The established knowledge
If you ask a trauma practitioner where pelvic fracture bleeding comes from you're likely to get the same answer: mostly venous. The figure varies with the source, but it is always large. The World Society of Emergency Surgery's 2017 pelvic trauma guidelines put it at 80% venous and 20% arterial [1]; a widely used StatPearls chapter says up to 85% is venous, most commonly from the presacral plexus and prevesical veins, with arterial bleeding in only 15–20% [2]. The number does real work. It underwrites the logic of the pelvic binder, the case for preperitoneal packing, and the familiar instinct that pelvic bleeding is a low-pressure ooze to be squeezed rather than a vessel to be found.
On 15 October 2026, a team from Leipzig University Hospital published a study in the European Journal of Trauma and Emergency Surgery that sets out to re-evaluate that paradigm [3]. Its most striking finding is that isolated venous bleeding was the least common source of pelvic haemorrhage they saw. The paper is open access, and it deserves to be read by anyone who teaches, or is about to be examined on, the C of the primary survey.
This post looks at what the study found, why the venous dogma may be less wrong than the headline suggests, and what, if anything, should change in the resuscitation room and in your exam answers.
What Leipzig found
Gerhardt, Höch and colleagues reviewed 512 consecutive patients with pelvic fractures admitted to their level I trauma centre between January 2020 and December 2022, all of whom had contrast-enhanced CT at presentation [3]. Two experienced pelvic surgeons independently reviewed every scan. Haemorrhage counted as "radiologically relevant" if there was active contrast extravasation, a contained intrapelvic haematoma, or asymmetric soft-tissue displacement suggesting one. Where bleeding was present, the team assigned a presumed predominant source (arterial, venous, cancellous bone or mixed) from the pattern and timing of extravasation and the haematoma's relationship to vessels and fracture surfaces.
Four findings stand out.
First, pelvic bleeding was less common than the textbooks imply. Only 95 of 512 patients (18.6%) had radiologically relevant haemorrhage [3].
Second, the sources were not what the 85% rule predicts. Among the 95, the presumed source was cancellous bone in 44.2%, mixed in 29.5%, arterial in 18.9%, and isolated venous in just 7.4%. Bleeding mapped mostly to iliac sites; presacral bleeding, the classic venous plexus story, was uncommon [3].
Third, the population was old. Mean age across the cohort was 68, and nearly half the fractures (47.3%) were fragility fractures of the pelvis from low-energy mechanisms [3]. Bleeding was more frequent after high-energy trauma (73% of bleeders versus 48% of non-bleeders) and in men, but age did not differ between the groups. The authors also report that arterial bleeding was seen more often in older patients after low-energy injury.
Fourth, bleeding mattered. Mortality was 14.7% in patients with pelvic haemorrhage and 5.3% in those without [3].
The authors conclude that the traditional concept of predominantly venous pelvic haemorrhage "may not fully reflect" contemporary CT findings, and argue for early CT-based localisation to guide individualised haemorrhage control.
Where the venous number came from
It is worth asking how the 80–90% figure became so durable. The Leipzig authors trace it largely to postmortem and early surgical series, many of which predate modern multidetector CT and drew on younger, higher-energy populations [3]. Those studies relied on what could be seen at operation or autopsy, and a surgeon packing a pelvis sees a diffuse, welling retroperitoneal ooze, not a labelled source.
There is also a selection problem running the other way. Much of the literature on arterial bleeding comes from angiography series, and patients who reach the angiography suite are, by definition, those already suspected of having an arterial source. The Leipzig team note that such studies may overestimate the clinically relevant arterial share [3]. Neither tradition was ever sampling the whole population of injured pelvises.
Steelmanning the old teaching
Before declaring the presacral plexus innocent, it is worth making the strongest case for the textbook view. There are at least four good arguments.
The first is taxonomic. Many statements of the classical teaching do not cleanly separate veins from bone; they describe pelvic bleeding as predominantly low-pressure bleeding from the venous plexus and fracture surfaces, as opposed to arterial bleeding. Add Leipzig's cancellous (44.2%), mixed (29.5%) and venous (7.4%) categories together and you get roughly 81% of bleeding that is wholly or partly non-arterial, against 18.9% arterial. That is strikingly close to the 80/20 split in the WSES guidance [1]. On this reading, the study does not overturn the paradigm; it re-describes its largest compartment. The bleeding is still mostly low-pressure. It is simply coming from bone rather than from named veins.
