Fluid overload in the septic patient

Panacea Financial Contract Review
This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.
F

fernandbteich

Full Member
10+ Year Member
Joined
Mar 13, 2011
Messages
15
Reaction score
0
Members do not see this ad.
Hi I have a question,
How do you manage patients in septic shock with limited cardiac reserve who got overloaded after multiple fluid boluses?
Do you have the guts to use furosemide?
Any studies or guidelines to help with the management?
Thank you
 
fernandbteich said:
Hi I have a question,
How do you manage patients in septic shock with limited cardiac reserve who got overloaded after multiple fluid boluses?
Do you have the guts to use furosemide?
Any studies or guidelines to help with the management?
Thank you
Click to expand...
It really depends on 100001 different factors.

Are they actually in septic shock? Note, for them to actually be in septic shock, they need to be *on pressors*, not just have low blood pressure. If they are actually in shock, what are their pressor requirements? (I've seen even ICU attendings misuse and abuse the definitions of sepsis vs severe sepsis vs septic shock, much less our general medicine guys)

Just how bad is their "limited cardiac reserve"? Whats their EF? How was their cardiac function prior to admission?

Just how bad is their "fluid overload"? Are they breathing OK? Are they intubated? Do they just have some edema or what exactly are you worried about?

And lastly, what's their renal function look like? I don't just mean their creatinine... How much urine are they making? What are their electrolytes doing?

The management is very different depending on the answers to these questions. If someone has functioning kidneys and minimal pressor requirements (hopefully no pressor requirements), of course I'd feel comfortable using furosemide. If they're intubated for their ARDS, there's excellent data to support drying them out to better allow you to extubate them. OTOH, if they have no renal function, are maxed on levophed and vaso, and are looking a bit poofy? I might consider initiation of CVVH if their electrolytes are going wonky, but I'm not exactly removing fluid at that time.
 
If your assessment is that they are fluid overloaded and you're no longer on the "happy" part of the frank starling curve then diurese. If the beans don't work then hook up an external kidney.

"Wet" patients do worse every single time it been studied. Patients in septic shock need crystaloid . . . Until they don't.

*Doctors* have to figure this out. There is no algorithm. Any "doctor" that can be replaced by an algorithm . . . Should be.
 
jdh71 said:
If your assessment is that they are fluid overloaded and you're no longer on the "happy" part of the frank starling curve then diurese. If the beans don't work then hook up an external kidney.

"Wet" patients do worse every single time it been studied. Patients in septic shock need crystaloid . . . Until they don't.

*Doctors* have to figure this out. There is no algorithm. Any "doctor" that can be replaced by an algorithm . . . Should be.
Click to expand...

Agreed. If they are truly in septic shock (to begin with) and you fluid overloaded them, then stop giving fluids and remove them (either via the beans or via machine). Pressors are your friends. And if that's not enough, sometimes they might need mechanical support. Sometimes you will need a Swan Ganz to help determine hemodynamics (although fallen out of favor, there is a role for them at times)
 
don't say swan in front of jdh. He will cite those ill designed studies and get his panties in a serious wad. I agree with PA cath if the resuscitation has lost its way and you know how to use one. a rare but valuable tool
 
sluggs said:
don't say swan in front of jdh. He will cite those ill designed studies and get his panties in a serious wad. I agree with PA cath if the resuscitation has lost its way and you know how to use one. a rare but valuable tool
Click to expand...

"Ill designed"? Talk about that. Let's have a journal club. No wad. They don't make a difference. I told you to show me benefit. Don't cry about it because you've been unable too. Until you do show me benefit it's all your own personal confirmation bias.
 
if you have a central line you can get an SvO2 and calculate a Fick to see how youre doing. if its really low i think i would challenge with lasix.
 
