Blunt cerebrovascular injury (BCVI) occupies an uncomfortable position in trauma practice. It is usually silent at presentation. It is invisible on the primary survey. Its natural history - an embolic or thrombotic stroke hours to days after an injury that looked survivable and unremarkable - is exactly the category of preventable catastrophe that organised trauma systems exist to abolish. The remedy is neither technically demanding nor expensive: a computed tomography angiogram (CTA) of the neck, followed in most cases by a single antithrombotic agent. The entire controversy is about who gets the scan.
For a quarter of a century the answer has been “patients who meet criteria”. The Denver criteria, their expanded successor, and the Memphis criteria all rest on the same premise: that clinical and radiological features available at the bedside can reliably select the patients worth imaging. Advanced Trauma Life Support has followed the same logic, and the 11th edition of the course manual, released in 2025, sets out a noticeably longer list of triggers than its predecessors did.
Two publications in the past week have brought the underlying tension into sharp relief. On 20 July, Emergency Radiology published a single-centre analysis of universal neck CTA screening in which a third of the injuries detected occurred in patients with no expanded Denver risk factor at all - while incidental abnormalities were found in roughly one in seven patients scanned, more than a third of whom were discharged with nothing in the record to suggest anyone had told them. The following day, the Journal of Trauma and Acute Care Surgery published a review of contemporary carotid artery injury management identifying expanded BCVI screening and earlier antithrombotic therapy as two of the principal forces reshaping practice.
Read together, the two papers frame the question this article addresses. The criteria-based approach that ATLS teaches is demonstrably leaky. The alternative carries costs that enthusiasm for universal imaging has been slow to quantify.
A disease that punishes delay - and rewards detection
BCVI describes non-penetrating injury to the carotid or vertebral arteries: intimal tear, dissection, intramural haematoma, pseudoaneurysm, occlusion or, at the extreme, transection. Historical series quoted an incidence of 0.1–3% of blunt trauma admissions. Contemporary series that image liberally consistently report figures towards or beyond the top of that range, which tells us something important before any argument about screening thresholds begins: the older estimates were partly an artefact of not looking.
The reason detection matters is the therapeutic window. In the classic Denver series, of 107 initially asymptomatic BCVIs left untreated, 23 patients (21.5%) went on to sustain a cerebrovascular accident, at a mean of 58 hours from injury. Among 282 asymptomatic injuries that were treated, one patient (0.5%) did. Other cohorts frame the same contrast as roughly a 25% stroke risk untreated against under 8% treated. Stroke risk climbs steeply with injury grade: unilateral vertebral artery occlusion carries a reported 9-20% risk, unilateral internal carotid occlusion in excess of 50%.
That 58-hour mean is the single most important number in the field. It describes a latent interval measured in days rather than minutes - wide enough that a diagnosis made in the emergency department genuinely alters the outcome, and wide enough that a diagnosis missed there is very unlikely to be made in time by anything other than the stroke itself. The move from digital subtraction angiography to CTA compressed the diagnostic interval dramatically: one influential series reported mean time from admission to diagnosis falling from 31.2 hours to 2.65 hours, with the stroke rate falling from 15.2% to 3.8% alongside it.
In Wilson and Jungner’s terms, BCVI is close to a textbook screening candidate: an important condition, a recognisable latent stage, an acceptable and accurate test, an available treatment, and a reasonably agreed policy on whom to treat. The live question is narrower and more interesting. Does case-finding by risk factor actually work?
The criteria keep missing patients - consistently, and by a lot
The most instructive studies are those from centres that adopted universal screening and then retrospectively asked how the published criteria would have performed. The answers have been unflattering and remarkably reproducible.
The University of Alabama at Birmingham reported 6,800 blunt trauma patients, of whom 6,287 (92.5%) underwent neck CTA; 480 (7.6%) had CTA evidence of BCVI. The expanded Denver criteria achieved the highest sensitivity at 74.7%, with a positive predictive value of 14.6%. The original Denver criteria managed 57.5% sensitivity and the Memphis criteria 47.3%. Applied prospectively, those three rule sets would have left, respectively, 42.5%, 25.3% and 52.7% of the injuries detected by universal screening without a neck CTA at all.
