Several approaches to the subclavian vein have been described, but most commonly an infraclavicular route is used ^{1, 2}. If the right side is used, the success rate for entry into the right ventricle is higher than for the external jugular approach but lower than for the right internal jugular approach. If the left side is used, right ventricular entry is easily achieved provided that the catheter is inserted with its natural curve following the line of the venous anatomy (Figure 1). The approach is suitable for use in infants and children ^{3, 4}.

The infraclavicular approach to cannulation can be performed as follows (Figure 2):

The patient is positioned in the head-down position, arms by their side with the head turned towards the opposite side of intended cannulation. A small sandbag or rolled-up sheet is placed longitudinally between the scapulae and an assistant applies gentle caudal traction to the arm on the same side as intended cannulation. These manoeuvres have the effect of minimising interference with the path of needle insertion by the humeral head and bringing the subclavian vein into close contact with the undersurface of the clavicle - thereby creating a predictable relationship between these two structures. (The effects of variation in shoulder position on the relationship between the clavicle and the subclavian vein have been elegantly analysed by Tan et al ¹³.)

After skin preparation and draping in accordance with institutional protocols, the skin is infiltrated with local anaesthetic at a point 1- 2cms below the inferior border of the clavicle in the midclavicular line. The infiltration is continued subcutaneously and infraclavicularly, using a 23g needle, whilst aiming at a point just above the sternal notch. If a more medial approach is chosen, passage under the clavicle tends to be more difficult, but the risk of inadvertent arterial or pleural puncture is reduced. In contrast, a more lateral approach facilitates infraclavicular passage, but because the needle is more posterior, increases the risk of either arterial or pleural damage (Figure 3).

An 18g, short-bevel needle (attached to an empty 10ml syringe) is now inserted at the skin infiltration point and again directed at a point just above the sternal notch. It is important that this needle not be directed posteriorly for fear of the risk of damage to the pleura or subclavian artery. The chance of inadvertent, posterior, direction of the needle can be reduced by ensuring that the shaft of the needle is maintained in the horizontal plane at all times and by putting a 10 degree upward bend on the needle adjacent to the hub (Figure 4). A constant, gentle negative pressure should be applied to the syringe so that a flash of blood will be observed on entry into the vein.

The practise of priming the aspirating syringe with saline (as recommended in some resuscitation protocols ¹⁵ ) is unwarranted and makes the differentiation of venous and arterial blood more difficult ¹⁴.

If the clavicle is encountered as the needle is advanced, the syringe and needle should be moved posteriorly as a unit by applying backward pressure to the shaft of the needle as it passes subcutaneously beneath the operators' thumb. If the vein is not entered after passing the needle a few centimetres beneath the clavicle, it should be withdrawn slowly whilst still aspirating, as the vein is occasionally encountered on withdrawal rather than advancement of the needle.

The process of advancing and withdrawing the needle in the subclavicular region should be repeated, whilst making small alterations in the cephalad or caudad direction of the needle, until venous entry is obtained. Clinicians should impose their own limit on the number of subclavicular passes which they are prepared to make as persistence with the approach in the face of repeated failure greatly increase the risk of inadvertent arterial or pleural puncture ⁶. The authors' limit is about four attempts.

Once the vein has been entered and the intravenous position of the needle has been confirmed (Figure 5), the Seldinger ⁵ wire is passed through the needle and threaded to a distance of not more than ~20 cms.

Andrews et al ⁷, have recently measured the distance from the skin puncture site to atrio-caval junction in 100 patients undergoing central venous cannulation by various routes. These distances were:

The pulmonary artery catheter introducer is now passed into the vein over the wire (Figure 6), (Video 2) and the catheter inserted (See section on catheter flotation).

Subclavian venepuncture is usually associated with a higher incidence of pneumothorax and haemothorax than the internal jugular approach, but the possibility of damage to the carotid artery is virtually non-existent and the incidence of catheter-related sepsis is much lower. If the patient is undergoing median sternotomy, the catheter can kink within the introducer sheath when the sternal retractor is in place. The incidence of this complication has been reported as up to 45% ⁸, although in our experience it can almost always be avoided if the introducer sheath is withdrawn 1 - 2 cms over the catheter after the catheter has been positioned. If this manoeuvre is used, the continuing intravascular position of the introducer sheath tip must be confirmed (by aspiration of blood from the side arm) before any drug or fluid is administered via the side arm. (See "Extravascular lumen placement" in the section entitled 'Miscellaneous Complications.)

