This one I got from a colleague of mine at a Scandinavian hospital, and it’s a scary reminder of the dangers of central cannulations. Placement of a large dialysis catheter went wrong – very wrong.

Standard procedure?
A standard procedure that by all standards went well: Easy cannulation of vein on first attempt. Good, non-pulsating backflow in the syringe. No problems on inserting the guide wire, and dilation of the vessel and insertion of the big dialysis catheter went easy – although more pain than usual was noted from the patient. Good, non-pulsatile backflow and easy injection was noted in both lumen. This was the subsequent CXR:

Anything out of the ordinary? Maybe, in retrospect, I would’ve liked to see the tip of the catheter inside the shadow of the heart – but it points in the right direction and follows the normal direction of the vessels. “There is no pneumothorax noted and the lung fields are clear, apart from a slight interstitial opaqueness in the lower right hemithorax. The catheter is well placed”, was the report from the radiologist.

But something’s wrong
The patient keeps complaining of pain that increased with respiration. She hadn’t received dialysis yet. The next day a new CXR including a lateral picture are taken:

Now, it’s easy to see the misplacement of the catheter. There is also sign of pleural fluid on the right lung. No pneumothorax. The pleural fluid could be blood, but by any standard a small amount. So, still nothing dramatic, the patient is stable and it’s almost 20 hours after catheter placement.

She’s taken down to the catheter insertion room and and the catheter is checked by attaching a bag of RL, showing a free running drip, and backflow on the line when the bag is lowered below patient level. After discussing the issue, it is first attempted to reposition the catheter. When this doesn’t work, it is decided to remove the catheter – and as a side effect they reveal the underlying problem.

Things just got real
The patient crashes. The patient quickly becomes tachycardic and BP is measured to 40/25. High flow clear fluid infusions are started until blood arrives and blood transfusions started. As the patient is slim, the right pleura is cannulated as a temporary measure. They get blood.

A small drain is placed and they get 750 ml of blood. After these initial clear fluid boluses and 2 PRBCs, the patient is again fairly stable. She’s taken to the perioperative ward, a thoracic surgeon is called and an OR made ready. She crashes again, is intubated on ketamine and sux and at the same time gets a proper chest drain which drains about a litre of blood. The patient gets somewhat stabilised again and is taken directly to a thoracic OR. During the resusitation effort, another CXR was taken:

In the OR the patient got a sternal split to get proper access to the pleural cavity and the vessels going up towards the head and arm on the right side. A lateral thoracotomy wouldn’t have given proper access in this case.

Catheter plug
They find the dialysis catheter has speared through the brachiocephalic trunk and had been sitting there as a plug. When the catheter was removed, the patient got a massive arterial bleed into her right pleural space, leading to hypovolemia as well as creating a sort of tension hemothorax pressing on the caval veins and the right heart – which were already volume depleted. This would explain the patient stabilising after draining the right pleural space. The blood that was aspirated from the catheter tip during the insertion procedure, could have been blood sitting in the pleural space.

This is a very special case where the holes in the cheese line up nicely to create a scary scenario.

It reveals the need for vigilance and listening to patient symptoms. It also shows how a bleeding chest needs a big tube – not a small catheter. In pre-hospital medicine we’re usually better at improvising. In-hospital, the big system often gets us thinking inside the box – like here, waiting for a chest tube to be avialable, when you have an endotracheal tube at hand that would do the job a lot better than a tiny drain catheter.

Still, quick and decisive resuscitation and operation saved the day. And to finish the case, here’s the post-OP CXR:

This entry was posted in Code Brown, Intensive Care. Bookmark the permalink.


  1. tony says:

    one wonders whether a bedside sono would have raised suspicion earlier, as opposed to CXR to confirm placement. a negative catheter flush during sono view of the heart would have suggested misplacement, and i suspect a hemothorax substantial enough to allow ‘free aspiration of blood’ from the cath would have shown up on sono of the RLL/RUQ.