A Virginia Commonwealth University cohort of 4,659 blunt trauma activations found 126 patients (2.7%) with 158 injuries. Against that reference standard, the American College of Surgeons Trauma Quality Improvement Program imaging criteria returned a sensitivity of 72.2%, a specificity of 64.9%, a positive predictive value of 6.8% and an overall accuracy of 65.2%. It is worth pausing on the specificity: these criteria are not merely insensitive, they are also poor at ruling injury in, which undercuts the usual defence that a low-yield rule is at least a precise one.
European data tell the same story in a different health system. Among 4,104 patients with suspected major trauma investigated with liberal CTA, 91 (2.2%) had 126 BCVIs, and the sensitivities of the various clinical screening criteria ranged from 57% to 84%. Crucially, the false negatives clustered anatomically - in the petrous segment of the carotid and in vertebral artery pseudoaneurysms. These are not radiological curiosities but lesions with genuine embolic potential, in segments that are difficult to infer clinically.
The systematic review published in Emergency Radiology on 8 June 2026 pooled eight comparative studies and put a range on the shortfall: selective criteria missed between 16% and 37% of BCVI cases relative to universal CTA. Stroke incidence across the included cohorts ran at 8-10%, with some events occurring in patients who carried no traditional risk factor. Imaging-related harm was low, with acute kidney injury around 1.4%, and cost-effectiveness analyses placed universal CTA below conventional willingness-to-pay thresholds at approximately $71,949 per quality-adjusted life year.
What ATLS 11 settles, and what it leaves open
The 11th edition of the ATLS manual widened the screening net appreciably. Beyond the expanded Denver criteria, it lists head or cervical CT performed because of the injury mechanism; traumatic brain injury with thoracic injuries; oculosympathetic palsy or Horner’s syndrome; skull fracture involving the petrous temporal bone, carotid canal, skull base, frontal bone, sinuses or orbit; neck injury, seatbelt sign or crepitus; and blunt cardiac rupture.
The first of those criteria is doing an enormous amount of work. In a mature trauma system where most blunt polytrauma patients receive whole-body CT, “head or cervical CT performed because of mechanism” approaches universal screening by stealth. For the patient already heading to the scanner, the practical gap between ATLS 11 and a formal universal protocol may be much narrower than the two documents make it appear.
But it is not zero, and the residual population is not a rounding error. The older adult after a ground-level fall who receives CT of the cervical spine but not the head, or the patient investigated for an isolated injury with no mechanism-based imaging, is precisely who the newest data identify as at risk of a missed lesion - and ground-level falls in older adults are the fastest-growing mechanism in most Western trauma registries, so the gap is widening rather than closing.
The point for examination candidates is straightforward: ATLS has widened the net, not abandoned the concept. The manual remains criteria-based, and answers to ATLS questions should reflect that.
The bill for imaging everyone: last week’s inconvenient paper
The study published on 20 July examined Priority 1 and 2 adult blunt trauma patients presenting to a regional Level 1 trauma centre between April 2023 and April 2024. Of 1,832 eligible patients, 1,554 (84.8%) underwent neck CTA under a universal screening protocol. Fifty-eight (3.73%) had CTA evidence of BCVI - and 20 of those 58 (34.5%) had no risk factor under the expanded Denver criteria.
That figure independently corroborates everything above. But the study’s primary purpose was to count what else the scans found. There were 265 incidental findings in 224 patients - 14.55% of those screened. They included 139 moderate-to-severe stenoses, 27 chronic arterial occlusions, 26 significant thyroid nodules and 9 aneurysms. Eighty-one of those patients (36.16%) were discharged with no evidence in their records that they had ever been told.