Trogghunter said:
Dobutamine.... 60% of the time it works everytime.
Click to expand...
I was thinking of bringing dobutamine up above. I like it as an adjunct agent too... except that when it's been studied it seems to increase mortality. That said, in the anecdata, it's a great tool to help compensate the ****ty hearts 😛

(OTOH, I've only seen a handful of patients with swans, and haven't really seen it make much of a difference in general. But part of that might be I really don't know what to do with much of the data, and neither do most of my co-residents. We use them so rarely because of the studies JDH cites)
 
smartparts said:
if you have a central line you can get an SvO2 and calculate a Fick to see how you're doing. if its really low i think i would challenge with lasix.
Click to expand...

You have to be careful with trying to do Fick. At baseline, their CO/CI will already be low. Most of the time clinical exam will correlate with your labs and you will know if your patient is wet, dry, or just right. Sometimes they contradict. Just keep in mind that the Fick Equation assumes a fixed metabolic activity and oxygen consumption (the 135xBSA component) which in reality is not really accurate (esp in the setting of sepsis)

Raryn said:
I was thinking of bringing dobutamine up above. I like it as an adjunct agent too... except that when it's been studied it seems to increase mortality. That said, in the anecdata, it's a great tool to help compensate the ****ty hearts 😛
Click to expand...

The problem with dobutamine (besides dropping SVR at low doses, which is good for heart failure but not good for sepsis) is the fact that it increases myocardial oxygen consumption - if the patient also has ischemic heart disease, that's additional demand (on top of the sepsis). Doesn't mean it doesn't have a role in this discussion. You have to tailor your treatment based on the patient's history as well as present situation.

As an aside, the Surviving Sepsis Campaign mentions it, stating "A trial of dobutamine infusion up to 20 micrograms/kg/min be administered or added to vasopressor (if in use) in the presence of (a) myocardial dysfunction as suggested by elevated cardiac filling pressures and low cardiac output, or (b) ongoing signs of hypoperfusion, despite achieving adequate intravascular volume and adequate MAP (grade 1C)"


Raryn said:
(OTOH, I've only seen a handful of patients with swans, and haven't really seen it make much of a difference in general. But part of that might be I really don't know what to do with much of the data, and neither do most of my co-residents. We use them so rarely because of the studies JDH cites)
Click to expand...

I think most physicians, including cardiologists as well as intensivists, have trouble interpreting the data from Swans - or more accurately, trouble deciphering what the data means for the patient and why I think the studies showed no benefit. It doesn't help that as there are fewer Swans being placed (even at large academic centers - with good reasons not to do them), there will be fewer cardiologists/intensivists in the future who can truly understand and interpret Swans (or even troubleshoot issues with the Swans). My personal experience also reflects what these RCTs show - that for most cases it neither helps or hurts. There are a few cases where Swans were useful (anecdotal) However, I still think it has a role in a very small select group of patients ... just have to be selective in who needs it.

Gastrapathy said:
How about intubate, pressors, CVVHD or, even better, get a consult from one of those pesky intensivists.
Click to expand...

Agreed. Perhaps having an intensivist help manage this patient might be helpful (or just consult Watson/Siri)
 
group_theory said:
You have to be careful with trying to do Fick. At baseline, their CO/CI will already be low. Most of the time clinical exam will correlate with your labs and you will know if your patient is wet, dry, or just right. Sometimes they contradict. Just keep in mind that the Fick Equation assumes a fixed metabolic activity and oxygen consumption (the 135xBSA component) which in reality is not really accurate (esp in the setting of sepsis)



The problem with dobutamine (besides dropping SVR at low doses, which is good for heart failure but not good for sepsis) is the fact that it increases myocardial oxygen consumption - if the patient also has ischemic heart disease, that's additional demand (on top of the sepsis). Doesn't mean it doesn't have a role in this discussion. You have to tailor your treatment based on the patient's history as well as present situation.

