Article Text

Download PDFPDF

Original research
Complications of femoral versus radial access in neuroendovascular procedures with propensity adjustment
  1. Joshua S Catapano1,
  2. Vance L Fredrickson1,
  3. Tatsuhiro Fujii2,
  4. Tyler S Cole1,
  5. Stefan W Koester1,
  6. Jacob F Baranoski1,
  7. Daniel D Cavalcanti1,
  8. D Andrew Wilkinson1,
  9. Neil Majmundar1,
  10. Michael J Lang1,
  11. Michael T Lawton1,
  12. Andrew F Ducruet1,
  13. Felipe C Albuquerque1
  1. 1 Department of Neurosurgery, Barrow Neurological Institute, Phoenix, Arizona, USA
  2. 2 Department of Neurological Surgery, Keck School of Medicine, University of Southern California, Los Angeles, Califronia, USA
  1. Correspondence to Dr Felipe C Albuquerque, Department of Neurosurgery, Barrow Neurological Institute, Phoenix, AZ 85013, USA; neuropub{at}


Background The transradial artery (TRA) approach for neuroendovascular procedures continues to gain popularity, but neurointerventionalists still lag behind interventional cardiologists in the adoption of a TRA-first approach. This study compares the complications and efficiency of the TRA approach to the standard transfemoral artery (TFA) approach at our institution during our initial phase of adopting a TRA-first approach.

Methods A retrospective analysis was performed on all consecutive neuroangiographic procedures performed at a large cerebrovascular center from October 1, 2018 to June 30, 2019. The standard TFA approach was compared with TRA access, with the primary outcome of complications analyzed via a propensity-adjusted analysis.

Results A total of 1050 consecutive procedures were performed on 877 patients during this 9-month period; 206 (20%) procedures were performed via TRA and 844 (80%) via TFA. The overall complication rate was significantly higher with the TFA procedures than with the TRA procedures (7% (60/844) vs 2% (4/206), respectively; p=0.003). A propensity-adjusted analysis showed that the TFA approach was a significant risk factor for a complication (OR 3.6, 95% CI 1.3 to 10.2, p=0.01). However, the propensity analysis showed that fluoroscopy times were on average 4 min less for TFA procedures than for TRA procedures (p=0.003).

Conclusion The TRA approach for neuroendovascular procedures appears to be safer than the TFA approach. Although a steep learning curve is initially encountered when adopting the TRA approach, the transition to a TRA-first practice can be performed safely for neurointerventional procedures and may reduce complications.

  • complications
  • TRA
  • transradial neuroendovascular
  • TRF
  • transfemoral neuroendovascular

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


The radial artery approach has become an increasingly popular alternative for arterial access in endovascular interventions.1 First described by Campeau2 nearly three decades ago, the use of a percutaneous radial artery approach for angiography has seen a significant increase in the field of interventional cardiology owing to comparative data demonstrating superiority in clinical outcome, patient satisfaction, and cost reduction compared with the standard femoral artery access.2–11 The association between radial access and reduced post-procedural bleeding, major vascular complications, overall adverse clinical events, and hospital length of stay is well documented in the cardiology literature.12–15 In light of this unequivocal evidence, the American Heart Association guidelines now recommend a 'radial-first' strategy for patients requiring treatment for acute coronary syndromes.16

Several advantages of the transradial artery (TRA) approach established in neurointerventional studies include a more direct access to the vertebrobasilar system, greater ease navigating complex arch anatomy such as bovine and type III arches, and robust collateral circulation from the ulnar artery leading to fewer ischemic (often clinically silent) complications.17–19 Variations to the traditional proximal radial access now exist, including distal radial artery access within the anatomic 'snuffbox' (radial fossa).20 Among its many advantages, this technique allows arterial access distal to the branch point of the superficial palmar branch supplying collaterals to the deep palmar arch of the hand; theoretically, this approach decreases the risk of an ischemic complication even further.20 Despite a smaller caliber vessel, the use of larger bore coaxial systems for intervention via the distal radial approach has been reported.21

The adoption of this paradigm shift away from the traditional transfemoral artery (TFA) access in the neurointerventional field has been lagging, however, in large part because of operator inexperience with the TRA approach and a general lack of large studies replicating the outcomes demonstrated by our cardiology colleagues.22 The present study, which is the largest to date, compares the risk of complications via a propensity match analysis in neuroendovascular procedures with a TRA versus TFA approach at a single center.