Two things follow. The first is arithmetical: at this centre, universal screening generated a downstream diagnostic workload roughly four times larger than the target condition itself. The second is a governance failure that enthusiasm for sensitivity has largely obscured. A significant carotid stenosis in a 68-year-old who fell off a ladder is either a stroke prevented or a cascade of surveillance, medication and anxiety, and which it turns out to be depends entirely on whether the follow-up pathway exists.
The other costs are more modest but real. Contrast-associated acute kidney injury runs at around 1.4%. Radiation is a minor consideration in adults but a substantial one in children, which is why paediatric practice has moved in the opposite direction, towards validated selective instruments such as the McGovern and Utah scores. And in systems already constrained by reporting capacity - the NHS prominently among them - adding an arterial-phase neck acquisition to every blunt trauma pathway is not a cost-free protocol change. It is a claim on radiologist time that has to come from somewhere.
Screening is only as good as what follows it
A screening programme inherits the uncertainty of its treatment, and here the evidence is softer than the diagnostic literature tends to imply.
That antithrombotic therapy beats no therapy is not seriously disputed - the 21.5% versus 0.5% contrast in the Denver data is about as stark as observational trauma evidence gets. Which agent remains genuinely unresolved. Aspirin and heparin appear broadly equivalent for stroke prevention; a pooled analysis of over 6,500 patients reported a lower stroke risk with antiplatelet therapy (odds ratio 0.57, 95% CI 0.33–0.96), and most syntheses report fewer bleeding complications with antiplatelets. Dose remains unsettled, with low-dose aspirin reported as a viable alternative to the historical 325 mg and heparin regimens. Duration is unsettled. Whether asymptomatic grade I lesions - many of which resolve spontaneously - require treatment at all is unsettled, and the value of routinely re-imaging them has been challenged on cost-effectiveness grounds.
The Eastern Association for the Surgery of Trauma’s 2020 practice management guideline recommends using a screening protocol in blunt polytrauma and recommends against routine endovascular stenting for grade II and III injuries, on the basis that in-stent thrombosis and the requirement for prolonged antithrombotic therapy outweigh any benefit over medical management alone.
The honest formulation of the counterargument is therefore this: universal screening will preferentially find low-grade lesions, and low-grade lesions are the ones where the benefit of treatment is least certain. Advocates are on their strongest ground when they point out that universal screening improves detection of high-grade injuries too, not merely grade I counts - the Birmingham data support that directly. But the marginal yield of moving from expanded criteria to universal imaging is inevitably weighted towards the injuries we understand least well.
A defensible middle position
The strongest practical argument for near-universal screening is marginal rather than absolute. In a patient already undergoing contrast-enhanced whole-body CT, extending arterial-phase acquisition to the neck adds little scanner time, little additional contrast and little additional radiation. In that population, the criteria-based approach becomes difficult to defend: it withholds an almost free test on the basis of rules with sensitivities somewhere between 47% and 75%.
Outside that population the calculus genuinely changes. The older patient with an isolated low-energy cervical fracture after a ground-level fall is not going to whole-body CT, and for them a neck CTA is a discrete decision with discrete costs. This is where criteria still earn their keep - and where better criteria, rather than no criteria, looks like the more promising research direction. Machine-learning-derived risk models trained on registry data may eventually outperform both the expanded Denver rule and a blanket protocol, though nothing of that kind has yet been prospectively validated.
Whichever policy a unit adopts, last week’s paper carries a lesson that is procedural rather than diagnostic. A screening programme without an incidental-findings pathway is an incomplete programme. Any department contemplating a move to universal CTA should build the communication and follow-up mechanism before it changes the imaging protocol, not after the first audit discovers that a third of the findings went nowhere.
Relevance to the ATLS examination
BCVI is a reliable source of ATLS written and viva material precisely because it sits at the junction of anatomy, mechanism and protocol. Candidates should be secure on the following:
• The expanded Denver criteria, together with the additional triggers listed in the 11th edition of the manual. These are directly testable list items and are among the most commonly examined content in the head and neck material.
