The CDC guidelines recommend placing subclavian lines to reduce the risk of catheter-related bloodstream infections. Meanwhile, mounting evidence suggests that we should probably be placing lines with ultrasound guidance. Unfortunately, the ultrasound-guided subclavian can be tricky. This post describes a slight modification that could make the technique easier and safer.
Theoretical rationale behind the shrug technique
Concept #1. Subclavian is the ideal site
The subclavian has numerous advantages:
- Easier to keep clean and dry, reducing infection risk (O’Grady 2011).
- Lower risk of catheter-associated DVT, compared to jugular or femoral sites (Parienti 2015).
- Avoids tracheostomy ties or neck collars.
- Doesn't impede venous drainage from the brain in neurocritically ill.
- Leaves the right internal jugular free for transvenous pacer wires, hemodialysis catheters, or ECMO catheters.
The main disadvantage is the risk of pneumothorax, which will be dealt with below.
Concept #2. Landmark subclavian isn't a viable solution for a teaching hospital
There are numerous reasons that ultrasound is desirable for central venous access. For the sake of brevity, only one will be discussed here: the Dunning-Kruger Effect. With falling popularity of the landmark subclavian, it is rare for trainees to master it. Most often, they are trained to the point of being dangerous (Mount Stupid, figure above).
Trainees would benefit more by learning to use ultrasound guidance at the ninja level. Ultrasonography has a relatively steep learning curve, so mastery can be reached relatively fast. Furthermore, identical ultrasound techniques can be used at any anatomic location (subclavian, jugular, femoral), allowing the trainee to simultaneously master every type of line insertion (1):
Concept #3. The short-axis is preferred
The short axis may be preferable for several reasons:
- The long-axis view is difficult in obese patients, when the target vein is deeper from the skin surface (see above).
- Fanning the ultrasound probe back and forth in short-axis can easily provide an accurate three-dimensional understanding of where the needle is relative to surrounding structures. This can allow for easy re-direction of the needle, if its initial direction isn’t perfect.
- Two randomized controlled studies both found the short-axis technique to have a higher success rate compared to long-axis technique for ultrasound-guided subclavian line placement (Vezzani 2017, Maddali 2017). The larger of these studies also found that short-axis technique reduced the risk of arterial injury, likely because it allows simultaneous visualization of the needle, vein, and artery (Vezzani 2017). Similar results have previously been found when cannulating the internal jugular vein (Chittoodan 2011).
- The short-axis view allows you to make sure the needle is aimed right at the middle of the vessel. This promotes successful passage of the wire, rather than hanging up on the edge of the vein:
The drawback of the short-axis approach is that it is possible to lose track of the tip of the needle (visualizing the shaft of the needle rather than the tip). Fortunately, this error is avoidable using proper technique:
- Incrementally advance the ultrasound probe and needle in short steps. Keep moving the ultrasound probe slightly ahead of the needle, then insert the needle to keep up. By walking the needle down towards the vein in this fashion, you are always visualizing the needle tip (not the shaft).
- If you get lost, wiggle the needle slightly while fanning the ultrasound probe back and forth. Find the tip of the needle, then continue.
Concept #4. Ribs can be used to prevent pneumothorax
Ribs run in-between the subclavian vein and the pleura. These can be used strategically to prevent pneumothorax. Senussi 2017 recently published a description of an ultrasound-guided axillary vein catheter placed over the second rib (figure below). The needle is aimed directly towards the rib, so that if it overshoots the vein it will hit the rib.
Concept #5. Shrugging opens up a window to puncture the subclavian vein over the first rib.
In neutral position, the clavicle lies over the proximal aspect of the subclavian vein. Shrugging the shoulder pulls the clavicle upwards, out of the way. This opens up a window where the subclavian vein can be seen with ultrasound, with the first rib behind it.
This position and window is ideal for placement of a central venous catheter:
- The subclavian vein is large, making it easy to cannulate and pass the wire.
- Pneumothorax is unlikely for several reasons:
- If the needle goes too far, it will usually hit the first rib (which protects the lung)
- If the needle doesn't hit the first rib, it will often arch over the lung. A medial subclavian puncture overlies the dome of the lung. Unless the needle is pointed downward at a steep angle, it will tend to pass over the lung (more on this below).
- The subclavian vein at this position is generally large. This reduces the likelihood of inadvertently puncturing through the entire vessel without getting blood return.
- Shrugging the shoulder increases the angle between the subclavian vein and the brachiocephalic vein as shown below. This may promote passage of the wire and catheter down the brachiocephalic vein (rather than taking an undesirable turn into the internal jugular vein).