The second argument is technical. Contrast-enhanced CT is good at showing arterial extravasation and relatively poor at showing venous bleeding, which may need delayed phases and is often indistinguishable from diffuse haematoma. Hussami and colleagues compared clinical CT in 49 polytrauma patients with multiphase postmortem CT angiography in 45 bodies with severe pelvic injury: the postmortem studies showed venous bleeding in 78% and arterial bleeding in 73%, and revealed more venous lesions than clinical CT had [4]. If routine CT under-calls venous bleeding, some of what Leipzig classified as cancellous (haematoma next to a fracture without a visible source) may have been venous all along. The authors acknowledge that their bleeding sources were inferred from imaging, not confirmed at angiography, surgery or autopsy [3].
The third is survivorship. Only patients who reached CT were included. Those who died at scene, in transit or in the resuscitation room, the very patients with the most catastrophic pelvic haemorrhage, are invisible in this cohort, and the authors say so [3]. If those deaths were disproportionately due to torn iliac vessels or massive venous disruption, the distribution among survivors-to-scanner tells us little about the lethal tail.
The fourth is relevance. This is a single-centre, retrospective cohort of all pelvic fractures, half of them fragility fractures in older people. It is not a cohort of haemodynamically unstable patients, which is the population ATLS teaching is actually about. A rule that describes the patient in shock need not describe the patient with a stable pubic ramus fracture.
These are serious objections. But they cut both ways. If CT cannot reliably distinguish venous from cancellous bleeding, then the confident claim that 85% of pelvic bleeding is venous rests on foundations at least as shaky as the new study's.
The fragile pelvis that bleeds
Whatever its limitations, the Leipzig study puts its finger on a demographic reality. In England and Wales, the Trauma Audit and Research Network reported in 2017 that falls from standing height had become the commonest cause of major trauma, and that older injured patients were less likely to be taken directly to a major trauma centre [5]. In Finland, the incidence of low-trauma pelvic fractures in older people rose by more than 3,000% between 1970 and 2013 [6].
The question is how often these low-energy fractures bleed dangerously. The answer is: not often, but not never. A single-centre study found clinically relevant bleeding in fewer than 1% of older patients with low-energy fragility fractures of the pelvis [7]. But in a nationwide analysis of the Japan Trauma Data Bank, Matsumoto and colleagues identified 1,354 patients aged 65 or over with pelvic fractures from ground-level falls (median age 85) and found that 80 (5.9%) needed therapeutic angioembolisation [8]. Systolic pressure below 90 mmHg on arrival was the strongest physiological predictor (odds ratio 5.43), alongside male sex, age, liver disease and cerebrovascular disease. Arterial bleeding requiring embolisation was independently associated with in-hospital death (odds ratio 4.3), and 28% of deaths occurred within 48 hours [8]. Registry selection inflates that 5.9%, as the authors concede. It is nonetheless a reminder that "low-energy" describes the mechanism, not the haemorrhage.
This matters because the UK guidance that governs binder use is written around mechanism. NICE's guideline on complex fractures recommends a purpose-made pelvic binder if active bleeding is suspected from a pelvic fracture "following blunt high-energy trauma" [9]. The older woman who falls in her kitchen and arrives grey and hypotensive with an unstable posterior ring injury sits awkwardly outside that sentence, even though everything in her physiology says she is bleeding.
What should change in the resus room?
Here is the contrarian point: if Leipzig is right, the pelvic binder becomes more defensible, not less. A binder does little for a lacerated superior gluteal artery. What it does do is reduce and stabilise the ring, bringing fracture surfaces back into apposition and limiting the movement that disrupts clot. That is precisely the intervention you would design for bleeding from cancellous bone. The rationale shifts from "tamponading the venous plexus" to "reducing the fracture", and that rationale only holds if the binder actually reduces the fracture.
That is where practice falls down. In Bonner and colleagues' study of 172 radiographs with a visible binder, only half were at the level of the greater trochanters; 39% sat too high, and in a subgroup the residual symphyseal gap was 2.8 times larger with a high binder [10]. A binder around the iliac crests is a belt, not a reduction device. If most pelvic bleeding comes from bone, the trochanteric position is not a detail for the skills station. It is the treatment.