• The Biffl grading scale (I–V) and how grade maps to management - including that grade V injuries demand immediate intervention while grades II and III are not routinely stented.
• The temporal pattern: asymptomatic at presentation, with stroke typically occurring within the first 72 hours. This is the entire rationale for screening rather than symptom-driven imaging.
• That CTA has replaced digital subtraction angiography as the first-line modality, and the reasons - comparable accuracy, far shorter time to diagnosis, and avoidance of catheter-related complications.
• The principle that antithrombotic therapy substantially reduces stroke risk, that antiplatelet and anticoagulant strategies are broadly equivalent, and that early therapy has been shown to be reasonably safe even in the presence of traumatic brain injury or solid organ injury.
• A point of exam technique: if a question asks what ATLS recommends, answer with the criteria-based approach as the manual sets it out. The universal-screening argument belongs in the discussion that follows, not in the answer itself - and candidates who can articulate the sensitivity-versus-resource trade-off when invited to discuss it tend to score well.
Conclusion
The BCVI screening debate illustrates how trauma evidence usually moves: not by a decisive randomised trial but by an accumulation of single-centre series that make the incumbent position progressively harder to defend. No randomised comparison of universal against selective CTA screening exists, and it is improbable that one ever will - the outcome is rare, the intervention is cheap, and equipoise is quietly eroding.
What the past six weeks of publication have added is symmetry. The June systematic review strengthened the case that criteria miss injuries and that universal imaging is cost-effective. The July incidental-findings study is a reminder that “cost-effective” and “consequence-free” are different claims, and that some of the costs of screening are borne by patients who never had the disease being screened for.
For the ATLS candidate, the examinable content remains the criteria. For the clinician, the more useful question is not whether to screen everyone, but whether your department knows what it will do with everything it finds.
References
1. Manka M, Kelleran K, Manka M, Drumsta D. Incidental findings identified by universal screening for BCVI using CT angiography. Emerg Radiol. 2026 Jul 20 [online ahead of print]. doi:10.1007/s10140-026-02517-8. https://doi.org/10.1007/s10140-026-02517-8
2. Othman Y, Nasef Y, Jayagopi K, Singh P, Prashar S, Elkbuli A. Universal computed tomography angiography screening for blunt cerebrovascular injury: a systematic review of improved detection, clinical outcomes, and cost-effectiveness. Emerg Radiol. 2026 Jun 8 [online ahead of print]. doi:10.1007/s10140-026-02495-x. https://doi.org/10.1007/s10140-026-02495-x
3. Carotid artery injury: contemporary evidence and best practices in management [review]. J Trauma Acute Care Surg. 2026;101(1):154–161. Published online 21 July 2026. https://journals.lww.com/jtrauma/pages/default.aspx
4. Black JA, Abraham PJ, Abraham MN, et al. Universal screening for blunt cerebrovascular injury. J Trauma Acute Care Surg. 2021;90(2):224–231. doi:10.1097/TA.0000000000003010. https://doi.org/10.1097/TA.0000000000003010
5. Leichtle SW, Banerjee D, Schrader R, et al. Blunt cerebrovascular injury: the case for universal screening. J Trauma Acute Care Surg. 2020;89(5):880–886. doi:10.1097/TA.0000000000002824. https://doi.org/10.1097/TA.0000000000002824
