Abstract
Objectives:
This study evaluates the diagnostic yield and utilization patterns of head and neck computed tomography angiography (CTAHN) in screening patients in the emergency department (ED) for blunt cerebrovascular injury. This study was conducted at a Level 1 trauma center serving a large geographical area with a high number of regional transfers. This study evaluates the prevalence of CTAHN overutilization in the ED and examines whether diagnostic yield correlates with adherence to the modified Denver criteria, transfer status, and trauma severity.
Material and Methods:
A retrospective chart review was conducted of patients presenting to the ED between January 01, 2021, and December 31, 2023. Patients over the age of 16 were included in this study if they presented with blunt traumatic injury and subsequently underwent CTAHN or neck CTA (CTAN) imaging during their ED evaluation. A multivariable regression analysis was performed to evaluate the association between CTA yield and factors including patient transfer status, trauma severity designation, and adherence to the modified Denver criteria.
Results:
A total of 643 patients met the inclusion criteria. A substantial proportion (74.2%) of CTAHN and CTAN examinations yielded negative results, with significant correlation to nonadherence to the modified Denver criteria. We found no significant association between CTAHN yield and transfer status or trauma level designation. Of the patients found to have trauma related actionable findings, 71.8% had adherence to the modified Denver criteria during triage screening.
Conclusion:
A high proportion of CTAHN and CTAN studies ordered from the ED yield low-value diagnostic results, with overutilization closely linked to nonadherence to the modified Denver criteria. Our findings carry important implications for quality improvement efforts, including unnecessary radiation exposure to patients and reduced efficiency in the ED. Enhanced guideline compliance may serve as a modifiable target to reduce diagnostic imaging overutilization.
Keywords: Blunt cerebrovascular injury, Computed tomography angiography, Denver criteria, Diagnostic imaging overutilization, Emergency department
INTRODUCTION
Screening patients presenting to the emergency department (ED) for blunt cerebrovascular injury (BCVI) with computed tomography angiography (CTA) can be a critical step in the prevention of lethal outcomes.[1,2] However, over the past two decades, the utilization of head and neck CTA (CTAHN) has increased disproportionately relative to other neuroimaging modalities, with an alarming number of studies yielding negative or clinically insignificant findings.[3] One study found that from 2001 to 2015, head and neck CTA utilization in the ED increased 14,600% and 17,781%, respectively, for medicare beneficiaries.[4,5] Some estimate that up to 85% of CTA screenings may have negative or insignificant findings, a trend likely driven by the accessibility, rapid acquisition, and diagnostic efficiency of CTA in acute care settings.[6]
The Denver criteria are widely regarded as the gold standard for assessing the need for CTA screening following blunt trauma involving the head and neck and are endorsed by the Western Trauma Association and used by numerous institutions. Since the introduction of BCVI screening guidelines some 20 years ago, identifying at-risk patients has been the subject of debate, with some reports suggesting that current guidelines fail to identify up to 37% of all BCVI.[7,8] Some studies have argued that ordering CTA based on clinical suspicion alone, without extensive survey and strict adherence to formalized screening criteria, may be justified out of precaution and urgency to avoid missing occult vascular injuries. Consequently, the optimal strategy for BCVI screening remains a contentious issue, with varying perspectives on whether current protocols sufficiently capture at-risk patients.[9]
Reducing the frequency of low-yield CTA orders from the ED has the potential to improve triage efficiency, enhance diagnostic accuracy, and mitigate unnecessary healthcare expenditures and provider workload, ultimately advancing patient care and decreasing morbidity and mortality.[1] While there is growing literature on diagnostic imaging overutilization, the relationship between CTA use and the detection of clinically significant BCVI in the emergency setting has yet to be thoroughly explored. To our knowledge, this study is the first comprehensive evaluation of CTAHN diagnostic yield at a Level 1 trauma center serving a large and predominantly rural population with substantial healthcare disparities and a high volume of regional transfers. We aim to evaluate the prevalence of clinically significant CTAHN imaging findings in blunt trauma patients and to determine whether diagnostic yield is associated with nonadherence to the modified Denver criteria, patient transfer status, and trauma severity designation.
MATERIAL AND METHODS
This retrospective cohort study was conducted at University Medical Center in Lubbock, Texas, USA -- a level 1 trauma center serving a large geographical area with a predominantly rural patient population. This study was approved by the Institutional Review Board, with waiver of the informed consent requirement, to include all trauma patients for a period of 3 years who presented to the ED between January 01, 2021, and December 31, 2023, and underwent a CTAHN or neck CTA (CTAN).
Inclusion and exclusion criteria
Patients were eligible for inclusion if they presented with traumatic injuries involving the head or neck and received either a CTAHN or CTAN during the ED evaluation. The following criteria were used to exclude CTA scans from this study:
CTAHN or CTAN for patients ≤16 years old
Non-diagnostic scans due to significant motion artifact or suboptimal image quality
Duplicate or repeat imaging following an initial CTAHN or CTAN ordered in the ED
Non-trauma related presentations, such as altered mental status, headache, weakness, focal neurological deficits, or other non-traumatic complaints.