As an aside, the Surviving Sepsis Campaign mentions it, stating "A trial of dobutamine infusion up to 20 micrograms/kg/min be administered or added to vasopressor (if in use) in the presence of (a) myocardial dysfunction as suggested by elevated cardiac filling pressures and low cardiac output, or (b) ongoing signs of hypoperfusion, despite achieving adequate intravascular volume and adequate MAP (grade 1C)"




I think most physicians, including cardiologists as well as intensivists, have trouble interpreting the data from Swans - or more accurately, trouble deciphering what the data means for the patient and why I think the studies showed no benefit. It doesn't help that as there are fewer Swans being placed (even at large academic centers - with good reasons not to do them), there will be fewer cardiologists/intensivists in the future who can truly understand and interpret Swans (or even troubleshoot issues with the Swans). My personal experience also reflects what these RCTs show - that for most cases it neither helps or hurts. There are a few cases where Swans were useful (anecdotal) However, I still think it has a role in a very small select group of patients ... just have to be selective in who needs it.



Agreed. Perhaps having an intensivist help manage this patient might be helpful (or just consult Watson/Siri)
Click to expand...

I think I hear the far off sound of bagpipes my friend!
 
LoudBark said:
LR or NS????
Click to expand...
No data one way or the other. Too much NS will cause a hyperchloremic metabolic acidosis, but on the grand scale of things, probably relatively harmless compared to the patient's underlying issues. Outside of one or two tiny trials in surgical patients wrt renal outcomes and some data showing that LR has less SIRS in pancreatitis than NS, there's no real evidence one way or the other. Especially in anyone with functioning kidneys, it probably makes no difference. And if they DON'T have functioning kidneys, they have bigger fish to fry.

That said, I probably use more LR than most of my co-residents. Both LR and NS are pennies a bag, and if someone is getting a mild hyperchloremic acidosis, I have no real reason to just blindly give them NS when I can switch them.

Edit: I lied, I forgot about one trial in ICUs. There was a JAMA paper in 2012 that looked at renal outcomes before and after their ICU switched from primarily chloride-heavy to chloride-restrictive fluids, and found that their outcomes improved after the switch. It had a LOT of possible confounders though. http://jama.jamanetwork.com/article.aspx?articleid=1383234#Abstract
 
Last edited:
Raryn said:
No data one way or the other. Too much NS will cause a hyperchloremic metabolic acidosis, but on the grand scale of things, probably relatively harmless compared to the patient's underlying issues. Outside of one or two tiny trials in surgical patients wrt renal outcomes and some data showing that LR has less SIRS in pancreatitis than NS, there's no real evidence one way or the other. Especially in anyone with functioning kidneys, it probably makes no difference. And if they DON'T have functioning kidneys, they have bigger fish to fry.

That said, I probably use more LR than most of my co-residents. Both LR and NS are pennies a bag, and if someone is getting a mild hyperchloremic acidosis, I have no real reason to just blindly give them NS when I can switch them.

Edit: I lied, I forgot about one trial in ICUs. There was a JAMA paper in 2012 that looked at renal outcomes before and after their ICU switched from primarily chloride-heavy to chloride-restrictive fluids, and found that their outcomes improved after the switch. It had a LOT of possible confounders though. http://jama.jamanetwork.com/article.aspx?articleid=1383234#Abstract
Click to expand...

The interesting thing about the end points in the Australian study were the lack of any teeth in the things that would ultimately "matter" to patients. Such as length of stays, 30 (or was it 90?) day mortality, and need for dialysis long term.

I was just talking to my resident about this yesterday.

Though it is largely because of this study (warts and all) that have slowly lead me to change my practice pattern. Why? Because of resource not patient factors. These days we all need to be good stewards of our domains. And CRRT (or intermittent dialysis) is expensive and ties up nursing. In the case of CRRT many places make these patients one to one and some places will even assign a nurse just to do the CRRT making the patient a two to one or at least partial two to one.

