Failing Oxygenator

Case Report-

Failing Oxygenator…


Aortic Valve Replacement, End to End Graft from Aortic Root to Ascending Aorta.

Pt. Data

  • 55 years old
  • Gender excluded
  • 71 inches
  • 95 Kilograms
  • 2.15 BSA
  • Baseline Hematocrit- 48%

Pertinant Hx:

  • Unremarkable.
  • No coagulopathic issues
  • No antiplatelet regimen
  • Severe AI
  • Mild AS
  • Dilated Root
  • No other issues-  patient appeared to be very robust and healthy.


AVR with possible root replacement, possible deep hypothermic circulatory arrest.

Prime Constituents: (After RAP)

  • 300 ml  Normosol
  • 5,000 units heparin
  • 25 g Albumin
  • 100 g Mannitol
  • 50 meq NaHCO3
  • Prime volume estimated at < 1000 cc’s.


  • Oxygenator: Good Hx
    • Membrane surface area- 2.5 m2
  • Circuit:  Customized
  •  RAP and ultrafiltration concomitant to CPB
  • 1% Forane on initiation of CPB
  • 85%  FiO2 3 liter sweep gas
  • 4.5 to 5.5  Blood Q
  • Cool to 32 C
  • Retrograde Cardioplegia for onset and continuation of myocardial preservation.
  • CSP Measured via CVP to be maintained at 50 mmHg with 1000 cc initial retrograde dose. KCL:  80 meq/L for initial arrest- reduced to 20 meq KCL/L for subsequent doses of 400 cc’s.



  • On Pump
  • Cooling
  • + 2 minutes:  Aortic crossclamp
  • + 5 minutes:  Cardioplegia delivered at 2 degrees C- x 4 minutes, Retrograde via coronary sinus- CSP pressure approximately 40-50 mmHg.
  • Heart arrests at 600 or so cc’s of cardioplegia- the balance to 1000 cc’s is given.
  • FiO2 is not adjusted for temp- remains at 85%.
  • + 10 Minutes:  the first blood gas drawn is unremarkable.
  • Observation:  CDI PO2’s seem to be in the 270’s or higher.
  • Observation Notice that CDI monitor shows PO2 dropping (<200) , SVO2 >70
  • + 12 minutes:  CDI monitor shows PO2 level dropping further- hovering around 150 or less.
  • Status:  4.8 lpm / 33 degrees C  / FiO2 @ 85% / SVO2 68%
  • + 13 minutes or so…  STAT ABG

ABG Results:

  • pH             7.47
  • pCO2        60
  • pO2           237
  • SO2%        99.9
  • Hct             33
  • HCO3        44.4 (initial Bicarb in prime)
  • BEb           18.6  (ditto)
  • K               6.48  (Arresting dose)
  • Gluc         148
  • ACT         434


I immediately bumped up my O2 sat to 100%, and drew another sample- not because the blood gas results were alarming- but because the CDI values kept dropping.  To be honest- I was concentrating so hard on the PO2- the PCO2 escaped me until the anesthesiologist on the case mentioned it.  His point was well made… “well the pO2 isn’t scary (the gas showed 237- but he CDI was less than 150)– but the pCO2 is 60″.

When he said the pCO2 is 60– well that got my attention.  At that point all the alarm bells were starting to go off.   I increased my gas flow from 3 liters to 5.

The inability to blow off CO2 in tandem with a substantial decrease in the ability to oxygenate properly, O2 lines intact and demonstrated to NOT be a delivery issue- well the next stop is to look at the oxygenator.

Now the oxygenator has my total focus.  Started running consecutive ABG’s one after the other.  Just wanted to be sure.  Continued my dialogue with the Anesthesiologist (a truly great resource- in this particular instance) and started to think of alternatives to the current oxygenation issue.

Drew another gas … (these results are 4 minutes after the previous ones)

ABG Results:

  • pH             7.37
  • pCO2        48
  • pO2           139  (from 237 in 4 minutes)
  • SO2%        99.5
  • Hct             34
  • HCO3        29.1
  • BEb           3.9
  • K               7.08
  • Gluc           151


It’s seems that the increased respiratory bump worked (in terms of blowing off the CO2).