Trauma classification and imaging outcomes
Patients were first classified by the mechanism of trauma into one of two categories: Blunt or penetrating injuries. The blunt trauma patients were then split into two groups: Adult and pediatric, pediatric being anyone 16 years or younger [Figure 1]. In this paper, we analyze the adult cohort for blunt trauma.
Figure 1:
Flow diagram illustrating patient selection and categorization for head and neck computed tomography angiography (CTA) performed in the emergency department. Adult blunt trauma patients (n = 643) were classified into negative, incidental, non–trauma related actionable, and trauma related actionable CTA findings. CTAHN: Head and neck computed tomography angiography, ED: Emergency department.
Mechanisms of blunt trauma were subclassified into falls (from ground or elevated levels), motor vehicle-associated accidents (including pedestrian versus auto events and motorcycle accidents), near hanging or strangulation, and assault using a blunt object or weapon.
CTA findings were classified into four categories according to the following scale:
Negative
- Incidental findings, including:
- Atherosclerosis <50% stenosis per North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria
- Minimal vessel calcification
- Evidence of previous stroke
- Vessel tortuosity (spiraling, looping, coiling) or hypoplasia
- Chronic small vessel disease.
- Non-trauma related actionable findings, including:
- Atherosclerosis ≥50% stenosis per NASCET criteria
- Extensive vessel calcification
- Previously identified or non-trauma related aneurysm.
- Trauma related actionable findings, including:
- Vascular dissections
- Aneurysms or pseudoaneurysms
- Vessel thrombosis or occlusions
- Emboli
- Arteriovenous fistulas
- Active contrast extravasations.
This classification scheme was formulated to isolate clinically significant and actionable findings from acute traumatic mechanisms. This was to distinguish these findings from chronic and incidental vascular pathology or trauma suffered at a previous time. Representative imaging examples are provided to illustrate the spectrum of CTA findings identified in this cohort. Figure 2 demonstrates a non-trauma related but clinically actionable finding (≥70% carotid stenosis by NASCET criteria), which frequently prompted outpatient vascular referral. Figures 3-5 depict trauma related actionable findings, including vertebral artery dissection with associated thrombus, intracranial aneurysm, and extracranial vertebral artery occlusion, respectively. These imaging findings directly influenced acute management decisions, including antithrombotic initiation, blood pressure control, and subspecialty consultation.
Figure 2:
Axial head and neck computed tomography angiography demonstrating a non-trauma related actionable finding. A large noncalcified atherosclerotic plaque at the origin of the right internal carotid artery (yellow arrow) produces approximately 70% luminal stenosis by North American Symptomatic Carotid Endarterectomy Trial criteria in a 65-year-old male.
Figure 3:
Axial head and neck computed tomography angiography demonstrating a trauma related actionable vascular injury. An acute displaced fracture of the left C1 lateral mass involving the foramen transversarium is associated with acute injury of the distal cervical left vertebral artery extending into the proximal intracranial segment (yellow arrow), consistent with arterial dissection and associated non-occlusive thrombus, in a 72-year-old female following blunt trauma.
Figure 5:
Coronal head and neck computed tomography angiography demonstrating a trauma related actionable vascular injury. Complete occlusion of the left extracranial vertebral artery extending from its origin (lower yellow arrow) to the level of the C4 vertebral body (upper yellow arrow) in a 19-year-old female following blunt trauma, findings consistent with posttraumatic vertebral artery dissection, with distal reconstitution and luminal opacification at the level of C4, extending into the intracranial segment.
Figure 4:
Coronal head and neck computed tomography angiography demonstrating a trauma related actionable vascular finding. A 0.5 cm focal contrast collection along the left temporal convexity (yellow arrow), located in the region of acute left subdural hemorrhage, is most consistent with a traumatic pseudoaneurysm in a 96-year-old female.
Further data collection
Additional demographic and clinical patient variables were collected, including patient age, sex, ethnicity, trauma severity designation, transfer status, and prior history of head or neck trauma or surgery. Our institution follows the modified Denver criteria for CTA screenings in the ED, to which adherence was assessed for the patients in the blunt trauma category per the guidelines seen in Table 1.