The second implication concerns the patient who does not respond. A correctly placed binder that fails to stabilise the circulation points away from bone and towards an arterial source, or bleeding elsewhere, and the response needs to be definitive. NICE recommends interventional radiology as first-line invasive treatment for active arterial pelvic bleeding when laparotomy is not needed, and pelvic packing when it is [9]. In practice the choice between angioembolisation and preperitoneal packing is driven as much by local availability and speed as by anatomy. In the American Association for the Surgery of Trauma's multicentre study, mortality for pelvic fracture patients admitted in shock was 32%, with significant variation in haemorrhage-control strategy between institutions [11]. Denver's eleven-year experience of preperitoneal packing reported a median of 44 minutes from arrival to operation and 21% mortality in refractory shock, with 13% needing subsequent embolisation for an arterial blush [12]. A 2022 meta-analysis found crude mortality of 23% with packing versus 32% with angioembolisation, but no significant difference in studies that compared both directly, and the groups differed in age, injury severity, arterial haemorrhage and time to intervention [13]. The honest summary is that both work, the faster option available to you is usually the better one, and many of the sickest patients need both.
The third implication is about CT. If the source matters and examination in the resuscitation room cannot tell you what it is, then contrast-enhanced CT in the patient who responds to initial resuscitation is not a luxury; it is how you choose between angiography, packing and fixation. NICE already recommends CT as first-line imaging for suspected high-energy pelvic fractures in adults [9]. The tension, familiar to every ATLS candidate, is the unstable patient who should not go to the scanner. The Leipzig data do not resolve that. They simply raise the value of the scan for those who can tolerate it.
The real lesson: physiology, not percentages
Perhaps the deepest lesson is that the 85% statistic was never a management algorithm. Nobody in a resuscitation room can see whether the haematoma in front of them is arterial, venous or osseous. What they can see is the mechanism, the pelvic X-ray, the FAST, and the response to a well-placed binder and to resuscitation. The percentages were always a teaching heuristic, built on autopsy and angiography series from a younger, higher-energy era.
ATLS 11 has already moved the emphasis. Commentary on the new edition describes an xABCDE primary survey that puts control of exsanguinating haemorrhage ahead of the airway, discusses pelvic binders alongside tourniquets as early haemorrhage-control measures, and highlights an expanded chapter on trauma in the older adult [14, 15]. The Leipzig paper nudges in the same direction: think less about which vessel, and more about whether the bleeding is responding to what you have done.
What this means for your ATLS exam
Do not rewrite your exam answers on the strength of one retrospective study. The ATLS MCQ tests the course's teaching, and you should expect questions to frame pelvic haemorrhage in classical terms: disruption of the posterior pelvic venous complex and fracture surfaces as the commonest source, with arterial bleeding a minority that matters most in patients who remain unstable. Know the nuance; answer the course.
The high-yield points are these:
Suspect pelvic haemorrhage in any hypotensive trauma patient without an obvious external, chest or abdominal source, whatever the mechanism. Falls from standing in older people can produce life-threatening pelvic bleeding [8].
Centre the binder over the greater trochanters, not the iliac crests. A high binder fails to reduce the ring [10].
Do not repeatedly compress or "spring" the pelvis to test stability. One gentle examination at most, and none once a binder is on.
The AP pelvis X-ray is an adjunct to the primary survey. It should not delay a binder in a shocked patient with a suspected pelvic injury.
A patient who remains unstable despite a correctly placed binder needs definitive haemorrhage control: angioembolisation, or preperitoneal packing with skeletal stabilisation, plus laparotomy if there is intraperitoneal bleeding. The choice depends on local resources and associated injuries [9, 12, 13].
- In older patients, apparently reassuring vital signs can mask significant blood loss, and a systolic pressure below 90 mmHg is an ominous sign [8].