However-  the PO2 dropped.  So now it becomes apparent that there is an issue.  The question is- which avenue and how to deal with it?

We are 20 minutes into a case that optimistically will be a 3 hour pump run, possibly longer.  Pretty serious problem when you are at the bottom end of knowing what the true issue is.

A relatively new (to me) oxygenator in use here, that has demonstrated some serious attitude from quite a few prior runs.  I have confidence in that.

However, things being what they are, being honest and believing your eyes and senses, are the quickest way to success.  There is absolutely nothing in my clinical experience that can identify what is going on with the oxygenator / patient relationship- other than it  is is beginning to fail in a really serious way- in a pretty serious hurry.

The fact that I don’t know why-  really doesn’t matter.  It’s not a Board question.  What are you going to do about it- is in the best of terms- the the clinical question that begs the answer.

The baseline considerations have in my mind been reviewed, back to the bottom line- the simple question is…  Can I get through this case continuing on the same course?

The answer  is- clearly- NO…

Here is the run down on that:

  • Notified the surgeon
  • He asked “what are you going to do? “
  • Switched to 100% O2
  • 6.0 LPM gas Q
  • bumped up blood Q to 5.6 LPM
  • Pt Pressures maintained at 60 mmHg
  • I have a plan- so I tell him- “I’ll fix it”.
  • SVO2 has dropped to 65%- (and we’re at 33C)


It’s always a great plan if you can fix it…


  • 1/4″ “Y” connector  x 1
  • LV vent tubing (has 3/8″ boot- with 1/4″ inch tubing on both ends)
  • 20 Blade
  • 2 clamps
  • 1 Oxygenator
  • 1 portable O2 tank


Since the oxygenator was functional in terms of blood flowing through it – (the ability to regulate patient temp and generate a cardiac output / pressure), my initial feeling was that if I didn’t have to change it out I wasn’t going to.

Much earlier in my career, a really good perfusionist (Chet) had convinced me that an in-tandem piggy back oxygenator was a solid option for dealing with oxygenator failures.  I had tried it a couple of times since his sage advice, and it has always worked.  The concept is simple.  Cut in an oxygenator via your recirc line, and basically hyper-oxygenate your venous side of the equation.  The rerouted blood is oxygenated and dumps back into your circuit via 1/4″ port to your venous reservoir.  Since it returns into your reservoir, air is NOT an issue.  So it becomes the simple task of splicing into the circuit 1 time, and making 2 connections.

Not exactly a veno-veno ECMO, but similar in concept.

And that’s exactly how it went.  It takes about 2 minutes once you have all your ducks in a row.  The total bonus is that there is no interrupted flow to the patient.

But it’s definitely a “gut-check” moment, as you wait to see if it works and that you made the right decision.


Well, It Worked.

ABG Results:

  • pH             7.41
  • pCO2        39
  • pO2           273
  • SO2%        99.9
  • Hct             33
  • HCO3        25.8
  • BEb           1.8
  • K               7.19
  • Gluc           160

The rest of the case went smoothly, the pump run was a little less than 2 hours, and the adjunct oxygenator performed very well.  ACT’s were maintained at >600 levels for the duration of the case.

Let me know what you think- if there was anything else that could or should have been done ?


Click image to enlarge

Space 1

32 thoughts on “Failing Oxygenator

  1. Is it possible that your Forane vaporizer might be venting gas to the atmosphere. Also, were you blowing CO2 onto the field? This might have augmented your PCO2s.
    Just a thought,
    Britt McILwain MPS, CCP

    1. I agee with Mr.McILwain, i had that happened to me once and when i checked the vaporizer, there was a leak. problem was fixed after i bypassed the vaporizer. Clogged up fibers or malfunctioning ones could be another thing. I like your piggy back idea……..

      1. Right on the spot in terms of checking things out. I would agree totally with that- and asked our O.R. director to have the vaporizer checked out.

        But don’t know if that happened.

        Good point – because that pump hasn’t been used since.

        I know there weren’t any leaks in the O2 line- because after the case I pressurized it to 8 LPM to test it- It worked fine.