Table 1:
Modified Denver criteria for head and neck computed tomography angiography (CTAHN) screening in suspected blunt cerebrovascular injury.
| The modified denver criteria |
|---|
|
Statistical analysis
The Stata/MP version 18.0 was used for statistical analysis, and listwise deletion was used to address missing data. Continuous variables, such as age, were presented as means with standard deviations or medians with interquartile ranges, as appropriate, using descriptive statistics to describe the study population. Categorical factors, including gender, trauma severity, patient transfer status, and adherence to the modified Denver criteria, were summarized using frequencies and percentages. Statistical tests such as Chi-square, Fisher’s exact test, t-tests, and analysis of variance were applied as appropriate to draw conclusions and identify correlations regarding the prevalence of low-yield imaging, diagnostic imaging overuse, adherence to guidelines, and diagnostic yield by trauma severity levels across various years. Factors contributing to low-yield or non-actionable CTA findings were examined using a multinomial logistic regression analysis. Potential contributing factors included patient demographics, trauma severity, timing of presentation, and compliance with the modified Denver criteria. To evaluate the degree of correlation between these variables and the likelihood of low-yield imaging, relative risk ratios (RRR) and confidence intervals (CI) were calculated. A significance criterion of p < 0.05 was applied to all two-tailed statistical tests. Assumptions regarding data normality were assessed, and when necessary, appropriate nonparametric methods were utilized. Adjustments for multiple comparisons, including the Bonferroni correction, were used as needed.
RESULTS
An overview of the clinical and demographic features of a sample of 643 adult patients who presented to the ED for blunt trauma and varying CTA results is shown in Table 2. Classification of this cohort revealed 477 negative findings (74.2%), 63 incidental findings (9.8%), 32 non-trauma related actionable findings (5.0%), and 71 trauma related actionable findings (11.0%) [Graph 1]. Of the 71 trauma related actionable findings in this cohort, we found evidence of 24 arterial dissections, 11 post-traumatic aneurysms or pseudoaneurysms, 15 arterial occlusions secondary to thrombus or other etiology, and the remaining cases were other significant findings such as stretch injuries, arteriovenous fistulas, or air emboli. A multinomial logistic regression analysis of the variables linked to CTA findings is shown in Table 3. It contrasts a basic category of negative findings with three categories of CTA findings: Incidental, non-trauma related actionable, and trauma related actionable findings.
Table 2:
Summary statistics and crude differences in clinical and demographic variables by CTA findings among patients presenting for blunt trauma to the emergency department.
| Variable | Full sample (n=643) | Negative (n=477) | Incidental finding (n=63) | Non-trauma related actionable finding (n=32) | Trauma related actionable finding (n=71) | p value |
|---|---|---|---|---|---|---|
| Mean (Standard Deviation) | ||||||
| Age (years) | 46.1 (19.9) | 41.8 (18.1) | 62.2 (16.9) | 69.6 (13.9) | 49.7 (20.9) | 0.001 |
| Frequency (Percentage) | ||||||
| Gender | 0.068 | |||||
| Male | 441 (68.6) | 338 (70.9) | 42 (66.7) | 22 (68.8) | 39 (54.9) | |
| Female | 202 (31.4) | 139 (29.1) | 21 (33.3) | 10 (31.3) | 32 (45.1) | |
| Ethnicity | 0.001 | |||||
| White/Caucasian | 285 (44.3) | 191 (40.0) | 32 (50.8) | 23 (71.9) | 39 (54.9) | |
| Hispanic/Latino | 251 (39.0) | 204 (42.8) | 20 (31.8) | 6 (18.8) | 21 (29.6) | |
| Afro-American | 28 (4.4) | 22 (4.6) | 4 (6.4) | - | 2 (2.8) | |
| Other | 6 (0.9) | 3 (0.6) | - | 2 (6.3) | 1 (1.4) | |
| No reply | 73 (11.4) | 57 (11.9) | 7 (11.1) | 1 (3.0) | 8 (11.3) | |
| Trauma type | 0.001 | |||||
| Fall from any level | 137 (21.3) | 78 (16.4) | 26 (41.3) | 19 (59.4) | 14 (19.7) | |
| Motor vehicle associated accident | 426 (66.3) | 330 (69.2) | 32 (50.8) | 12 (37.5) | 52 (73.2) | |
| Assault with blunt object | 35 (5.4) | 29 (6.1) | 3 (4.8) | - | 3 (4.2) | |