All patients undergoing a neuroangiographic procedure at a single center from October 1, 2018 to June 30, 2019 were retrospectively reviewed. The study was approved by the St Joseph’s Hospital and Medical Center Institutional Review Board for Human Research (Phoenix, Arizona, USA) and complied with the Health Insurance Portability and Accountability Act. The need for patient consent was waived by the institutional review board because of the retrospective nature of the study. Electronic medical records were analyzed for age, sex, procedure, angiographic pathology, amount of contrast, fluoroscopy time, complications, sheaths/catheters used, and access site which was confirmed via angiographic imaging. Procedures were grouped as (1) diagnostic angiography and/or vasospasm treatment that only included verapamil infusion or (2) specific treatment. Access sites were categorized as either TRA (both distal or 'snuffbox' and proximal or 'wrist') or TFA. Access site crossover was recorded based on both the operative report and saved angiographic imaging, and outcomes were assigned to the actual final procedure performed. Complications were categorized as minor (access site hemorrhages that did not require treatment) or major (vessel dissection/perforations, thromboembolic events, intracranial hemorrhages, retroperitoneal hemorrhages, and vessel occlusions).

Statistical analysis was performed using SPSS Statistics for Windows, Version 26.0 (IBM Corp, Armonk, New York, USA) along with Excel (Microsoft, Redmond, Washington, USA). Independent t-test and χ2 analyses were performed to compare cohorts. Additionally, a propensity match analysis (matching for age, sex, sheath size, catheter size, procedure, and angiographic pathology) was performed to analyze the risk of access site complications via a multivariate logistic regression analysis and a linear regression for efficiency analysis (efficiency measured as fluoroscopy time and volume of contrast used). P values <0.05 were considered to be statistically significant.


During the 9-month study period a total of 877 patients had 1050 consecutive neuroangiographic procedures performed at a single center, with 43 patients (5%) having multiple procedures that included both radial and femoral access. Of the 1050 procedures, 206 (20%) were performed via TRA and 844 (80%) via TFA. There was no statistically significant difference between patients in the TRA and TFA groups with regard to sex (62% (127/206) vs 59% (498/844) female, respectively, p=0.53) or mean age (56±17 vs 59±17 years, p=0.07) (table 1). The types of procedures performed between the two groups were statistically different (p<0.001): the percentage of diagnostic/vasospasm procedures was greater in the TRA group (72%, 148/206) than in the TFA group (62%, 522/844), while a higher percentage of TFA procedures (16%, 138/844) than TRA procedures (4%, 9/206) were performed for a thrombectomy, thrombolysis, carotid artery stent, and/or stent for atherosclerosis/stroke (table 1). There was no significant difference in catheter sizes used for the TRA and TFA procedures, but significantly larger sheath sizes were used in TFA approaches compared with TRA (p=0.004) (table 1).

Table 1

Characteristics of patients undergoing angiographic procedures by access site*

Table 2 shows a comparison of the intraprocedural variables and outcomes for the TRA and TFA procedures. The mean total fluoroscopy time (TRA, 23±17 min; TFA, 20±17 min; p=0.09) and mean total contrast amount (TRA, 121±64 mL; TFA, 113±64 mL, p=0.18) for the two groups were not significantly different. Additionally, among the subgroup of diagnostic/vasospasm procedures there was no significant difference between access routes in terms of either mean fluoroscopy time (TRA, 15±8 min; TFA, 14±13 min; p=0.30) or mean total contrast agent administered (TRA, 106±55 mL; TFA, 102±76 mL; p=0.61). However, mean fluoroscopy time and contrast total administered in interventional cases were significantly higher in TRA versus TFA procedures (40±20 min vs 30±19 min, p=0.001; 159±69 mL vs 131±57 mL, p=0.002; respectively). The percentage of access site crossover was similar in both groups (TRA, 1.5% (3/206) vs TFA, 1.3% (11/844), p=0.89).