- Compared to axillary vein insertion, more proximal catheter insertion might reduce the risk of thrombosis (Buzancais 2016).
Technical details
#1. Site selection
The left subclavian vein is usually preferable for the following reasons:
- Lines placed in the left subclavian are less likely to be malpositioned (Tarbiat 2014, Boon 2007). The right subclavian makes a sharp turn when merging with the internal jugular (figure above), which can promote malposition. In contrast, the left subclavian takes a diagonal glide path directly into the superior vena cava, without any sharp turns. This also explains why the left subclavian site is favored for placement of transvenous pacemakers or Swan-Ganz catheters.
- A 20-cm central line can generally be advanced to the hilt on the left side without risking placement in the right ventricle (unless the patient is unusually small). Advancing the catheter completely allows it to be secured with only two sutures and a small dressing, which keeps things neat and reduces the risk of infection.
Of course, the right side may be preferred in some patients (e.g. unilateral right-sided lung disease).
#2. Patient positioning & adequate vein length
Ensure with ultrasound that the target is large enough to cannulate and wire. If the target vessel looks small, consider the following maneuvers:
- Trendelenburg position.
- Shrugging the shoulder more may help open up the ultrasound window better.
Just because it is possible to get a view of the subclavian vein doesn't mean that the procedure is safe. You should be able to obtain vision of the subclavian vein over a couple centimeters with the probe oriented vertically (as it would be during the procedure itself). This is usually easy, but it can be impossible in severely obese patients. If you can't achieve a good landing zone, choose another site (e.g. the internal jugular).
The patient should maintain this shrugging position for the entire procedure. If the patient changes position before insertion of the dilator, dilation may be difficult (because you will be trying to dilate a non-linear tract).
#3. Target selection – Avoiding the lung
When selecting your target, pay attention to structures behind the target vein. There are roughly four things that may be seen underneath the vein:
- Bone: There is often a considerable length of subclavian vein which is overlying the first rib. This is an ideal target for cannulation, because the rib blocks the needle from potentially causing a pneumothorax.
- Medial soft tissue sign: Very close to the sternum, you may encounter a view where the vein is overlying soft tissue. This is a reasonable site for cannulation. The soft tissue will provide you with some buffer if you advance too far. Furthermore, in this location you're often overlying the dome of the lung such that you cannot hit the lung (the needle will arch over the lung tissue into the mediastinum; figure below). Use caution, however, because if you go exceedingly deep there is a risk of injuring mediastinal structures.
- Vein directly overlies pleura or subclavian artery: Try to find a safer view elsewhere.
Ultrasonography provides two-dimensional images. However, by scanning back and forth you can combine these images in your mind to generate a three-dimensional understanding of the vein and surrounding anatomy. Identify dangerous structures (subclavian artery, lung) and avoid them.
#4. Needle insertion
The needle should be inserted with caution and precision. Even if bone is underlying the subclavian, do not rely on the bone to prevent injury. The first and best line of defense against complication is precise control of the needle tip. The most important technique here is coordinated advancement of the needle and ultrasound probe in small steps, with the ultrasound probe always ahead of the needle (discussed above).
Another trick is helpful here. Ideally the linear probe should be perfectly aligned along the axis of the target vessel (figure below). If the probe and vessel are aligned, then it's OK if you enter the vessel at a slightly different spot than you were initially aiming for (your needle will remain on track to hit the vessel whether the needle takes a slightly sharper or shallower angle). Alternatively, if the probe and the vessel are misaligned then it is easy to miss the vessel if you miscalculate the depth of your needle slightly (the vessel veers out of the pathway of your needle).
A simple way to make sure that your probe is aligned with the vessel is the fan test. Fan the transducer back and forth slightly. If the vessel is aligned correctly, it will stay in the center of the ultrasound screen. If the vessel is misaligned, it will shift left and right on the screen. If misaligned, take a minute to rotate the probe until it is aligned with the vessel.
Your goal is to line up the needle and vessel, using the probe as an intermediary:
Video showing the fundamentals of the shrug technique (positioning, identification of the vein & first rib, alignment of the probe).
#5. Wire placement
Threading the wire is occasionally tricky. Sometimes the wire will veer into the internal jugular vein or the contralateral subclavian vein. This may be suspected when the wire initially passes smoothly but then hangs up around a depth of ~15 cm.
