Retrospective review of traumatic brain injury: is skull X-ray still clinically relevant?

Introduction

Traumatic brain injury (TBI) is defined as an acute injury to the brain resulting from mechanical energy transferred to the head by an external force.¹ It represents a major global public health problem and remains one of the leading causes of death and long-term disability worldwide.² Epidemiological data in Europe demonstrates variability in TBI incidence across countries and ranges from approximately 47 to 694 cases per 100,000 population annually.³

Imaging plays a central role in the evaluation of patients with head injury.^2,4 The primary objectives of imaging are to identify lesions requiring immediate neurosurgical intervention, detect injuries that require close observation or medical management, and assist in prognosis.⁴ Computed tomography (CT) is currently regarded as the gold standard imaging modality in the acute assessment of head trauma due to its high sensitivity for detecting skull fractures and intracranial haemorrhage (ICH).^4,5 Current clinical guidelines, including those from the National Institute for Health and Care Excellence (NICE), the Scandinavian Neurotrauma Committee, Centers for Disease Control and Prevention (CDC), and the American College of Radiology (ACR) recommend CT as the primary imaging modality in patients with suspected TBI.^6,7,8,9 In addition to its diagnostic accuracy, CT is widely available, rapid, and well-suited for use in emergency settings.⁴

To optimize the use of CT and reduce unnecessary imaging, several clinical decision rules have been developed to guide its indication. Among the most widely used are the Canadian CT Head Rule, the New Orleans Criteria, and the National Emergency X-Radiography Utilization Study II (NEXUS-II).^10,11,12 These tools aim to identify patients at risk of clinically important intracranial injury while minimizing unnecessary exposure to ionizing radiation.⁵

Historically, skull radiography was widely used in the evaluation of head trauma, primarily for the detection of skull fractures.¹³ The presence of a skull fracture was considered a marker of increased risk for intracranial injury.^13,14 However, advances in imaging technology and new evidence have demonstrated the limited diagnostic value of skull X-ray in clinical practice.^15,16,17 Compared with CT, skull radiography has lower sensitivity for detecting fractures and is unable to identify intracranial pathology. As a result, modern clinical guidelines no longer recommend skull radiography as part of the routine diagnostic workup of head injury.^6,7,8,9 For example, current recommendations explicitly advise against the use of skull X-ray in the evaluation of mild TBI, particularly in paediatric populations, due to its low diagnostic yield.⁹

Despite these recommendations, skull radiography continues to be used in some clinical settings, which may reflect local institutional practices or other contextual factors.^17,18,19 Furthermore, data from studies on the use and diagnostic performance have demonstrated the limited reliability of skull X-ray.^15,16,17

The aim of this study was to evaluate the diagnostic performance of skull X-ray in patients presenting with head injury and to assess its association with intracranial haemorrhage and clinically significant traumatic brain injury (csTBI), using CT as the reference standard.

Patients and methods

Results

Study population

The study cohort consisted of 12,392 patient encounters recorded between January 1, 2018, and December 31, 2023. Among these, 10,875 (87.8%) were classified as head injuries without associated brain injury, while 1,517 (12.2%) were classified as traumatic brain injury (TBI). The cohort included 7,793 (62.9%) males and 4,599 (37.1%) females. The median age was 42.0 years (interquartile range [IQR] 11.1–70.3).

A total of 6,146 skull X-rays and 3,103 head CT scans were performed during the study period, corresponding to approximately 1,024 X-rays and 517 CT scans per year. Regarding imaging utilization: 31.5% (n = 3,902) of encounters had no imaging performed, 43.5% (n = 5,387) underwent skull X-ray only, 18.9% (n = 2,344) underwent head CT only and 6.1% (n = 759) underwent both skull X-ray and head CT. The distribution of imaging utilization is visualized in Figure 1. Detailed imaging findings are presented in Table 1.

FIGURE 1.

TABLE 1.

Imaging findings in skull X-ray and head CT

Head CT findings	N (%)	Skull X-ray findings	N (%)
No evidence of bone or brain injury	2,093 (67.5)	No reliable signs of injury	5,960 (97.0)
Fracture without intracranial haemorrhage	261 (8.4)	Skull fracture	153 (2.5)
Fracture with intracranial haemorrhage	308 (9.9)	Isolated nasal bone fracture	33 (0.5)
Intracranial haemorrhage without fracture	441 (14.2)

Discussion

In this retrospective study, we evaluated the diagnostic performance of skull radiography in patients with head injury, using CT as the reference standard. Our findings demonstrate three key observations: (1) skull X-ray has low sensitivity for detecting skull fractures, (2) the absence of a fracture on X-ray does not reliably exclude intracranial injury, and (3) despite its limited diagnostic value, skull radiography remains frequently used in clinical practice.

The sensitivity of skull X-ray was low, indicating that more than half of fractures identified on CT were missed (Table 2). These findings confirm the limited reliability of skull radiography as a diagnostic tool. In contrast, the specificity was relatively high, indicating that a fracture on X-ray was uncommon in patients without a fracture on CT (Table 2). This may partly reflect the low prevalence of radiographically visible fractures and the large proportion of normal examinations in our cohort.