Table 2

Comparison of intraprocedural variables and outcomes by access site during angiography*

A significantly greater percentage of complications was noted among TFA procedures (TFA, 7% (60/844) vs TRA, 2% (4/206); p=0.003). Of the 64 complications reported, 50 (78%) were minor complications (groin hematomas without need for treatment) with a significantly higher percentage of minor complications occurring in TFA compared with TRA procedures (6% (47/844) vs 2% (3/206), respectively, p=0.01). A total of 14 of the 64 (22%) observed complications were major. No significant difference between access site groups (p=0.33) was observed in terms of the percentage of major complications; among 206 TRA procedures, one (0.5%) thromboembolic event occurred compared with 13 (2%) major complications occurring among 844 TFA procedures (6 vessel dissections/perforations, 2 intracranial hemorrhages, 2 thromboembolic events, 2 occlusive femoral arteries, and 1 retroperitoneal bleed). When evaluating the subgroup of procedures performed for interventional treatment, there was no significant difference in complication rates between the two access sites (TRA, 5.2% (3/58) vs TFA, 9% (30/321), p=0.45). However, when evaluating the subgroup of patients undergoing diagnostic/vasospasm procedures, significantly more complications (p=0.007) occurred in the TFA group (6%, 30/522) than the TRA group (1%, 1/148).

As outlined in table 3, among the 206 TRA procedures, 89 (43%) were performed via distal ('snuffbox') access and 117 (57%) were performed via a proximal (wrist) puncture. There was no significant difference in sex, mean age, type of procedure, angiographic pathology, total contrast amount administered, access site crossovers, or complications between distal and proximal TRA access site procedures. Proximal compared with distal TRA procedures required significantly larger sheaths (>5 F sheath used in 41% (48/117) of proximal vs 21% (21/89) of distal TRAs, p=0.04) and larger catheters (>5 F catheter used in 38% (44/117) of proximal vs 19% (19/89) of distal TRAs, p=0.02). Additionally, mean fluoroscopy time was significantly shorter in the distal TRAs than in the proximal TRAs (19±13 vs 24±21 min, respectively; p=0.045).

Table 3

Characteristics and comparison of radial access procedures categorized as either distal or proximal

A propensity match analysis (matched for sex, age, sheath size, catheter size, type of procedure, and angiographic pathology) was performed and included in a logistic regression analysis for access site complications. The results presented in table 4 show the TFA procedure to be a major risk factor (adjusted OR 3.6, 95% CI 1.3 to 10.2; p=0.01) for complications when compared with TRA access. Additionally, an adjusted linear regression analysis for efficiency was performed. Data showed a significant difference in mean fluoroscopy time (p=0.003) with TFA procedures associated with 4 min less time than TRA procedures, but no significant difference (p=0.09) in mean total contrast amount used (9.8 mL less in TFA vs TRA approaches).

Table 4

Propensity-adjusted* predictors of outcomes for femoral access (vs radial access)


The TRA approach for neuroendovascular procedures has continued to gain traction over the past several years with a wide array of benefits reported, including a reduced risk of severe complications.22 23 However, unlike in interventional cardiology, the use of TRA still remains secondary to a TFA approach for neurointerventional procedures, largely due to the general lack of large studies in this field.4 22 This paper presents the largest known single-center neurointerventional study comparing the TRA with the TFA approach. Among the procedures analyzed, only 2% (4/206) of TRAs resulted in complications compared with 7% (60/844) when using a TFA approach (p=0.003). Furthermore, a propensity-adjusted analysis found TFA procedures to be a major risk factor for a complication (OR 3.6; p=0.01) compared with TRA procedures. Additionally, there were no life-threatening complications with the TRA procedures while five major complications were found with the TFA approach (2 occlusive/near occlusive femoral arteries, 1 retroperitoneal bleed, and 2 large intracranial hemorrhages). In 2018, Snelling et al 22 found a similarly low rate of complications in 148 procedures using a TRA approach; no major complications but 10 (~6%) minor complications were noted. More recently, in 2019, Khanna et al 23 found 10 minor complications in 233 patients with a TRA neuroendovascular procedure. When complications were broken down by procedure type in our study, we found no significant difference in complication rates between TRA and TFA when performed for treatment; however, when analyzing diagnostic and vasospasm angiography procedures, the TFA approach had a 6% complication rate versus a 1% complication rate for the TRA approach (p=0.007). Diagnostic angiography is viewed as a relatively benign procedure and our findings suggest that a TRA approach makes these procedures even safer for patients.