| Other/Unknown | 36 (5.6) | 31 (6.5) | 2 (3.2) | 1 (3.1) | 2 (2.8) | |
| Near hanging or strangulation | 9 (1.4) | 9 (1.8) | - | - | - | |
| Trauma severity | 0.525 | |||||
| Trauma 1 | 312 (48.5) | 235 (49.3) | 26 (41.3) | 11 (33.4) | 40 (56.3) | |
| Trauma 2 | 282 (43.9) | 204 (42.7) | 31 (49.2) | 18 (54.3) | 29 (40.9) | |
| Non-activated trauma | 12 (1.8) | 9 (1.9) | 1 (1.6) | 1 (3.1) | 1 (1.4) | |
| Not recorded | 37 (5.8) | 29 (6.1) | 5 (7.9) | 2 (6.2) | 1 (1.4) | |
| Transfer status | 0.188 | |||||
| Non-transfer | 401 (62.4) | 306 (64.2) | 32 (50.8) | 22 (68.8) | 41 (57.8) | |
| Transfer | 242 (37.6) | 171 (35.8) | 30 (49.2) | 10 (31.2) | 30 (42.2) | |
| Prior Head/Neck surgery or trauma? | 0.134 | |||||
| No | 468 (72.8) | 349 (73.2) | 47 (74.6) | 26 (81.3) | 46 (64.8) | |
| Yes | 35 (5.4) | 21 (4.4) | 6 (9.5) | 3 (9.4) | 5 (7.0) | |
| Unknown | 140 (21.8) | 107 (22.4) | 10 (15.9) | 3 (9.3) | 20 (28.2) | |
| Type CTA Performed | 0.002 | |||||
| Head & Neck | 557 (86.6) | 399 (83.6) | 62 (98.4) | 30 (93.8) | 66 (93.0) | |
| Neck only | 86 (13.4) | 78 (16.4) | 1 (1.6) | 2 (6.2) | 5 (7.0) | |
| Modified Denver criteria followed? | 0.001 | |||||
| No | 285 (44.3) | 212 (44.4) | 30 (71.9) | 23 (71.9) | 20 (28.2) | |
| Yes | 358 (55.7) | 265 (55.6) | 33 (52.4) | 9 (28.1) | 51 (71.8) | |
| Number of points followed on modified Denver criteria | 0.005 | |||||
| One | 308 (47.9) | 230 (48.2) | 27 (42.9) | 10 (31.3) | 41 (57.8) | |
| Two or More | 52 (8.1) | 36 (7.6) | 6 (9.5) | - | 10 (14.1) | |
| Not followed | 283 (44) | 211 (44.2) | 30 (47.6) | 22 (68.7) | 20 (28.1) | |
p<0.05 were considered statistically significant, CTA: Computed tomography angiography.
Graph 1:
Distribution of CTAHN and CTAN findings among adult blunt trauma patients undergoing emergency department screening.
Table 3:
Multinomial logistic regression on CTA findings considering negative findings as the base outcome for patients presenting for blunt trauma to the emergency department.
| Variables with labels | Negative | Incidental Findings | Non trauma related actionable findings | Trauma related actionable findings |
|---|---|---|---|---|
| Relative risk ratio (RRR) with 95% CI | ||||
| Age (years) | Base Outcome | 1.06*** (1.04, 1.07) | 1.08*** (1.05, 1.11) | 1.02*** (1.01, 1.04) |
| Gender | Base Outcome | |||
| Male | Reference Group | |||
| Female | 1.00 (0.53, 1.88) | 0.50 (0.20, 1.27) | 2.13** (1.24, 3.67) | |
| Trauma type | Base Outcome | |||
| Fall from any level | Reference Group | |||
| Motor vehicle associated accident | 0.90 (0.45, 1.81) | 0.66 (0.26, 1.68) | 1.21 (0.57, 2.54) | |
| Assault with blunt object | 1.24 (0.30, 5.07) | 4.12e-07 (NE) | 0.94 (0.23, 3.85) | |
| Other/Unknown | 0.74 (0.15, 3.70) | 0.65 (0.07, 6.35) | 0.66 (0.13, 3.31) | |
| Near hanging or strangulation | 1.87e-07 (NE) | 3.28e-07 (NE) | 1.81e-07 (NE) | |
| Severity | Base Outcome | |||
| Trauma 1 | Reference Group | |||
| Trauma 2 | 1.04 (0.54, 2.00) | 1.03 (0.42, 2.55) | 0.79 (0.44, 1.43) | |
| Non-activated trauma | 0.74 (0.07, 8.07) | 1.70 (0.14, 20.93) | 0.62 (0.07, 5.81) | |
| No information | 1.24 (0.34, 4.13) | 0.79 (0.14, 4.58) | 0.23 (0.03, 1.85) | |
| Prior Head/Neck surgery or trauma? | Base Outcome | |||
| No | Reference Group | |||
| Yes | 1.14 (0.38, 3.40) | 0.74 (0.17, 3.25) | 1.57 (0.52, 4.74) | |
| Unknown | 0.89 (0.38, 2.13) | 0.95 (0.23, 3.86) | 1.28 (0.65, 2.54) | |
| Type CTA performed | Base Outcome | |||
| Head & Neck | Reference Group | |||
| Neck only | 0.07** (0.01, 0.53) | 0.35 (0.07, 1.67) | 0.40 (0.15, 1.06) | |
| Denver criteria (Yes) | Base Outcome | 1.55 (0.01, 208.85) | 0.02 (0.0002, 2.89) | 2.06 (0.09, 45.63) |
| Criteria | Base Outcome | |||
| One followed | Reference Group | |||
| two or more criteria followed | 1.16 (0.40, 3.40) | 3.39e-07 (NE) | 1.26 (0.54, 2.94) | |
| Not followed | 1.70 (0.01, 229.89) | 0.06 (0.0005, 6.70) | 1.10 (0.05, 24.47) | |
Significance level: ***p < 0.001, **p < 0.05. NE: Not estimable due to complete separation, as no patients in this subgroup demonstrated incidental or actionable CTA findings, CTA: Computed tomography angiography, CI: Confidence interval.