    • Thomas D says:

      I don’t think US was used for placement. I often (but not always) use it myself, but I haven’t thought of viewing the heart while flushing the central line – excellent idea! Do you use doppler, or do you just get a flow pattern with ‘bubbles’ showing up, or do you use air/saline mix as US contrast? I’d love to hear more about that!

      Regarding blood aspiration, I’m not sure what was going on there. If the line might have been in a vessel, and then dislodged, or what. From the second CXR set with the side view, it doesn’t look probable that the catheter is in any vessel. I agree it also sounds improbable to aspirate blood on a catheter placed inside the pleural space without a huge hemothorax. I guess we’ll never know.

      • nfkb says:

        for the bubbles saline might do the job pretty well. Macromolecules (HEA) are sometimes better in difficult patients

        it’s really easy to see bubble flowing throught the right heart during an echocardiography

        • Thomas D says:

          Yes, air/saline bubble mix is usually a great contrast medium for echo. For macromolecules, did you mean HES or something different?

          • nfkb says:

            my mistake, i used the french acronym

            amidon instead of starch -> HEA = HES

          • Thomas D says:

            Gotcha. So HES actually makes for better visualisation on ultrasound?! I’ve got to try that out. Then HES might have a place in ICU after all 😉

          • tony says:

            hunt for a bag of hetastarch for a squirt of echo contrast?

            if you can dig up a 3-way stopcock, you can get improved echo contrast with the equipment from the line kit you have open, plus a ml of blood.

            short story: instead of 1ml air, 9ml saline for agitated contrast, use 1ml air, 1 ml blood, 8ml saline and agitate per usual.

            article: “Superiority of the combination of blood and agitated saline for routine contrast enhancement” pubmed ID 9950967

          • Thomas D says:

            After 6S and CHEST, the HES is just lying around not getting used anyway 😉

            No, of course I won’t use HES on a regular basis, but it would be interesting to see if it visualises better than saline flush on ultrasound. Usually, saline alone will probably do the trick for verifying CVC placement. Thanks for the link to agitated saline vs saline/blood mix.

      • tony says:

        re: confirming CVL placement with flush.

        i use this regularly to confirm deep ultrasound-placed peripheral IV catheters because i cant see infiltration and aspiration is not always reliable (could be aspirating hematoma or may not be able to aspirate freely). a brisk flush with saline (not using bubbles) is generally glaringly obvious in the proximal vein.

        i recently discussed confirming CVL placements with the folks who run the ‘ultrasound podcast’, off the air. i have little expereince with this, but can pass on waht i learned frmo them. they say that they like this a lot- and that you dont need to introduce bubbles or microbubbles like we do for echo contrast (presumably this would be even better, but then it also takes extra equipment and fiddling around to do right). seeing contrast in the right heart confirms venous placement. further, they note that as you gain experience with this you can get a feel for the timing- giving you a clue that you might have a diverted venous placement.

        • Thomas D says:

          Thanks for the update! Trust the awesome ultrasoundpodcast-geeks to be in the know! This is certainly something I will have to look into. Great tips!

      • I am currently researching my mother’s case where we believe a CVC was misplaced in the Right Femoral Artery instead of the Right Femoral Vein. If I am reading the medical records correctly the mistake wasn’t recognized until approximately 16 hours after insertion. The attending ED Doctor noted the need for “5 or more” attempts at insertion of the CVC and still didn’t use the Ultrasound method to confirm proper placement of the catheter. My mother was 68 at the time of the incident and subsequently spent 6 months in the CTICU where she needed to have her right leg amputated. Not one doctor would talk to us about why my mother’s leg was basically black and ultimately needed to be amputated. She went through an initial amputation surgery removing her right leg below the knee, and ultimately needed another amputation surgery above the knee due to the presence of infection in and around her knee. I am looking for input on what the “Standard Practice” is for insertion of a CVC and what the standard practice is when so many attempts are needed to “successfully” insert the CVC in a critically ill patient. Please email me at I am in desperate need of education on this subject

  2. Mark says:

    Interesting case, good that the patient survived. It is a reminder of the risks of such catheter insertion and that every procedure carries risk.

