Brachial artery access may be riskier than doctors think, study finds

When interventional radiologists need to thread catheters through the arterial system, they usually reach for the groin or the wrist. Femoral access has long been the workhorse of endovascular procedures, and radial access has grown in popularity, particularly in coronary work. But for some patients, neither route works well. Hostile groins scarred by surgery or disease, difficult anatomy, or the need for a larger sheath than the radial artery can accommodate all push operators toward a third option: the brachial artery at the elbow. A new study from a tertiary centre suggests this fallback route may carry a higher risk than many clinicians assume.

The study, published in CVIR Endovascular, reviewed every procedure using percutaneous brachial artery access at a single institution over a 12.5-year period, from September 2013 to May 2025. Researchers identified 100 confirmed brachial punctures from radiology records and electronic patient notes, then recorded patient demographics, sheath sizes, closure techniques and any complications, grading each event according to the CIRSE classification system used widely in interventional radiology.

The headline figure is stark. Twenty-three of the 100 procedures resulted in a complication, an overall rate of 23%. Nine of those, or 9% of all cases, were grade 3 events requiring additional treatment. Five pseudoaneurysms were treated successfully with a single thrombin injection during admission, while one pseudoaneurysm failed thrombin therapy and needed surgical closure. One patient with prolonged bleeding and an expanding haematoma went to theatre for surgical repair. Another developed a large haematoma requiring blood products and extra follow-up, and one symptomatic arterial occlusion demanded unplanned anticoagulation and prolonged monitoring. All but one complication were diagnosed during the same admission; a single asymptomatic occlusion surfaced months later when the patient returned for another procedure.

That rate sits well above the roughly 11% often cited in the published literature, although the authors stress that reported figures vary widely because studies differ in how they define complications, in access techniques, sheath sizes and closure methods. The mismatch between the literature and this single-centre experience was precisely what prompted the audit. The team had noticed an increasing number of patients whose complex vascular disease made femoral or radial access unsuitable, and they perceived more complications than the literature suggested. The data confirmed that perception.

The patients were typical of a tertiary vascular practice: 64 men and 36 women with a mean age of 69 years. Sheaths ranged from 3 to 9 French in diameter, with 57% of cases using a 6 Fr sheath and twelve cases using something larger. Most procedures, 84%, were aortic, mesenteric or renal, or peripheral arterial interventions. Every access was obtained with ultrasound-guided puncture of the brachial artery in the antecubital fossa, and every puncture and closure was performed under the direct or indirect supervision of a consultant interventional radiologist.

Perhaps the most striking finding concerns vascular closure devices. These devices, designed to seal arterial punctures without prolonged manual compression, are not approved for brachial use, yet they were deployed in seven cases. Five of those seven, 71%, resulted in complications, including three grade 3 events: two pseudoaneurysms requiring thrombin injection and one surgical repair. In a secondary analysis, closure device use was associated with grade 3 complications at a rate of 42.9% versus 6.5% in cases without a device, a difference that reached statistical significance with Fisher’s exact test at p = 0.03. The authors urge caution in interpreting this, given that only seven devices were used, but the signal is hard to ignore, particularly alongside the devices’ off-label status at this access site.

The search for other risk factors came up largely empty. No convincing association emerged between complication rates and patient age, gender or sheath size. Among the 77 uncomplicated cases, sheath sizes ranged from 3 to 9 Fr with a median of 6 Fr; among the 23 complicated cases, sizes ranged from 4 to 6 Fr, again with a median of 6 Fr. Mean age was 68 years in patients without complications and 71 in those with, while patients with grade 3 complications had a mean age of 76. There was a trend toward increasing age among grade 3 cases, with a median of 78 versus 68 years and p = 0.055, just short of conventional significance. Female patients made up 26 of 77 uncomplicated cases and 10 of 23 complicated ones, and four of the nine grade 3 events occurred in women, a pattern the authors note is consistent with earlier work identifying female gender as a demographic risk factor, though the small sample precluded robust statistical analysis.

One reassuring absence was median nerve injury, a recognised but rare complication of brachial puncture that did not appear once in this series. The authors attribute this at least in part to routine ultrasound guidance and operator familiarity with upper-limb neurovascular anatomy, and they argue that ultrasound guidance should be considered standard practice for this access route. The retrospective design, however, limited what could be examined. Documentation of needle gauge, compression time, elbow immobilisation and pressure dressing use was inconsistent, and the team could not reliably capture anticoagulation status, peri-procedural blood pressure, operator experience or the exact puncture level along the artery, all potential confounders. The case mix of a tertiary referral centre, where patients often have complex anatomy and few alternative access options, may also inflate the observed rate.

The study’s practical consequences are already visible at the institution involved. The audit directly informed a new departmental standard operating procedure and a structured reporting template covering access technique, closure, documentation and follow-up. Patients are now counselled during informed consent when brachial access may be needed. The department has reiterated the importance of ultrasound-guided single-wall puncture, avoiding double-wall or multiple punctures, atraumatic sheath insertion, and careful attention during sheath removal and manual pressure. Haemostasis, the authors emphasise, is a key part of the overall procedure and should not be left to an unsupervised junior trainee.

The broader message is one of humility about a technique often treated as routine. Brachial access retains real advantages: a larger vessel than the radial artery, a shorter working distance to target anatomy, and value in carotid stenting and complex aortic aneurysm repair, along with faster ambulation and better patient comfort than femoral access. But the authors conclude that brachial artery access may be less safe than commonly perceived and should be reserved for carefully selected cases. They call for wider auditing and transparent reporting across centres to define best practice, and they suggest that standardised procedural protocols and post-procedural care pathways, including their own practice of at least 10 minutes of manual compression, arm splinting for 4 to 6 hours and a duplex scan before discharge, may help reduce complications in the meantime.

Subject of Research: Safety and complication rates of percutaneous brachial artery access for endovascular procedures

Article Title: Brachial artery access for endovascular procedures: how safe is it?

Article References: Howroyd, L. R., Khattar, A. S., Tam, H.-Y. C., Morgan, R., & Chun, J.-Y. (2026). Brachial artery access for endovascular procedures: how safe is it?. CVIR Endovascular, 9(1), Article 133. https://doi.org/10.1186/s42155-026-00741-4

Image Credits: AI Generated

DOI: 10.1186/s42155-026-00741-4

Keywords: brachial artery access, endovascular procedures, access-site complications, vascular closure devices, CIRSE classification, interventional radiology, ultrasound guidance, retrospective study, pseudoaneurysm, arterial occlusion, sheath size, tertiary centre