Criteria adherence
There were significant differences in the adherence to the modified Denver criteria; patients in the trauma related actionable findings group were more likely to adhere to guidelines during triage (p = 0.008) and meet the modified Denver criteria (p = 0.001). Additionally, meeting ≥2 criteria on the modified Denver criteria was associated with an even higher likelihood of actionable CTA findings (p = 0.005). Formal sensitivity and specificity calculations for the modified Denver criteria were not performed, as all patients in this cohort underwent CTA and no independent gold standard was available as a comparator. However, among patients with trauma related actionable findings, 71.8% met the modified Denver criteria at triage, demonstrating strong association with guideline adherence for clinically significant injury detection. Notably, 28.2% of trauma related actionable findings occurred in patients who did not meet criteria, consistent with prior literature demonstrating that rigid application of screening criteria may miss a subset of BCVI cases.
Age
Older age demonstrated an increased likelihood of incidental and non-trauma related findings. The mean age in this cohort was 46.1 ± 19.9 years, with the incidence of CTA findings being clearly correlated with age. The probability of incidental findings, non-trauma related actionable findings, and trauma related actionable findings rose with increasing age. Older individuals more commonly presented for falls as the mechanism of trauma and were more likely to have non-trauma related actionable findings that required care, likely reflecting age-related comorbid vascular disease or chronic degenerative changes.
Mechanism of trauma
Motor vehicle accidents (MVAs) were the predominant cause across all trauma mechanisms (69.2% overall) and accounted for 73.2% of trauma related actionable findings. In contrast, falls were more often associated with non-trauma related actionable findings (59.4%). Differences in trauma mechanism distribution were statistically significant (p = 0.001); for example, falls were more common in the incidental and non-trauma related actionable findings. Assault with a blunt object and near-hanging or strangulation injuries were rare and did not significantly contribute to outcomes.
Ethnicity
Significant variations were found by ethnicity (p = 0.003). Hispanics/Latinos were more prevalent in the negative group with 42.8% representation in that category. Caucasians/ whites were more prevalent in the non-trauma related actionable findings category. Those of Afro-American ethnicity accounted for a modest percentage of trauma related actionable findings and an intermediate proportion of negative and incidental results.
Other findings
Patients for whom only a CTAN scan was performed were 93% less likely to have incidental findings when compared to those for whom a CTAHN was performed, indicating that incidental findings are more frequent when the intra and extracranial vessels of the head are visualized. Women had a higher likelihood of trauma related actionable findings with twice as much representation than men (RRR = 2.13; 95% CI, 1.24– 3.67; p < 0.05). Trauma severity designation (p = 0.5), transfer status (p = 0.2), and prior head or neck surgery (p = 0.1) were not significant predictors of CTA outcome.
Clinical care point
Age, severity of vascular injury, and vessel injured were key predictors of clinical management. The majority of dissections, pseudoaneurysms, and occlusions were managed conservatively. A common regimen across all etiologies was antiplatelet management (e.g., aspirin ± clopidogrel) or frank anticoagulation (e.g., IV heparin bridged with warfarin ± factor Xa inhibitor if indicated). Beta blockers were used in conjunction with blood pressure elevation significantly increased the risk for mural stress. Therapy duration depended on persistence of vascular insult which was monitored with imaging on both an inpatient and outpatient basis, depending on the patient. Symptom management varied, with acetaminophen and opioids being used in varying degrees. Cases that required neurosurgery and/or neurointerventional radiology to assess the need for intervention encompassed, but were not limited to, cranial vessel rupture, expanding pseudoaneurysm, and basilar hypoperfusion. Age was the most significant predictor of mortality across all variables.
DISCUSSION
BCVI is used to characterize non-penetrating injuries to the carotid and vertebral arteries, as well as other major vessels of the head and neck.[10] BCVI can result from the pathological stretching, compression, and impingement of vessels because of high-force flexion, extension, rotation, or direct impact.[10,11] If unrecognized, BCVI can progress to thromboembolism, vessel occlusion, dissection, pseudoaneurysm formation, and ultimately stroke or death. Although historically considered rare and an uncommon cause of cerebral ischemia and ischemic stroke, emerging studies estimate that BCVI may occur in up to 2% of all hospital trauma admissions and up to 9% of patients with severe traumatic head injuries.[10] Early recognition of injuries and timely intervention remain salient and are crucial in preventing the progression of BCVI to hemorrhage, vessel occlusion, or stroke.[12] The carotid and vertebral arteries are deemed higher risk in developing cerebral ischemic effects compared to other vessels, and thus are often prioritized in initial assessments.[10,13]
CTA utilizes contrast material to produce images of intra and extracranial vasculature, and can play an indispensable role in the diagnosis, management, and prognostication of BCVI.[11] CTAHN is commonly used to assess clinically significant BCVI and a statistically insignificant incidence of strokes and other pathological outcomes have been observed in trauma patients with negative findings.[10] Its lower invasiveness, cost-effectiveness, and greater availability compared to other imaging modalities make CTA a preferred choice in acute traumatic settings. However, despite its advantages in availability and diagnostic yield, its use is not always essential or appropriate, and it may be its ease of access in the ED that has resulted in substantial overutilization.[6,11]
To the best of our knowledge, this is the first study to evaluate CTAHN and CTAN diagnostic yield for BCVI within an adult trauma cohort drawn from a large, predominantly rural catchment area, where triage variability and transfer patterns uniquely influence imaging practices. By delineating how nonadherence to BCVI screening criteria drives imaging overuse, this study reframes diagnostic yield as a quality of care metric rather than a purely radiologic outcome. Moreover, by integrating epidemiologic, operational, and imaging level data, it bridges the gap between radiologic performance metrics and systems-based clinical decision making in trauma care.