Our results are consistent with prior literature. A meta-analysis by Hofman et al. reported a pooled sensitivity of 0.38 and specificity of 0.95, while Nekuda et al. demonstrated extremely low sensitivity of 0.00 and specificity of 0.94.^15,16 In contrast, the higher sensitivity of 0.71 reported by Chawla et al. likely reflects difference in study populations and medico-legal case selection.¹⁷ Clinical decision rules such as Canadian CT Head Rule and the New Orleans Criteria demonstrate substantially higher sensitivity of 0.90 and 0.97, respectively, further supporting the superiority of CT-based assessment.¹⁸

As shown in Table 3, the presence of a fracture on X-ray was associated with a significantly higher rate of ICH, indicating a meaningful association between fracture and ICH. These findings are consistent with prior studies. Dunning et al. reported that skull fracture identified on CT was associated with a substantially increased risk of ICH (relative risk (RR) 6.13, 95% CI 3.35–11.2).¹⁹ Similarly, Andronikou et al. found that more than three-quarters of children with intracranial abnormalities had an associated skull fracture.²⁰

However, most patients with ICH (72.3%) in the analysed subgroup were without a fracture on X-ray, highlighting the limited sensitivity of skull radiography as a screening tool. Furthermore, a substantial proportion of all CT examinations demonstrated ICH without skull fracture, making these cases inherently undetectable by skull X-ray (Table 1). Although csTBI was more frequent in patients with fracture on X-ray, this association did not reach statistical significance, likely due to the low number of events.

From a clinical perspective, an important finding of this study is that 11.2% of patients without a fracture on X-ray had ICH on CT. Thus, a normal skull radiograph does not exclude intracranial injury, as even in this high-risk subgroup approximately one in nine patients without a fracture on X-ray still had ICH. A skull radiograph may therefore provide false reassurance and delay CT, diagnosis and treatment. It should not be used as a triage or rule-out test. The decision for CT imaging should be guided by validated clinical decision rules rather than by the presence or absence of a fracture on X-ray and a normal radiograph should never be used to withhold CT when clinical criteria for imaging are met.

Despite its limited diagnostic value, skull radiography was performed in many cases (n = 6,146), representing almost half of all observed encounters, with most examinations yielding negative results (97.0%). This suggests that its continued use is not driven by diagnostic effectiveness, but rather by factors such as availability, perceived lower radiation exposure, and its continued inclusion in routine clinical workflows. In the subgroup of patients who underwent both imaging modalities, most CT scans were performed in patients without a fracture on X-ray (86.0%), indicating that clinicians may already rely more on clinical judgment than on radiographic results. In this context, skull X-ray may represent a redundant step that does not meaningfully influence patient management.

The continued use of skull radiography may also reflect historical practice patterns. Earlier guidelines recommended skull X-ray, particularly in paediatric patients, as a tool for detecting fractures and indirectly assessing the risk of intracranial injury.²¹ However, even early studies highlighted its limited diagnostic value.²² In addition, more recent evidence suggests that the accuracy of skull radiograph interpretation is variable and dependent on physician experience, with subtle or linear fractures particularly prone to misinterpretation.²³ In a paediatric study by Chung et al., physicians correctly identified 63% of fractures shorter than 5 cm, compared with 93% of fractures longer than 5 cm.²³ Together, these findings indicate that the limitations of skull radiography are multifactorial, reflecting both the limitations of the imaging modality and variability in interpretation.

In our institution, the persistent use of skull X-ray most likely reflects established institutional practice and clinician habit rather than diagnostic necessity or limited CT availability, as CT is accessible on site at all times. In other settings and institutions, restricted access to CT and medico-legal considerations may also contribute to its continued use. Due to the retrospective design of our study, the specific reasons for ordering skull X-ray were not recorded, so these explanations should be interpreted as plausible contributors rather than established causes.

This study has several important strengths, including a large sample size and the use of CT as the reference standard, which enhances the reliability of the findings. In addition, the analysis of diagnostic performance and the inclusion of real-world clinical data increase its practical relevance. However, its limitations include the retrospective study design, the single-centre patient cohort, and the fact that not all patients underwent both CT and skull X-ray imaging, which may have affected the assessment of diagnostic accuracy. Furthermore, because CT was not performed systematically but rather as a decision of the treating physician, the patients who underwent both skull X-ray and CT represent a selected, higher-risk subgroup. This verification (selection) bias means that lower-risk patients, in whom findings would more often be truly negative, are underrepresented, so the observed prevalence of ICH (13.3%) is likely higher than in the overall cohort. Consequently, the estimates of sensitivity, specificity and predictive values should be interpreted as applying to a higher-risk population and may not be directly generalizable to all patients presenting with head injury.

In conclusion, skull radiography demonstrates low sensitivity for detecting skull fractures and limited utility in excluding intracranial injury. While the presence of a fracture increases the likelihood of ICH, its absence does not provide sufficient reassurance. These findings support current guideline recommendations that CT should remain the primary imaging modality in patients with head injury and suggest that routine use of skull X-ray may be unnecessary.