An additional risk of a left-sided approach is the risk of a thoracic duct injury. This has been reported after both left internal jugular and left subclavian venous cannulation ^{9, 10}, but the complication is extremely rare. The point of damage to the thoracic duct is usually at its site of insertion at the confluence of the internal jugular and subclavian veins (Figure 7) and chylothorax may then result. This condition is notoriously difficult to treat and is the major reason for selecting the right-sided approach to the subclavian vein in preference to the left.

The particular advantages of the infraclavicular subclavian approach are that the site is easily dressed; that it is associated with a much lower rate of sepsis than the jugular route ¹¹ ; that it is well tolerated by patients and that the possibility of damage to the carotid artery is virtually non-existent. It is also suitable for use in small children ^{3, 4}.

Supraclavicular approaches to the subclavian vein have been described. However, in the opinion of one of the authors (DAP), they are more likely to be associated with pneumothorax and, on the left side, the possibility of thoracic duct injury than other approaches. In a recent study by Muhm et al ¹² , a high success rate and low pneumothorax rate were reported, but the incidence of thoracic duct puncture was just under 1%.

1. Aubaniac R: L'injection intraveineuse sousclaviculaire; advantages et technique. Presse Medicale 60:1456, 1952

2. Shanahan MX; Tynan A Infraclavicular subclavian vein catheterization. Aust N Z J Surg, 40:119-23, 1970

3. Pybus DA, Poole JL, Crawford MC. Subclavian venous catheterisation in small children. Anaesthesia 1982: 37,451.

4. Bonventre EV; Lally KP; Chwals WJ; Hardin WD Jr; Atkinson JB. Percutaneous insertion of subclavian venous catheters in infants and children. Surg Gynecol Obstet, 169:203-5, 1989

5. Seldinger SI. Catheter replacement of the needle in percutaneous arteriography. Acta Radiol 39:368, 1953

6. Lefrant JY, Muller L, Nouveoon E et al When subclavian vein cannulation attempts must be stopped? Anesthesiology Supplement 1998. ASCCA abstract B11.

7. Andrews RT, Bova DA, Venbrux AC. How much guidewire is too much? Direct measurement of the distance from subclavian and internal jugular vein access sites to the superior vena cava-atrial junction during central venous catheter placement. Crit Care Med 2000 Jan;28(1):138-42

8. Mantia AM, Robinson JM, Lolley DM et al Sternal retraction and pulmonary artery catheter compromise. J Cardiothorac Anesth 2:430, 1988

9. Ruggiero RP; Caruso G Chylothorax--a complication of subclavian vein catheterization. J Parenter Enteral Nutr, 9:750-3, 1985

10. Khalil KG; Parker FB Jr; Mukherjee N; Webb WR Thoracic duct injury. A complication of jugular vein catheterization. JAMA, 221:908-9, 1972 Aug

11. Mermel LA, McCormick RD, Springman SR, Maki DG The pathogenesis and epidemiology of catheter-related infection with pulmonary artery Swan-Ganz catheters: a prospective study utilizing molecular subtyping. Am J Med 1991 Sep 16;91(3B):197S-205S

12. Muhm M, Sunder-Plassmann G, Apsner R, Kritzinger M, Hiesmayr M, Druml W Supraclavicular approach to the subclavian/innominate vein for large-bore central venous catheters. Am J Kidney Dis 1997 Dec;30(6):802-8

13. Tan BK, Hong SW, Huang MH, Lee ST Anatomic basis of safe percutaneous subclavian venous catheterization. J Trauma 2000 Jan;48(1):82-6.

14. Ho AM, Chung DC, Tay BA, Yu LM, Yeo P Diluted venous blood appears arterial: implications for central venous cannulation. Anesth Analg 2000 Dec;91(6):1356-7.

15. Cummins RO. In 'Advanced Cardiac Life Support.' American Heart Association. Dallas, TX 1997.

Last edited on: 09/12/2000