Supraclavicular versus infraclavicular approach for ultrasound-guided right subclavian venous catheterisation: a randomised controlled non-inferiority trial

Kim. Anaesthesia 2022,77(1)59-65;

Clinical Question

  • In patients undergoing elective neurosurgery, is a supraclavicular in-plane ultrasound guided approach, compared with an infra-clavicular out-of-plane ultrasound guided approach, non-inferior for catheterisation-related complication?


  • The internal jugular, femoral and subclavian veins are commonly used for central venous lines
  • The use of subclavian lines is associated with a lower risk of infection but a higher risk of pneumothoraces
  • Subclavian lines can be placed either supra or infra-clavicular. Previous single operator RCTs (1, 2) have demonstrated that for a single operator ,ultrasound guided supraclavicular approach was quicker, with similar success rates and fewer complications, compared with an infraclavicular approach
  • When an infra-clavicular approach is used the use of the short axis out-of-plane ultrasound has been reported to have significant advantages over the use of long axis in-place ultrasound


  • Randomised controlled trial
  • Non-inferiority trial
  • Computer generated random allocation sequence, 1:1 ratio
  • Use of sealed opaque envelopes to maintain allocation concealment
  • Sample size calculation: 416 patients. With a false positive rate set at 5%, false negative rate set at 20%, and non-inferiority margin set at 5.9%, 416 patients required assuming a drop-out rate of 5%


  • Single centre, Seoul National University Hospital
  • September 2018 – May 2020


  • Inclusion:
    • Age 20-79
    • Required central venous catheterisation
    • Undergoing elective general anaesthetic for neurosurgery
  • Exclusion:
    • Infection or medical device at the puncture site
    • Right subclavian venous thrombosis
    • Haemostatic disorders or receiving anticoagulant
    • Previous history of surgery that might distort the anatomy of the right subclavian vein
  • 416 patients randomised. 8 patients excluded as not appropriately allocated, 4 withdrew consent, and 3 did not meet inclusion criteria, leaving 401 patients included in analysis
  • Comparing baseline characteristics of intervention vs. control group
    • Male: 45% vs 49%
    • Age: 53 vs 51 years
    • BMI: 24 vs 25
    • ASA 1/2: 79% vs 80%
    • ASA 3: 17% vs 18%
    • ASA 4: 4% vs 2%


  • Supra-clavicular approach
    • Long-axis in-plane ultrasound


  • Infra-clavicular approach
    • short-axis out-of-plane ultrasound

Management common to both groups

  • All other peri-operative management was the same
  • Central line inserted after induction of general anaesthesia. Bed in flat position, with pillow removed and head, neck and arms in a neutral position
  • VscanTM portable ultrasound machine
  • One of four anaesthetists who all had at least 10 successful cannulations with both approaches
  • Mechanical ventilation stopped prior to venous puncture, and re-started after successful catheterisation


  • Primary outcome: proportion of patients with catheterisation-related complications, including catheter misplacement and mechanical complications (arterial puncture, haematoma formation, pneumothorax and haemothorax)
    • 3% in supra-clavicular vs 13.4% in infra-clavicular group
      • Mean difference -10.4% (95% CI -15.7 to -5.1%), OR 0.20 (95% CI 0.8-0.49), p<0.001
  • Secondary outcomes:
  • Comparing intervention vs. control group
    • No significant difference in
      • Overall success
        • 99.5% vs 99.5%
      • Success at 1st attempt for catheterisation
        • 87% vs 88%
      • Time required for catheterisation
        • 79 vs 78 seconds
      • Incidence of mechanical complications
        • 2% vs 3%
      • Arterial puncture
        • 2% vs 1.5%
      • Pneumothorax
        • 0% vs 1%
      • Rescue ventilation
        • 14.5% vs 12.4%
    • Significantly less in supra-clavicular group
      • Catheterisation related complication
        • 3% vs 13.4%
        • Mean difference -10.4 (95% CI -15.7 to -5.1%), OR 0.2 (0.08-0.49), p<0.001
      • Catheter misplacements
        • 1% vs 10.4%
        • Mean difference -9.4 (95% CI -13.9 to -5), OR 0.09 (0.02-0.37), p<0.001
      • Catheter misplacement into ipsilateral internal jugular vein 0.5% vs 8%
        • Mean difference -7.5 (95% CI -11.4 to -3.6%), OR 0.06 (0.01-0.44), p<0.001
      • Time required for venous puncture
        • 9 vs 13 seconds

Authors’ Conclusions

  • The supraclavicular approach seems to be a better approach for ultrasound-guided right subclavian venous catheterisation


  • Randomised controlled trial
  • Multi-operator, all operators were experienced with both techniques
  • Allocation concealment maintained


  • The supraclavicular approach used an inplane ultrasound where as the infaclavicular approach used out-of plane ultrasound. Although i understand the reasons for this, this may be a confounding variable
  • This was a non-inferiority study and and was not designed to address superiority
  • Single centre study in South Korea
  • Only included patients who were mechanically ventilated and having elective neurosurgery. Results may not be generalisable to other settings
  • Blinding of those inserting lines not possible
  • 3.6% of patients randomised not included in analysis

The Bottom Line

  • In elective patients undergoing subclavian central line insertion, the use of a supra-clavicular approach with in-plane ultrasound was non-inferior to an infra-clavicular approach with out-of-plane ultrasound.
  • The success rates were similar for both approaches where as the rate of catheter related complications was significantly less with the supra-clavicular approach. This was predominately due to a reduction in catheter misplacement

External Links


Summary author: @davidslessor
Summary date: 14th Jan 2022
Peer-review editor: @avkwong

Picture by: [author/site]


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.