    With the retrospectoscope, simply removing the catheter seems a little hasty. We know it is incorrectly positioned, therefore further clarification of its actual location prior to removal may have been useful. The ability to obtain blood was obviously falsely reassuring. Tony’s point about US is good. The other option may have been CT angio of some sort, although with a patient requiring dialysis, implying poor renal function, the contrast would have been risky. Having said that, the damage to organs may have been less and more recoverable than two cardiac arrests.

    At the end of the day, a good save though. Thanks for sharing.

    • Thomas D says:

      Hmmm, I usually don’t hesitate to remove a catheter if I think it’s misplaced. Of course, a big cath like a dialysis catheter might make me more wary (especially after reading this story!). On the other hand, a big lumen catheter like that makes you feel more confident with non-pulsative flow. I guess the blood aspiration and the fact they were two specialists doing the replacement procedure, made them feel comfortable removing it. And like you say, contrast for a freak misplacement like this would be overkill.

      In retrospect it might seem like a good idea, but I don’t think many clinicians would order a CT scan with contrast before removing a misplaced catheter. I wouldn’t. Not even a dialysis catheter. Or are there other views on this? Do you have guidelines for imaging before removing misplaced catheters?

      • Mark says:

        I hear you Thomas, I know I have removed a few misplaced catheters without imaging, although vascaths/dialysis cathethers make me more nervous if they are potentially arterial because of the calibre of the hole.

        I think the fact that this was so clearly misplaced and being aware of at least one other similar story of errant lines in the thorax and haemorrhage, had me more twitchy than I would normally be and in need of greater confirmation of actual location.

        I am not aware of any guidelines in my hospital pertaining to this.

        It’s a good teaching case and your website is also very good. Mark.

  3. dr brian makamure says:

    Nice example of fancy decisive inexperienced guy would have lost the patient.hooray for quick thinking which comes with working with trauma patients.

  4. nfkb says:

    wo, scary case !

    i really don’t understand how it’s possible !? collagen disease ? diabetes ?

    an idea : plug the cell saver on the chest tube when you highly suspect massive hemothorax


    • Thomas D says:

      Hi Remi!
      I’m not sure how this could’ve happened? Could the needle have been placed through the artery, and thus subsequent dilators and catheter as well? I don’t have the background on the patient, so I’m not sure.

      Cell saver on chest tube is a great tip! Something to think about for trauma as well. Thanks for commenting.

  5. Rahul says:

    Scary indeed and I also think there must be some REALLY thin arterial walls for that to happen.
    Great teaching points none the less.

  6. mrbee says:

    A blood gas, performed on insertion of the device would clearly have shown whether the blood aspirated was indeed venous, arterial, or mixed venous. The absence of pulsatile flow is a very poor measure of line placement.

    • Thomas D says:

      I agree. Although I would think pulsative flow is slightly better for a big line like a dialysis catheter, and in a young person. But clearly not 100%. I guess the operators where also calmed down by what seemed to be a straight forward procedure from their point of view (apart from patient pain). After hearing the case, I would’ve love to have a blood gas available – just to see what it would’ve shown. Blood gas is definitely one of the good ways to affirm placement, but could the catheter have gone through the artery, and actually been in the vein and got displaced later? I don’t know. This is obviously a very special case, and you might not have been able to avoid it. But I agree, several available routine investigations might have given you a good hint.

  7. Tim says:

    Unfortunate (and almost disasterous) complication, but glad to hear of the positive outcome for the patient. Also an excellent teach scenario to highlight various other techniques that could be employed during the procedure.
    A couple of Q’s;
    Was the device was actually ‘placed’ under U/S guidance or using ECG guidance (or a combination of both), as malposition of the guidewire (on ECG) would be noted before the proceduralist even got to dilate the vessel.
    Also, tranduction of the cannulation needle prior to guidewire insertion/dilator/catheter insertion would possibly have averted this problem.
    Was the length of the catheter noted on the post insertion CXR? Generally, we should be able to see the full length of the catheter once insitu in the correct position. Its shorter length than the actual catheter length might give away a possible posterior travelling direction.
    Food for thought and thank you for the interesting post.