References
1. Coccolini F, Stahel PF, Montori G, et al. Pelvic trauma: WSES classification and guidelines. World J Emerg Surg. 2017;12:5. DOI: https://doi.org/10.1186/s13017-017-0117-6
2. Khaliq F, Rodham P. EMS pelvic binders. In: StatPearls. Treasure Island (FL): StatPearls Publishing; updated 1 January 2024. https://www.ncbi.nlm.nih.gov/books/NBK598968/
3. Gerhardt F, Höch A, Notov D, Henkelmann J, Zeidler S, Einhorn S, Kleber C. Pelvic hemorrhage in the context of demographic change: frequency and CT-based characterization of pelvic bleeding. Eur J Trauma Emerg Surg. 2026;52(1):320. Published online 15 October 2026. DOI: https://doi.org/10.1007/s00068-026-03363-6 Full text (PMC): https://pmc.ncbi.nlm.nih.gov/articles/PMC13672563/
4. Hussami M, Grabherr S, Meuli RA, Schmidt S. Severe pelvic injury: vascular lesions detected by ante- and post-mortem contrast medium-enhanced CT and associations with pelvic fractures. Int J Legal Med. 2017;131(3):731–8. DOI: https://doi.org/10.1007/s00414-016-1503-4 Full text (PMC): https://pmc.ncbi.nlm.nih.gov/articles/PMC5388710/
5. University of Manchester. Falls now commonest type of major trauma in England and Wales, report reveals [press release on the Trauma Audit and Research Network report Major Trauma in Older People]. April 2017. https://www.manchester.ac.uk/about/news/falls-now-commonest-type-of-major-trauma-in-england-and-wales-report-reveals/
6. Kannus P, Parkkari J, Niemi S, Sievänen H. Low-trauma pelvic fractures in elderly Finns in 1970–2013. Calcif Tissue Int. 2015;97:577–80. DOI: https://doi.org/10.1007/s00223-015-0056-8
7. de Herdt CL, et al. Clinically relevant bleeding risk in low-energy fragility fractures of the pelvis in elderly patients. Eur J Trauma Emerg Surg. 2024;50:1585–9. DOI: https://doi.org/10.1007/s00068-024-02492-0
8. Matsumoto S, Aoki M, Shimizu M. Prevalence and risk factors for major arterial bleeding in fragility pelvic fractures in the aging population. Sci Rep. 2025;15:38257. DOI: https://doi.org/10.1038/s41598-025-22076-1
9. National Institute for Health and Care Excellence. Fractures (complex): assessment and management. NICE guideline NG37. London: NICE; 2016 (updated 2022). https://www.nice.org.uk/guidance/ng37/chapter/recommendations
10. Bonner TJ, Eardley WGP, Newell N, et al. Accurate placement of a pelvic binder improves reduction of unstable fractures of the pelvic ring. J Bone Joint Surg Br. 2011;93(11):1524–8. DOI: https://doi.org/10.1302/0301-620X.93B11.27023
11. Costantini TW, Coimbra R, Holcomb JB, et al.; AAST Pelvic Fracture Study Group. Current management of hemorrhage from severe pelvic fractures: results of an American Association for the Surgery of Trauma multi-institutional trial. J Trauma Acute Care Surg. 2016;80(5):717–25. DOI: https://doi.org/10.1097/TA.0000000000001034
12. Burlew CC, Moore EE, Stahel PF, et al. Preperitoneal pelvic packing reduces mortality in patients with life-threatening hemorrhage due to unstable pelvic fractures. J Trauma Acute Care Surg. 2017;82(2):233–42. DOI: https://doi.org/10.1097/TA.0000000000001324 Full text (PMC): https://pmc.ncbi.nlm.nih.gov/articles/PMC5250563/
13. McDonogh JM, Lewis DP, Tarrant SM, Balogh ZJ. Preperitoneal packing versus angioembolization for the initial management of hemodynamically unstable pelvic fracture: a systematic review and meta-analysis. J Trauma Acute Care Surg. 2022;92(5):931–9. DOI: https://doi.org/10.1097/TA.0000000000003528
14. The Skeptics' Guide to Emergency Medicine. SGEM Xtra: This One Goes to 11 – ATLS 11th Edition. March 2026. https://thesgem.com/2026/03/sgem-xtra-this-one-goes-to-11-atls-11th-edition/
15. Mayo Clinic. ATLS update: what trauma professionals need to know. Mayo Clinic Medical Professionals (Trauma). 2026. https://www.mayoclinic.org/medical-professionals/trauma/news/atls-update-what-trauma-professionals-need-to-know/mac-20602992