Of the 643 blunt trauma cases that met the inclusion criteria, 74.2% of the CTA screenings yielded negative results. This is consistent with previous studies reporting that up to 85% of CTA studies ordered in the ED yield insignificant or negative results.[6] According to our results, age, trauma type, and modified Denver criteria adherence emerged as the strongest predictors of diagnostic yield.
Our findings indicate that a significant (71.8%) percentage of patients with trauma related actionable findings met the modified Denver criteria (p < 0.005). Furthermore, meeting two or more criteria on the modified Denver criteria was associated with an even higher yield of trauma related actionable findings. These results suggest that guideline-concordant CTA use continues to provide meaningful diagnostic discrimination, even within heterogeneous trauma populations consistent with prior studies.[14] However, of the total patients in the cohort that met the modified Denver criteria (n = 358), only 14.2% (n = 51) were categorized as trauma related actionable findings, suggesting that improved adherence and targeted refinement of existing screening practices, rather than expansion, restriction, or neglect of current criteria, may best enhance utilization efficiency without compromising diagnostic safety.
In this study, not all patients (28.2%) with trauma related actionable findings met the modified Denver criteria during triage. The presence of trauma related actionable findings in patients who did not meet the modified Denver criteria likely reflects atypical injury mechanisms, delayed symptom evolution, or limitations inherent to criteria-based screening tools. For example, prior studies have demonstrated that BCVI may occur in the setting of low-mechanism trauma, suggesting the role of clinical judgment as a necessary adjunct to guideline-based screening.[8] Importantly, however, no increased morbidity attributable to traumatic vascular pathology was observed among patients with trauma related actionable findings who did not meet the modified Denver criteria in this cohort.
Previous studies have found that MVA accounts for nearly 50% of BCVI in acute traumatic settings. Our study agrees with these findings, as a larger proportion (73.2%) of patients in this study with trauma related actionable findings were victims of a motor vehicle associated accident. Conversely, it has been shown that lower mechanisms of injury also have a high contribution to the incidence of BCVI. Harper et al. wrote that no association exists between mechanisms of injury in patients who suffer BCVI and meet criteria for CTAHN or CTAN versus those who do not. Thus, low mechanisms of injury and little to no identifiable risk factors may demonstrate low utility in assessing patients for whom CTA screening is warranted. This may be especially pertinent when screening elderly patients in the ED.[8,14]
A greater portion of elderly patients fell into the incidental and non-trauma related actionable findings categories. Age has been documented as an independent and essential risk factor in the decline of vascular function. Vascular density in the cortex decreases with age, which may be congruent with and a causative factor in cortical atrophy and degeneration. The aged brain is more likely to exhibit arteriolar tortuosity, in the form of spirals, loops, and coils and can demonstrate unfavorable responses in counteracting stress and damage, with a propensity for hypertrophic, eutrophic, and hypotrophic vascular remodeling. Thus, vascular structures in an aged brain have a decreased caliber and intrinsic capability to counteract degenerative changes and trauma.[15] A significant portion of the elderly patients represented in this study were indicated for CTA screening as a result of falls, most of which were from ground level. This supports findings from Harper et al., who reported that elderly patients with ground-level falls have an elevated risk of BCVI despite low impact mechanism of injury.[8]
In this study, female sex was independently associated with higher rates of trauma related actionable findings compared to their male counterparts, a finding not widely reported in previous studies. Magnotti et al. found that gender is not independently associated with mortality after blunt trauma.[16] Whether this is due to a true biological phenomenon or sampling artifact unique to this study may require further investigation.
Our study revealed no significant association between a patient’s transfer status and CTA findings. Transfer status at a level 1 trauma center has been described as an associated factor for repeat imaging since prior imaging is not always sent with patients and is often of poor quality.[17] Our findings indicate that transfer status is a poor predictor of vascular injury risk and contributes to overutilization. Due to the retrospective and single-center nature of this study, we were unable to determine what percentage of transfer patients received duplicate imaging.