    • Thomas D says:

      I agree, it’s a case for safety measures and against complacency. I don’t know how the hospital from the case is set up with regards to ultrasound and other investigations for central line placement. Also, at our institution, it’s acceptable practice to place a central line without any investigations apart from CXR after placement. And we do that all the time.

      To your questions: US wasn’t used, and of course, especially when used in-plane, it would show entering into the lumen of the vein by both the cannula and guidewire. To my knowledge, ECG didn’t show arrhythmia. I know the guidewire wasn’t hooked up to the ECG (to show raised P waves).

      The length of the catheter was most probably noted, but in my hospital we seldom use that for measuring on the CXR. It sounds interesting, but would it be sensitive enough? Maybe for a jugular vein insertion, but wouldn’t a brachial vein make the catheter course winding and imprecise for measuring on CXR? Interesting point, though.

      Thanks for your thoughts and tips. All the posts here have been interesting and many of them have taught me new thinkings and techniques for line placement and trauma care. It’s what FOAMed is all about!

  8. Matt says:

    Very interesting case!!!

    I understand that clinicians place CVCs all the time without ultrasound guidance, but shouldn’t cases like this change the era of blind non-emergent CVC placement (ie this cases ? Ultrasound makes it nearly impossible to cannulate an artery after CVC placement and ultrasound training. Moreover, once a clinician is an expert at ultrasound guided placement it literally takes seconds to use it. I’m no expert (ie I’m just a EM resident going into critical care) but I will plan to use ultrasound guidance in all make cases non-emergent CVC placement of because cases just like this one.

    I had never heard or seen “anyone!!!” look at the heart while flushing a catheter to confirm intravenous placement (or even mentioning to simply get a quick ABG). The routine is non-pulsatile blood flow and a CXR that’s it. This is going to be part of my CVC placement regimen from now on. Thanks for a great post!!!

    • Tim says:

      I agree with you completely on your point in regards to once a clinician has gained expertise with US, then the procedure is quicker for the inserter and safer for the patient.
      However, just using non-pulsatile blood flow is NOT good enough for checking the vessel.
      With imaging technology and various accessories available today, there is no reason to be using old-fashioned (and unreliable) techniques.
      Vessel assessment PRIOR to device placment is advocated by many current guidelines (AVA, NICE, ESPEN, etc) and this should be the standard of care prior to a device even being chosen/placed.
      The proceduralist SHOULD be checking the vessel with US prior to insertion. This, in my opinion, is a mandatory process, and in the case of the inexperienced inserter, pressure transduction of the vessel through the cannulating needle, cannula or Raulerson syringe PRIOR to guidewire feed and dilation should be performed every time, until a higher level of experience and proficiency is is gained.

      • Thomas D says:

        I agree with the need to check your position, but in my opinion, you’re also setting yourself up for complications the longer you wait before inserting the guidewire. Any manipulation of your needle after perforating the vessel, like draw an ABG or attach a pressure transducer to your needle, brings a risk of dislodging the needle from the lumen. Especially if you’re inexperienced.

        I like to insert the guidewire to have a more secure line into the vessel. Then check it with ultrasound or an ECG lead attached to the guidewire to look for peaked P waves.

        And I agree with Matt. The routine is mostly getting backflow, check for non-pulsative flow, have a steady flow of blood from the syringe while attaching the guidewire, then easy insertion of guidewire. The rate of complications are low, but of course, the need for having a higher level of confirmation might be appropriate. Some places have mandatory ultrasound for CVCs, and we’ll definitely see more of that in the near future.

  9. Pingback: BEST OF 2013 |

  10. Pingback: The LITFL Review 099 - LITFL

  11. Pingback: Does central line position matter? Can we use ultrasonography to confirm line position?

Leave a Reply

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