6. Muther M, Sporns PB, Hanning U, et al. Diagnostic accuracy of different clinical screening criteria for blunt cerebrovascular injuries compared with liberal state of the art computed tomography angiography in major trauma. J Trauma Acute Care Surg. 2020. PMID: 32195997. https://pubmed.ncbi.nlm.nih.gov/32195997/
7. Kim DY, Biffl W, Bokhari F, et al. Evaluation and management of blunt cerebrovascular injury: a practice management guideline from the Eastern Association for the Surgery of Trauma. J Trauma Acute Care Surg. 2020;88(6):875–887. doi:10.1097/TA.0000000000002668. https://doi.org/10.1097/TA.0000000000002668
8. Cothren CC, Biffl WL, Moore EE, Kashuk JL, Johnson JL. Treatment for blunt cerebrovascular injuries: equivalence of anticoagulation and antiplatelet agents. Arch Surg. 2009;144(7):685–690. https://pubmed.ncbi.nlm.nih.gov/19620551/
9. Eastman AL, Muraliraj V, Sperry JL, Minei JP. CTA-based screening reduces time to diagnosis and stroke rate in blunt cervical vascular injury. J Trauma. 2009;67(3):551–556. https://pubmed.ncbi.nlm.nih.gov/19741399/
10. Biffl WL, Moore EE, Offner PJ, Brega KE, Franciose RJ, Burch JM. Blunt carotid arterial injuries: implications of a new grading scale. J Trauma. 1999;47(5):845–853. doi:10.1097/00005373-199911000-00004. https://doi.org/10.1097/00005373-199911000-00004
11. Burlew CC, Biffl WL, Moore EE, Barnett CC, Johnson JL, Bensard DD. Blunt cerebrovascular injuries: redefining screening criteria in the era of noninvasive diagnosis. J Trauma Acute Care Surg. 2012;72(2):330–337. doi:10.1097/TA.0b013e31823de8a0. https://doi.org/10.1097/TA.0b013e31823de8a0
12. 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(1):e000924. doi:10.1136/tsaco-2022-000924. https://doi.org/10.1136/tsaco-2022-000924
13. Schmidt JC, Huang DD, Fleming AM, et al. Missed blunt cerebrovascular injuries using current screening criteria — the time for liberalized screening is now. Injury. 2023;54(5):1342–1348. doi:10.1016/j.injury.2023.02.019. https://doi.org/10.1016/j.injury.2023.02.019
14. Ali A, Broome JM, Tatum D, et al. Cost-effectiveness of universal screening for blunt cerebrovascular injury: a Markov analysis. J Am Coll Surg. 2023;236(3):468–475. doi:10.1097/XCS.0000000000000490. https://doi.org/10.1097/XCS.0000000000000490
15. Kik CC, Slooff WBM, Moayeri N, de Jong PA, Muijs SPJ, 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(4):2727–2738. doi:10.1007/s00330-021-08379-7. https://doi.org/10.1007/s00330-021-08379-7
16. Brommeland T, Helseth E, Aarhus M, et al. Best practice guidelines for blunt cerebrovascular injury (BCVI). Scand J Trauma Resusc Emerg Med. 2018;26(1):90. doi:10.1186/s13049-018-0559-1. https://doi.org/10.1186/s13049-018-0559-1
17. Biffl WL, Cothren CC, Moore EE, et al. Western Trauma Association critical decisions in trauma: screening for and treatment of blunt cerebrovascular injuries. J Trauma. 2009;67(6):1150–1153. doi:10.1097/TA.0b013e3181c1c1d6. https://doi.org/10.1097/TA.0b013e3181c1c1d6
18. Susai CJ, Alcasid NJ, Banks KC, et al. Is repeat computed tomography angiography for asymptomatic grade 1 blunt cerebrovascular injuries cost-effective? J Surg Res. 2024;303:409–414. doi:10.1016/j.jss.2024.09.015. https://doi.org/10.1016/j.jss.2024.09.015
19. Venkataraman SS, Herbert JP, Ravindra VM, et al. Multi-center validation of the McGovern Pediatric Blunt Cerebrovascular Injury Screening Score. J Neurotrauma. 2023;40(13–14):1451–1458. https://pubmed.ncbi.nlm.nih.gov/36517974/
20. Choice of antiplatelet vs. anticoagulant for blunt cerebrovascular injury. Life (Basel). 2025;15(9):1364. doi:10.3390/life15091364. https://doi.org/10.3390/life15091364
21. Advanced Trauma Life Support (ATLS) Course Manual. 11th ed. Chicago, IL: American College of Surgeons; 2025. https://www.facs.org/quality-programs/trauma/education/advanced-trauma-life-support/atls-11/