Delayed wire hang-up can often be resolved by using the plastic introducer catheter to hold position in the vein and then re-inserting the wire a few times (similar to the concept of re-floating a pacer wire). With some trial-and-error (e.g. rotating the head towards the operator, the Abesh manevuer, inserting the wire with the J-tip pointing in a different direction) the wire can often be encouraged to pass more smoothly. If the wire repeatedly hangs up ~15 cm the central line can still be placed and often will be OK (either lying in the superior vena cava or contralateral subclavian, which is generally adequate).
#6. Passage of the central line
For patients with severe obesity, this can be tricky. There may be several centimeters of tissue between the skin and the vessel. If the catheter is advanced too abruptly, it can herniate into the soft tissue and end up outside the vessel:
This can be avoided with the following techniques:
- Gently compress the soft tissue as you advance the catheter. This reduces the distance that the catheter needs to pass through the soft tissue before entering the vein.
- Feed the catheter in gently.
- Occasionally palpate to ensure that the catheter isn't looping within the soft tissue.
#7. Catheter depth
This depends on three factors:
- Patient height
- Amount of soft tissue between the skin and the vein (increased in morbid obesity). For patients with excess soft tissue, the line should be placed deeper, to prevent being pulled out of the vascular system by movement of the soft tissue.
- Location in the right versus left subclavian.
In most cases, a left-sided subclavian can be advanced to the hilt (using a 20-cm catheter). This may result in the line tip sitting in the atrium, which is fine. For patients who are short and thin, a depth of ~18 cm may be better on the left side (this can still be secured using two sutures, without a lot of line flopping around). A right-sided subclavian will need to be advanced less (perhaps ~2-3 cm less deeply than on the left side).
#8. Quality control
The goal shouldn't be to perform competent procedures, but rather to perform flawless procedures. If you shoot for flawless and fall short, you'll still be doing an excellent job. Every single procedure should be critiqued using the following metrics:
Indicators of high quality:
- Low number of needle passes.
- Needle-tip control is so meticulous that you can see exactly when the needle is about to enter the vessel (e.g. you see it tenting the proximal vessel wall before it pierces though, then you watch it pop into the vessel).
- Needle enters the center of the vessel, allowing guidewire to pass easily.
Indicators of low quality (2):
- The first sign that you are in the vessel is a flash of blood in the syringe. Ideally you should already know that you're in the vessel based on ultrasonography, before blood enters the syringe (#2 above).
- The needle accidentally passes through the vessel and crashes into the first rib. This is a sign of very poor technique, with the needle-tip outrunning your ultrasound probe.
- You puncture through the entire vessel without realizing it, and only obtain blood return when pulling back the needle.
Obviously, every procedure won’t be perfect. The key is continuously critiquing technique and aiming for perfection. Don't be satisfied merely by performing the procedure successfully.
Genius General experience
The shrug line has been in use at Genius General Hospital for over a year. It has likely been placed about a hundred times, mostly by pulmonary fellows. A couple catheters were malpositioned into the subclavian or internal jugular veins, which is an inevitable consequence of subclavian catheterization (catheters inserted into the contralateral subclavian vein were left in situ and used without problems)(3). One catheter was inserted in subcutaneous fatty tissue as described above (see #7), and this was removed without sequelae. No hematoma, hemothorax, pneumothorax, or other serious complications occurred. Procedural difficulty was mostly limited to patients with severe obesity (although in such patients it is generally possible for an expert operator to place the line)(4).
- The ultrasound-guided subclavian is a widely recommended and well-accepted procedure. The shrug technique represents a minor modification on this procedure, which may improve speed and safety.
- Shrugging the shoulder upwards opens up a proximal window where the subclavian vein can be visualized and cannulated.
- The success of this procedure may be easily estimated before starting, using ultrasonography. If the views are of dubious quality, try finding a better site.
- The ultrasound-guided subclavian central line placement depends on very precise ultrasound-guided needle control, so this isn’t a novice-level procedure. However, this is a technique well worth learning. Once mastered it will allow the operator to place any central line with ease (subclavian, jugular, or femoral).
Thanks to Scott Weingart for encouraging me to video this and producing the video above.
Notes
- A similar argument can be made for anyone who isn’t constantly putting in central lines. For example, many attendings at training hospitals may put in 25-50 lines per year. If these are all placed in the same technique, it will be easier to maintain competency in that single technique. If these lines are divided between three or four different techniques, it will be difficult to maintain competency in all of the different techniques.
- Please note that “low quality” indicators are still compatible with a safe and effective line insertion procedure without any harm to the patient. However, if you are consistently hitting low-quality indicators, that suggests that your technique needs work.