    2. It was a non-coronary case- no CO2 was flushed up-

      I understand the direction you are going in- but the environment was totally familiar in terms of what to expect- Even when we do flush with CO2- it isn’t something we see (elevated CO2) during a bypass run…

      I don’t believe it was an issue here 🙂

  2. I agree with the other guys regarding the vaporizer. I have seen this occur and that was the issue, can’t say if that is true in this case.

    At first, the pO2 wasn’t alarming to me, especially if you were running a possibly undersized oxygenator, like the RX15, with this patient. The CO2 was alarming, whcih being an AVR I would think of CO2 filed flooding pushing the CO2 up.

    Anyways, I haven’t seen an oxygenator fail that fast, which would lead me to believe it was a gas delivery propylene to the oxygenator, hence the gas line and anything inline which would be the vaporizer.
    My first thing would have been to shut off vaporizer or just bypass it. That fixed the problem when I saw this.

    Recic. line for re-oxygenating blood works fine if you just need to supplemental oxygenation, but if the oxygenator totally failed, you would have had to replace it. Had I had a pO2 near 100mmHg at FiO2 of 1.0, I would have elected to come off and change it out. At 32’C, you would have plenty of time should you have to do it. Fortunately, placing a oxygenator in reciec line worked in this case.
    Let us know how the vaporizer checks out.

  3. It’s always much easier to think these things through when you’re sitting back plucking away at a keyboard! I think that’s a great idea cutting the tandem oxygenator in. There are two things though that I would consider.
    First, I run RX15’s on about 95% of my cases and I admittedly push them to the limit. What I have found is this oxygenator does not perform well when you push it’s upper end (flow) with Hct > 30. I have found the optimum Hct is around 26-28. Having a minimal boundary layer is critical to maximizing the performance of this oxygenator. I have even diluted patients down to a Hct of 28 when I have experienced this. We get so focused on running high Hct’s I think we forget about capillary perfusion and oxygenator performence in the setting of high Hct’s.
    Second: did you actively cool to 32 degrees? If so you may have had platelet aggregation lay down on the membrane. Warming a bit may have reversed the platelet aggregation and then just allow your temp to drift down to your desired target.
    Thanks for sharing your case.

  4. Great case report Frank. Thank you for sharing your experience. Never tried the tandem oxygenator like that but will keep it in mind as it sounds like a great option. I’ve had the same problems the others have listed above about the vaporizer clogging the system and as soon as we turned it off there was a visible difference in the blood even before we were able to test it.

  5. glad this worked for you. was involved directly with one case where this worked and indirectly with another where this didn’t work (failure was too complete in primary oxy). i think chet might have been around for one of these.

    i first heard of the idea for cutting an oxy into the recirc line from an excellent female perfusionist in London (UK). she told the story of having a registrar trip on her oxy and snap off the plastic connector for the water lines. she added a 2nd oxy in the recirc not for oxy failure, but for the heat exchanger.

  6. These are moments when you seriously wonder why it was you chose to go into this profession. The bottom line is whatever corrective measure you chose, if it solves the problem and the patient gets off bypass safely … you made the right decision.
    In my opinion, it probably was not the vaporizor because you had a 210 pound patient with a gas flow of only 3 liters. If it was the vaporizor, which was turned on as soon as you went on bypass, you would have noticed color changes in the arterial blood and a more rapid drop in PO2 long before the 15 minute mark. Also, by increasing your sweep gas flows from 3 to 5 liters you had an immediate response (2 minutes) which dropped the PCO2 from 60 to 48.
    It would also not have been an oxygenator high pressure excursion because you used albumin in your prime. If it was HPE, the platelet clumping on your membrane would have caused the gases to be much worse, much quicker, and your systemic pressures would have been dropping dramatically because the blood flow would not have been getting by the oxygenator. This could have been verified by doing a CBC and checking your platelets because they drop by 80-90% during HPE (they all clump in the oxygenator). This is one of the reasons why changing out an oxygenator during HPE gets an immediate response in gases … the oxygenator that was thrown out has all their platelets.
    CO2 flushing would also not have had a dramatic effect on the PaO2 and arterial saturations.
    I place my money on the side of anaesthesia. If this was a valve patient and your anaesthetist was using fast tracking, the patient may have been getting light.
    They don’t have to actually wake up and wink at you to consume vast amounts of oxygen, especially a guy who is over 200 pounds. They simply have a much greater oxygen consumption and CO2 production.
    If you had a BIS monitor you could have seen them getting light. I have seen many times over the years where I asked the anaesthetist to give a little more sedation and the venous sats would slowly come back up.
    As you know, there is also a drug that is becoming very popular with anaesthesia. This is Methylene Blue. If this is given during the case you can watch as venous sats drop to the floor (it is transient) and the blood turns black. But it does not sound like it was used here, especially that early.