Trauma severity designation and previous head or neck surgery or trauma also showed no significant impact on variations in outcome and should thus not be considered in determining whether CTA screening is appropriate. These findings suggest that reliance on these factors does not enhance BCVI detection or diagnostic yield accuracy and should not be considered among other measures.
Previous studies have reported that nonphysician practitioners (NPPs) and ED residents order a significant proportion of CTA exams with non-trauma related findings or negative results. More experienced providers, e.g., ED attendings, have been found to order studies with relatively lower rates of negative or insignificant yield.[18] This study found no such correlation, as all neuroimaging studies ordered by residents or NPPs at our institution are overseen by attending physicians.
Defensive medicine, staff inexperience, and operational constraints, such as time pressures and overcrowding, are among the factors that drive diagnostic imaging overutilization, especially in the ED. Our findings of high nonadherence to the modified Denver criteria and the prevalence of negative or incidental results in the setting of trauma demonstrate that easy access to imaging without strict screening measures can exacerbate overuse.[19-21] Given the significant associations our study identified between patient age, mechanism of trauma, number of points met on the modified Denver criteria and CTA yield, these factors should be integrated into clinical decision making in the emergency setting to improve screening practices, minimize unnecessary imaging, and improve patient care efficiency.
This study has several limitations. The retrospective design introduces potential for information bias, particularly in the assessment of criteria adherence, which relied on documentation available in the electronic medical record. Because this study was a single-center analysis conducted on a largely rural population, generalizability to other settings such as urban academic institutions may be limited. While we included transfer patients, the absence of standardized access to outside imaging or referring documentation may have influenced repeat imaging rates or assessments on CTA appropriateness. Although adherence to the modified Denver criteria was systematically assessed, the retrospective nature of this study may not fully reflect real-time clinical judgment, triage constraints, or other nuances in the emergency setting.
CONCLUSION
Our findings suggest that nonadherence to screening guidelines is a modifiable factor contributing to the overuse of CTAHN imaging in the ED. Despite our institution’s designation as a Level 1 trauma center serving a large rural and medically underserved population, it shares similar constraints in diagnostic efficacy and driving variables in CTA overutilization as urban counterparts. Implementing interventions such as enhanced training for emergency providers, improved access to prior imaging, and stricter adherence to established clinical criteria could help mitigate diagnostic imaging overuse. These measures would not only reduce unnecessary imaging but would also maintain diagnostic accuracy and enhance patient care.
Funding Statement
Financial support and sponsorship: Nil.
Footnotes
How to cite this article: Nagy KK, Orton JB, Awal A, Jacob R, Nagy L. Diagnostic yield and utilization patterns of head and neck computed tomography angiography. J Clin Imaging Sci. 2026;16:24. doi: 10.25259/ JCIS_275_2025
Ethical approval:
The research/study was approved by the Institutional Review Board at Texas Tech University Health Sciences Center, number IRB-FY2024-232, dated September 05, 2024.
Declaration of patient consent:
The authors certify that they have obtained all appropriate patient consent forms. In the form, the patients have given their consent for their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
Conflicts of interest:
There are no conflicts of interest.
Use of artificial intelligence (AI)-assisted technology for manuscript preparation:
The authors confirm that there was no use of artificial intelligence (AI)-assisted technology for assisting in the writing or editing of the manuscript and no images were manipulated using AI.
References
- 1.Müskens JL, Kool RB, Van Dulmen SA, Westert GP. Overuse of diagnostic testing in healthcare: A systematic review. BMJ Qual Saf. 2022;31:54–63. doi: 10.1136/bmjqs-2020-012576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Decker JH, Mazal AT, Bui A, Sprenger T, Skare S, Fischbein N, et al. NeuroMix with MRA: A fast MR protocol to reduce head and neck CTA for patients with acute neurologic presentations. AJNR Am J Neuroradiol. 2024;45:1730–6. doi: 10.3174/ajnr.A8386. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Tu LH, Malhotra A, Venkatesh AK, Taylor RA, Sheth KN, Forman HP, et al. Head and neck CTA utilization: Analysis of ordering frequency and nonroutine results communication, with focus on the 50 most common emergency department clinical presentations. AJR Am J Roentgenol. 2022;218:544–51. doi: 10.2214/AJR.21.26543. [DOI] [PubMed] [Google Scholar]
