edward

. Blast Injury Mechanism in Blast Injury Science and Engineering. Eds Bull

Anthony

, Clasper

, Mahoney

Peter

, McGregor

Alison

, Masouros

Spyros

, Ramasamy

Arul

. 2nd Edition. Springer Nature; 2022.

Centres for Disease Control and Prevention. Explosions and Blast Injuries: A Primer for Clinicians. 2006.

Kocsis

, Tessler

. Pathology of blast-related brain injury. J Rehabil Res Dev, 2009; 46(6):667–672; doi: 10.1682/jrrd.2008.08.0100

10.

Lanska

. Historical perspective: Neurological advances from studies of war injuries and illnesses. Ann Neurol, 2009; 66(4):444–459; doi: 10.1002/ana.21822

11.

Singleton

JAG

, Gibb

, Hunt

NCA

, et al. Identifying future ‘unexpected’ survivors: A retrospective cohort study of fatal injury patterns in victims of improvised explosive devices. BMJ Open, 2013; 3(8):e003130; doi: 10.1136/bmjopen-2013-003130

12.

, Tubb

, Poepping

, et al. Dismounted blast injuries in patients treated at a role 3 military hospital in Afghanistan: Patterns of injury and mortality. Mil Med, 2016; 181(9):1069–1074; doi: 10.7205/MILMED-D-15-00264

13.

Eastridge

, Mabry

, Seguin

, et al. Death on the battlefield (2001–2011): Implications for the future of combat casualty care. J Trauma Acute Care Surg, 2012; 73(6 Suppl 5):S431–437; doi: 10.1097/TA.0b013e3182755dcc

14.

Regasa

, Agimi

, Stout

. Traumatic brain injury following military deployment: Evaluation of diagnosis and cause of injury. J Head Trauma Rehabil, 2019; 34(1):21–29; doi: 10.1097/HTR.0000000000000417

15.

Smith

, Hodgetts

, Mahoney

, et al. Trauma governance in the UK defence medical services. J R Army Med Corps, 2007; 153(4):239–242; discussion 243; doi: 10.1136/jramc-153-04-04

16.

Ashworth

, Baxter

, Gibb

, et al. Injuries in underbody blast fatalities: Identification of five distinct mechanisms of head injury. J Neurotrauma, 2022; 40(1–2):141–147; doi: 10.1089/neu.2021.0400

17.

Gibb

, Delaney

, Murphy

, et al. Post-mortem computed tomography in the investigation of conflict and terrorist related deaths: UK military experience of developing a multidisciplinary service. Clin Radiol, 2022; 78(11):804–811; doi: 10.1016/j.crad.2022.07.007

18.

Wilson

, Ashworth

, Hutchinson

, British Neurotrauma Group . A proposed novel traumatic brain injury classification system—an overview and inter-rater reliability validation on behalf of the Society of British Neurological Surgeons. Br J Neurosurg, 2022; 36(5):633–638; doi: 10.1080/02688697.2022.2090509

19.

Armstrong

. When to use the Bonferroni correction. Ophthalmic Physiol Opt, 2014; 34(5):502–508; doi: 10.1111/opo.12131

20.

Rothman

. No adjustments are needed for multiple comparisons. Epidemiol Camb Mass, 1990; 1(1):43–46.

21.

Şahin

, Delen

, Korfali

. Perimesencephalic subarachnoid hemorrhage: Etiologies, risk factors, and necessity of the second angiogram. Asian J Neurosurg, 2016; 11(1):50–53; doi: 10.4103/1793-5482.165793

22.

Matsuyama

, Okuchi

, Seki

, et al. Perimesencephalic nonaneurysmal subarachnoid hemorrhage caused by physical exertion. Neurol Med Chir (Tokyo), 2006; 46(6):277–281; doi: 10.2176/nmc.46.277 discussion 281–282.

23.

Wilson

. Monro-Kellie 2.0: The dynamic vascular and venous pathophysiological components of intracranial pressure. J Cereb Blood Flow Metab, 2016; 36(8):1338–1350; doi: 10.1177/0271678X16648711

24.

Monro

. Observations on the Structure and Functions of the Nervous System. Illustrated With Tables. By Monro

Alexander

,. Creech and Johnson: Edinburgh; 1783.

25.

Cai

, Mossop

, Diaconu

, et al. The “Crumple Zone” hypothesis: Association of frontal sinus volume and cerebral injury after craniofacial trauma. J Craniomaxillofac Surg, 2017; 45(7):1094–1098; doi: 10.1016/j.jcms.2017.04.005

26.

Cavalcante

, Oka

SC-R

, de Santana Santos

, et al. Influence of helmet use in facial trauma and moderate traumatic brain injury victims of motorcycle accidents. J Craniofac Surg, 2012; 23(4).

27.

Pappachan

, Alexander

. Correlating facial fractures and cranial injuries. J Oral Maxillofac Surg, 2006; 64(7):1023–1029; doi: 10.1016/j.joms.2006.03.021