    1. very interesting comment- and now has me wondering if it wasn’t a platelet aggregation issue. Since we RAP I have started priming with crystalloid only- and adding the prime constituents after chasing out the crystalloid- that means blood is hitting the fibers before the albumin does. It has been a subtle shift as when we first started aggressively rapping- I still kept about 10 g of albumin circulating in the prime- but then I switched to all crystalloid. ???

      The albumin hits the membrane about 40 seconds after the blood does

      But as you also mention- if that were the case- blood flow would have been impeded- which it was not. I may revise the RAP method- and put a small amount of protein into the circuit- before RAP.


  7. 1. Recic. line for re-oxygenating blood .-> Wrong! a L->R Shunt is the last thing you want, which is what you are doing by using the recirc line in this manner.

    2. Doubling up oxygenators in series? Try in parallel next time, it works without increasing the pressure drop. And can be employed easier IMHO.

    Failing oxygenators during a routine non ecmo case? In 25 yrs, the few I have had to troubleshoot are NOT oxygenator failures, but as others have pointed out, high demand, roller pump under occluded (O2 delivery issue), vaporizer loose fitting, gas line filter occluded

    1. Well I didn’t use a roller- the gas line was NOT clogged- Knowing the type of patient, the gravity of the case, and the anesthesiologist as I do- there was sufficient anesthesia on board. He doesn’t mess around.

      The 3 LPM of 85 % was certainly enough to keep up with this guy at a 33 C temp- and the oxygenator was not undersized for the patient demand (2.5 m2).

      I appreciate your insight regarding a left to right shunt- but it worked just fine. It’s not really a series either (by pure definition of the concept)- i just bumped up Q to accommodate the shunt.

      I agree with you that a parallel is preferable to a straight series.

      Q and Line pressures to Pt were not significantly altered. So your “pressure drop” was a non issue.

      But thank you again for your straight forward assessment.

      1. Reread the case synopsis, and I do realize that sometimes the time elements do get skewed when you are trying to trouble shoot things.

        But IMHO, I would not have blinked with a pCO2’s of 60 with a ~ 0.6:1 GBFR and a saturation of 99.5% with less than 100% FiO2.

        Much less making a decision to to splice in an oxygenator with blood gas data taken 4 mins after a barely getting to a 1:1 GBFR and 100% FiO2…

        Sorry just my opinion

        1. No apologies necessary. I wanted opinions and options and all of you have given those on both counts. 🙂

          Thanks again for taking the time to respond. I have learned something from it- and I hope others have as well.

        2. I agree with Skeptical. I would have done nothing. Well, not NOTHING, but the steps you already took. Did you notice any plasma fluid or exudate in the gas exit? Or “SIGH” the oxygenator? This is from old membrane days, but works with hollow fibers too. I never ran silicone membranes, but old guys told me this, and it works.Turn it up to around 9 lpm gas flow, and gently tap on it, and sometimes lots of fluid will dribble out and improve transfer. While 60 PCO2 is troubling, it’s not dangerous. In fact, it’s pH stat at some places.The added risk of changing something on bypass, as well as the risk of infections, outways the gains in this situation. If it’s not really REALLY broke, don’t fix it.

          I was puzzled as to why you did not trust your CDI? It’s as accurate as a blood gas analyzer when calibrated with gas and zero’ed with a first gas. Always trust your instruments. After the x-clamp is off, I would have asked the surgeon if anesthesia could shallowly ventilate without interfering with his work.