A major limitation of the TRA approach discussed in the literature is the steep learning curve observed.22 Online supplementary figure 1 illustrates the number of transradial procedures performed each month during the study, with the majority performed in the last 2 months of the study (137 of 206, 67%). Although the complications in the present study occurred in the first 60 cases, there was no significant difference in the percent of procedures that could not be completed and that needed to cross over between the TRA and TFA approach (1.5% vs 1.3%, respectively). Snelling et al 22 found comparable results with only 4.7% crossover in TRA procedures. Moreover, they noted that TRA took slightly longer than TFA approaches for diagnostic angiograms. Although the mean fluoroscopy time for diagnostic angiography in the present study was 1 min longer for the TRA than for the TFA procedures (15 min vs 14 min), this was not a significant difference (p=0.30); this group also included angiograms for vasospasm treatment. Interestingly, this slight difference in fluoroscopy time did not hold true for treatment procedures, where we found significantly longer procedures with the TRA approach (40 min vs 30 min, p=0.001). Additionally, in our linear regression propensity-adjusted analysis, mean fluoroscopy time was found to be 4 min less with the TFA procedures (p=0.003). However, mean TRA fluoroscopy time was greater in the first 40 treatments (41 min) than the last 18 treatments (36 min), suggesting that with continued TRA use, the interventionalist becomes more efficient.

Recently, Brunet et al 20 reported the use of the distal TRA approach via the anatomic snuffbox for neurointerventional procedures, which reportedly decreased the rate of radial artery occlusion and improved patient comfort. The present paper is the first in the neurointerventional literature to compare the distal (snuffbox) TRA and standard proximal (wrist) TRA approaches. A greater percentage of diagnostic/vasospasm angiograms were performed using the distal rather than the proximal TRA approach in this study. Interestingly, we found no significant differences in access site crossovers or complication rates between the two TRA approaches (the one complication in the distal TRA was slight swelling at the access site with no clinical significance). However, total mean fluoroscopy time was 5 min shorter with the distal TRA approach than with the proximal TRA procedures (p=0.045). Although the difference was not significant, for diagnostic procedures the mean fluoroscopy time was 2 min less using the distal versus proximal TRA approach (14 min vs 16 min, p=0.10). This shorter time was likely due to the distal procedures being performed during the latter part of the study and greater operator expertise. We also found a significant difference with regard to sheath and catheter size, with larger sizes used in the proximal versus distal TRA approaches; this is likely due to a larger number of treatments performed in those procedures.


The TRA approach for neuroendovascular procedures appears to be safer than the standard TFA approach. The distal (snuffbox) TRA approach has a similar low risk of complications to the more proximal (wrist) TRA approach. In particular, the complication rate is significantly lower in diagnostic procedures using the TRA approach compared with the TFA approach. Although a steep learning curve is initially encountered when using the TRA approach, our results suggest that the transition to a TRA-first practice can be performed safely, with a minimal loss in efficiency.


The authors thank the Neuroscience Publications staff at Barrow Neurological Institute for help with manuscript preparation.



  • Contributors All authors made substantial contributions to the conception or design of the work. JSC: study design, data collection, data analysis, manuscript editing and writing, and statistical analysis. ML: study design and data analysis. SK: data collection and manuscript writing. NM: study design and analysis. DAW: study design and analysis. JFB: manuscript editing and data analysis. VLF: manuscript editing and data analysis. TSC: manuscript editing, data analysis, statistical anaylsis. TF: manuscript writing. DDC: data collection. MTL: study design and manuscript editing. AFD: study design and manuscript editing. FCA: study design, manuscript editing, and guarantor.

  • Disclaimer The institutional review board waived the need for patient consent because of the retrospective nature of the study. Thus, no attempt was made to contact patients whose data were included in the analysis. Personal data have been sufficiently anonymized per the journal requirements.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement No data are available.