- It's debatable whether a central line in the contralateral subcalavian vein needs to be replaced. If the catheter isn't being used for anything really irritating (e.g. chemotherapy, hydrochloric acid, TPN) then it is probably fineto leave it in place and use it. For any individual patient the risk of repositioning the line (increased infection) or inserting a new line must be weighed in comparison to the risk of using the catheter in situ (which seems to be negligible).
- An ultrasound-guided subclavian isn't the recommended site for a patient with supermorbid obesity. However, patients are encountered without other good options (e.g. internal jugular veins are occupied with a hemodialysis catheter, stenosed, or clotted).
- PulmCrit wee: Why I like central lines for GI bleed resuscitation - March 13, 2024
- PulmCrit wee: Polypharmacy in the ICU – when in doubt, deprescribe - January 30, 2024
- PulmCrit hot take: VAP prophylaxis (PROPHY-VAP trial) - January 22, 2024
Hi, Great post. Enlightening and worth emulating for sure. I will also see in next few lines if I can ”shrug” it. till now i have either used the traditional landmark approach or the following technique. I have used a long axis approach for subclavian or better said the axillary vein. Near the edge of the pectoralis major the linear probe is aligned first in short axis, the artery and vein are identified by pressing on vessels and colour doppler. then carefully, the vein is brought in center of view and the probe rotated to ”open up” the vein into… Read more »
Thanks. I prefer the short-axis technique, but any technique that you can master is great. Also, there is no law that says you can’t go back and forth between short-axis and long-axis during the same procedure! The shrug maneuver should help, regardless of what technique you prefer (short-axis or long-axis).
Great post, thanks! As a sidenote: the Dunning-Kruger effect seems not to be real, or at least not supported by the original studies, see http://www.talyarkoni.org/blog/2010/07/07/what-the-dunning-kruger-effect-is-and-isnt/
Thanks, that’s really interesting. Dunning-Kruger is undoubtedly an oversimplification. It’s possible that this has largely to do with sampling error: – Trainees will often have success with their first few subclavian lines, this may cause them to assume that they will continue to have the same success with subsequent lines (thereby over-estimating their competence) – Some trainees may *fail* at their first few subclavian lines, this may cause them to over-estimate the task difficulty and under-estimate their competence. – Only after doing a LOT of procedures is it possible to get an accurate sense of the procedural difficulty and potential… Read more »
Great post. I would like to add that the position of the J-tip is also important.
If the J-tip points down towards the heart it seems that 97% of the cathethers goes towards the heart, whereas only 57% if the J tip points up. (Anesth Analg 2005;100:21–4)
Excellent point!! I was always trained to point the J-tip that direction, and that’s always what I’ve taught folks. I’m glad to see there is some evidence to back it up (honestly I wasn’t sure whether it was voodoo or not).
Thanks for a great article on CVC techniques!
Actually, by using real-time US guidance, “J-tip directions” and “right-vs-left-sided-approach” become outdated, landmark-related stories. With US, a misplaced guidewire can be repositioned immediately.
As part of a research project on mechanical complications related to CVC insertions, we’ve created 3 videos showing our preferred CVC insertion techniques. By using the approaches we show, misplaced catheters will hopefully become remnants of the past!
https://youtu.be/DyeLjNF-PtA Microconvex probe, subclavian CVC
https://youtu.be/XVJZ7QYlKjg Microconvex probe, IJV
https://youtu.be/YoeazT4us1o Linear Probe, subclavian CVC
All the best from Sweden
Ola Borgquist
An excellent piece, thank you for the clarity of explanation. I will look at the application in my paediatric patients. A plea. The paragraph on Dunning Kruger. Unfortunately you are amongst many who are wrongly using a graph from a comic as a visual representation of the DKe not a graph from the paper as shown below. . Neither Dunning nor Kruger have EVER used the phrase “Mount Stupid” for the very simple reason that “peak” does not exist in their research. The correct graph (from their research paper) shows the disconnect between true and perceived abilities for ALL subjects.… Read more »
See above comment on dunning-kruger: https://emcrit.org/pulmcrit/shrug-subclavian/#comment-274739. This isn’t intended to be taken literally.
Like always, thoroughly enjoyed the post. Appreciate your hard work.Thank you.