          The only time I every changed an oxr was a 6 hour pump run,on a young 2.4m2 patient, during which the PO2’s dropped to high 50’s. I was sure we would go on if he crashed, being so young. The surgeon asked me if the oxygenator was failing, I said no, it’s sub-optimal (heh heh). Lot’s of people live on PO2’s that low and high PCO2’s every day. We came off, and I changed the Oxygenator only, in case we went back on. What really suprised me was how damn slow I was, even though I knew the procedure and had everything ready. I pretended I was desperate, and timed myself to see how fast I could do it. First, I couldn’t get the oxr to come off the cardiotomy like it was supposed to. So I did what you did, lay it on the floor and use longer tubing. I think that’s a much more realistic approach, which saves a lot of valuable time in a crisis that might only happen once or twice in 20 years. .

  8. I read through all the blogs and i enjoy listening to all of you fellows and i try learn from you.
    I liked franks idea as i mentioned in my last blog about using the recirc.line although never thought of it before and i don’t see why not if it works, this is almost similar to partially opening the recirc.line to hyperoxygenate the blood as one of the alternatives when PO2 drops after exhausting all other means to raise the PO2.
    As far as L to R shunting using this technique, i would diagree with Skeptical, because this is in vitro and not in vivo. L to R shunt inside the human body will cause and alter hemodynamics but this could be avoided in our circuit when you place your flow probe past the ALF and measures exactly the flow to the pat.
    Changing out an oxgenator is an option (last option) when you are certain you are having a failed oxygenator and the pts heart is beating ( before X-clamp),then you can put the volume back in the pt and ask anesthesia to ventilate until you exchage the old oxygenator with a new one or you may parrellel it. othewise i would limb off if the heart is arrested and cool down.
    my two cents and i might be wrong, once again thank you for sharing this case with us.

  9. Great idea since I work alone. I have had to ask anesthesia for a little help prior to coming off CPB on a really large patient (300+lbs). Quick easy and no real air worries. Thanks for teaching this old dog a new trick.

    Sent from my iPhone

  10. This is a pretty good idea. One person could do it and not have to worry about any systemic blood flow interuption.

  11. Frank,

    Believe it or not, I attended a meeting (probably 15 or 16 yrs ago) where an individual said he had accomplished the same thing using a hemoconcentrator. He said he simply placed it in-line as you describe (recirc line to open venous reservoir) and attached 100% oxygen to the effluent line (removing the plug at the other end, allowing for gas venting) and said that he was successful in meeting the patient’s oxygen requirements to the end of the pump run. Although your method is more expensive, significantly increase surface exposure and displaces more patient volume, I think I would be inclined to go your route, especially with today’s new generation of low prime oxygenators with integrated filtration (like the Terumo FX 15).


  12. I too have had similar issues, pO2 and SvO2 low while pCO2 on high side on each sample (no cdi at this hospital), but managed to wean off. The problem seemed to reoccur over time randomly. I checked with every possible cause, changed blender had gas sources anaylzyed, and finally isolated the cause as human error from lab personnel who failed to calibrate their machines.

    1. I think the cause of the malfunction is due to an incorrect assessment of the case and can be read in this data. High Ht , poor hemodilution due to the RAP (low and priming ? Ultrafiltration on? ) High HCO3 , poor air washing , decrease in temperature. All have contributed to the creation of the problem.

  13. Many people have mentioned vaporizer blocked or clogged, however I had witnessed a similar incident a few years ago where the holder was just slightly decoupled from base and caused a leak rather than full block. This may have explained the somewhat gradual rather than acute decline. If the bottom of your vaporizer is even slightly resting on something (in our case the S3 control panel) and gets bumped it can break the seal.

  14. I placed extra oxygenator in the same way that you mentioned in the article In the middle of EEC I clamped main Ox and run the secondary Ox. I had to interrupt this way because the venous line was full of air. When I established normal Ecc the venous line was without air. I did this several times and always happened the same. I think if you use recirculation line it interfere with venous drainage. I use a Quadrox oxy.

  15. Sorry I wrote bad my email. I placed extra oxygenator in the same way that you mentioned in the article In the middle of EEC I clamped main Ox and run the secondary Ox. I had to interrupt this way because the venous line was full of air. When I established normal Ecc the venous line was without air. I did this several times and always happened the same. I think if you use recirculation line it interfere with venous drainage. I use a Quadrox oxy.

Leave a Reply