- 4.ElHabr A, Merdan S, Ayer T, Prater A, Hanna TN, Horny M, et al. Increasing utilization of emergency department neuroimaging from 2007 through 2017. AJR Am J Roentgenol. 2022;218:165–73. doi: 10.2214/AJR.21.25864. [DOI] [PubMed] [Google Scholar]
- 5.Prabhakar AM, Gottumukkala RV, Hemingway J, Hughes DR, Patel SS, Duszak R., Jr Increasing utilization of emergency department neuroimaging in medicare beneficiaries from 1994 to 2015. Am J Emerg Med. 2018;36:680–3. doi: 10.1016/j.ajem.2017.12.057. [DOI] [PubMed] [Google Scholar]
- 6.Mehan WA, Jr, Shin D, Buch K. Effect of provider type on overutilization of CT angiograms of the head and neck for patients presenting to the emergency department with nonfocal neurologic symptoms. J Am Coll Radiol. 2024;21:890–5. doi: 10.1016/j.jacr.2023.08.042. [DOI] [PubMed] [Google Scholar]
- 7.Poyiadji N, Beauchamp N, 3rd, Myers DT, Krupp S, Griffith B. Diagnostic imaging utilization in the emergency department: Recent trends in volume and radiology work relative value units. J Am Coll Radiol. 2023;20:1207–4. doi: 10.1016/j.jacr.2023.06.033. [DOI] [PubMed] [Google Scholar]
- 8.Harper PR, Jacobson LE, Sheff Z, Williams JM, Rodgers RB. Routine CTA screening identifies blunt cerebrovascular injuries missed by clinical risk factors. Trauma Surg Acute Care Open. 2022;7:e000924. doi: 10.1136/tsaco-2022-000924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ebrahimzadeh SA, Manzoor K, Edlow JA, Selim M, Chang YM, Bhadelia RA, et al. Diagnostic yield of CT angiography performed for suspected cervical artery dissection in the emergency department. Emerg Radiol. 2022;29:825–32. doi: 10.1007/s10140-022-02065-x. [DOI] [PubMed] [Google Scholar]
- 10.Brommeland T, Helseth E, Aarhus M, Moen KG, Dyrskog S, Bergholt B, et al. Best practice guidelines for blunt cerebrovascular injury (BCVI) Scand J Trauma Resusc Emerg Med. 2018;26:90. doi: 10.1186/s13049-018-0559-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kik CC, Slooff WM, Moayeri N, De Jong PA, Muijs SP, Öner FC. Diagnostic accuracy of computed tomography angiography (CTA) for diagnosing blunt cerebrovascular injury in trauma patients: A systematic review and meta-analysis. Eur Radiol. 2022;32:2727–38. doi: 10.1007/s00330-021-08379-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Patel J, Huynh TJ, Rao D, Brzezicki G. Vascular trauma in the head and neck and endovascular neurointerventional management. J Clin Imaging Sci. 2020;10:44. doi: 10.25259/JCIS_96_2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Biffl WL, Moore EE, Ryu RK, Offner PJ, Novak Z, Coldwell DM, et al. The unrecognized epidemic of blunt carotid arterial injuries: Early diagnosis improves neurologic outcome. Ann Surg. 1998;228:462–70. doi: 10.1097/00000658-199810000-00003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wagner MJ, Hussein I, Low G, Samji KB. Comparing the Denver criteria sets for blunt trauma: A retrospective study of cases in Edmonton, Alberta. Br J Radiol. 2023;96:20221116. doi: 10.1259/bjr.20221116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Xu X, Wang B, Ren C, Hu J, Greenberg DA, Chen T, et al. Age-related Impairment of Vascular Structure and Functions. Aging Dis. 2017;8:590–610. doi: 10.14336/AD.2017.0430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Magnotti LJ, Fischer PE, Zarzaur BL, Fabian TC, Croce MA. Impact of gender on outcomes after blunt injury: A definitive analysis of more than 36,000 trauma patients. J Am Coll Surg. 2008;206:984–91. doi: 10.1016/j.jamcollsurg.2007.12.038. discussion 991-2. [DOI] [PubMed] [Google Scholar]
- 17.Tung M, Sharma R, Hinson JS, Nothelle S, Pannikottu J, Segal JB. Factors associated with imaging overuse in the emergency department: A systematic review. Am J Emerg Med. 2018;36:301–9. doi: 10.1016/j.ajem.2017.10.049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Christensen EW, Liu CM, Duszak R, Jr, Hirsch JA, Swan TL, Rula EY. Association of state share of nonphysician practitioners with diagnostic imaging ordering among emergency department visits for medicare beneficiaries. JAMA Netw Open. 2022;5:e2241297. doi: 10.1001/jamanetworkopen.2022.41297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bruls RJ, Kwee RM. Workload for radiologists during on-call hours: Dramatic increase in the past 15 years. Insights Imaging. 2020;11:121. doi: 10.1186/s13244-020-00925-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kocher KE, Meurer WJ, Fazel R, Scott PA, Krumholz HM, Nallamothu BK. National trends in use of computed tomography in the emergency department. Ann Emerg Med. 2011;58:452–62.e3. doi: 10.1016/j.annemergmed.2011.05.020. [DOI] [PubMed] [Google Scholar]
- 21.Kwee RM, Toxopeus R, Kwee TC. Imaging overuse in the emergency department: The view of radiologists and emergency physicians. Eur J Radiol. 2024;176:111536. doi: 10.1016/j.ejrad.2024.111536. [DOI] [PubMed] [Google Scholar]