Hmm, I am actually in favor of teaching the blind technique. With the new guidelines, it is quite possible that enough subclavian lines will once again be performed to ensure competency. And, I feel that it is important for doctors to be able to place at least one type of central line without ultrasound guidance. As an alternative, I propose this solution: the use of a micropuncture kit and it’s 21 gauge needle and .018 guidewire from the vascular surgery operating room. With this smaller needle acting as the equivalent of a “finder” needle used in the old blind IJ… Read more »
Was hoping to avoid the US vs blind argument as this has been done a lot. With the existence of intraosseous access and ubiquitous access to ultrasonography, I think landmark guided lines are going to become a thing of the past in the next 20-30 years. Ultrasound beats landmark in terms of speed and reduced complication rates. Furthermore, ultrasound teaches a single technique (needle-tip control) which facilitates rapid learning of how to place a line anywhere & facile retention of this skill set. In contrast, it’s unrealistic to expect every provider to learn several separate techniques for how to place… Read more »
Hmm, Perhaps my points are specific to trauma codes. Blood CAN be given IO, but it is really hard to bolus it in rapidly. U/S may be available, but in a critical trauma with 10 people around the bed and femoral access not an option due to penetrating thoracoabdominal trauma, how do you wheel the u/s machine into place? You can’t- there is no room and no time. You have to attempt a blind subclavian if the nurses can’t get an IV. I think the key is NOT to expect expert level skills at getting it in perfectly on the… Read more »
I don’t do trauma, but I’ve been in a lot of crowded ED & ICU rooms. If there are so many people that you can’t cram in an ultrasound machine, then there are too many damn people in the room and the extraneous staff should be kindly dismissed.
I don’t buy the “no time” argument either. The amount of time it takes to set up an ultrasound will generally pay off in the form of fewer attempts, reduced procedure time at the first site attempted, and reduced need to perform a procedure at another site if the first fails.
I realize this is 2 years late…..but in a trauma/code situation where you can’t get an US machine into an ergonomic-ish position, the whole new generation of handhelds (eg Butterfly IQ) are very handy and work very well for vascular access….
Thanks, great post. In the past I have tended to use the IJ over the subclavian in the awake patient as achieving adequate local anaesthesia is difficult particularly with a landmark technique. Using this technique would certainly facilitate using local anaesthesia making it more tolerable and safer. In the intubated patient I wonder how easy it is to maintain the shoulder in the “shrug” position? Generous head down (if able) would probably facilitate as would another set of hands.
I actually find it easier to maintain the shrug position in intubated patients. I will often deepen their sedation (e.g. by increasing the propofol infusion or giving some ketamine). You can pull their shoulder up and it tends to stay there. There isn’t a need for someone else to hold the shoulder.
This may actually be harder to do on awake patients, as they need to be reminded not to move their shoulder during the procedure.
Do you prefer not to use this technique in awake but delirious patients? Often from shock requiring vasopressors but don’t yet need to be intubated. Or do you give some midazolam and go for it in those cases?
Could you post a video or a drawing of the best way to achieve and maintain this shrug position? Seems to me this is the biggest issue in our patients who may be in extremis or unconscious…..
We’re working on better videos. Achieving the shrug position is easier the more somnolent/unconscious the patient is. The position you’re moving the patient into isn’t inherently uncomfortable or very extreme, so once you shrug up the shoulder it will tend to stay there if the patient is unconscious.
Great piece, here are a few thoughts. I use long acid for the following reasons :, veins can curve, seeing it in LA helps you see the “runway” better. I also think visualizing the needle pop into the vein is much safer in LA. I very commonly see the anterior wall tent down until it opposes the posterior wall, and when the needle pops it is at the posterior wall. Pulling the needle back to the middle allows the GW to thread better and safer. When there is a lot of tenting I also flatten the needle to almost horizontal,… Read more »
There are many ways to skin the central line cat. If you really love the long axis, this is potentially compatible with the shrug maneuver as well (i.e. you could use the same target & same position, with a long-axis orientation instead of a short-axis orientation).
Thanks for the great post! The shrug position seems useful, but may be difficult to use my patients (intubated/polytrauma/critically ill/septic shock patients in ER).
What other modifications would work in your experience and opinion? What do you think about passive leg raise 40 degrees, trendelenber position?
Just discovered this great post! A couple of questions though.. (1) you reference Kitagawa for the advantages of shrugging – but the conclusion from their article was that having the shoulder in the DOWN position was safer and more reliable than UP or neutral. (2) also, how do we reconcile this with the anecdotal observation that pulling the arm and shoulder down often helps our guidewire to thread when it hangs up?
Excellent article. Thank you Josh. Agree w/ everything above.
Some observations from my own practice having done over 250 of these;
Hello,
In short axis how do you suggest to move the probe in order to track the needle tip? sliding or fanning?
I am used to slide the probe in the IJV but in the subclavian is this still possible having the clavicle in the field? I understood that in the subclavian landing zone is crucial.
Thank you,
Giacomo