So these days I use a lot of LR. Hell. I'd use plasmalyte if it wasn't so relatively expensive.
 
jdh71 said:
The interesting thing about the end points in the Australian study were the lack of any teeth in the things that would ultimately "matter" to patients. Such as length of stays, 30 (or was it 90?) day mortality, and need for dialysis long term.

I was just talking to my resident about this yesterday.

Though it is largely because of this study (warts and all) that have slowly lead me to change my practice pattern. Why? Because of resource not patient factors. These days we all need to be good stewards of our domains. And CRRT (or intermittent dialysis) is expensive and ties up nursing. In the case of CRRT many places make these patients one to one and some places will even assign a nurse just to do the CRRT making the patient a two to one or at least partial two to one.

So these days I use a lot of LR. Hell. I'd use plasmalyte if it wasn't so relatively expensive.
Click to expand...


Do you commonly follow serial lactic acids in these patients? And if so, do you use it to guide your fluid resuscitation?
It seems there is a big push to follow serial lactic acid every 6 hours or so until it is normal, but what is the point of following it if it doesn't cause you to manage your fluid resuscitation based on the results?
 
LoudBark said:
Do you commonly follow serial lactic acids in these patients? And if so, do you use it to guide your fluid resuscitation?
It seems there is a big push to follow serial lactic acid every 6 hours or so until it is normal, but what is the point of following it if it doesn't cause you to manage your fluid resuscitation based on the results?
Click to expand...
Well, if it isn't getting better (or goodness forbid if it's getting worse)... it means you must be doing *something* wrong.
 
LoudBark said:
Do you commonly follow serial lactic acids in these patients? And if so, do you use it to guide your fluid resuscitation?
It seems there is a big push to follow serial lactic acid every 6 hours or so until it is normal, but what is the point of following it if it doesn't cause you to manage your fluid resuscitation based on the results?
Click to expand...

I do follow lactates. But lactate is just one piece of the puzzle. And if it's clearing, then good. You've at minimum done an asaquate resuscitation from an oxygen delivery perspective but this doesn't tell you if you've given too much fluid.

Where I work we use a NICOM which uses impedance to derive flow through the aorta. It's not a perfect measurement and there can be confounders but the absolute difference in numbers you get as a percentage are meaningful, especially when your data in (a fluid bolus for instance) is a known quantity and the only changing variable over the course of a very short time window - essentially it lets me know if I'm on the steep part of the frank starling curve. I direct a lot of my resuscitation based on differences in NICOM readings before and after a fluid bolus. It's quick and dirty. And while not absolutely perfect, it is something "objective" and in my opinion better than a guess. Anecdotally, it almost always works.

If a patient is no longer fluid responsive by NICOM and still producing lactate, I do a bedside echo (will probably order a formal one too) and if cardiac function is down I'll use some dobutamine or milrinone (which ever I think makes more sense). And if cardiac function looks good I ask myself if it could be the gut (old guy with PVD?) and perhaps image the abdomen. But if I don't think it's the gut then I know the patient is in a lot of trouble. Could be such a bad gram negative sepsis with mitochondrial dysfunction from endotoxins and I just need to hang in there or something is dead somewhere (maybe more than just a single somewhere) and I let the decision makers know it's really bad and their loved one may not make it despite our best efforts.
 
Last edited:
Despite my recent pissing match with JDH over PA catheters I agree with all the principles he articulates.
1. You need a measure or measures of fluid responsiveness. There are a few. NICOM works.
2. You need a measure of whether your overall resusc is working (lactate clearance works most of the time)
3. In select cases where you resusc is not going well, you need to assess whether cardiac dysfunction is impairing your resusc. Echo is a good way of doing this, with the caveat that it gives you a single point in time, and you can't see the result of your interventions unless you are doing serial echo
All I'd add is that in select cases a PA cath can really help with 1 and 3. If everything is going ok, you likely do not need a PA cath, so a study that randomizes patients to PA or no PA is not going to show a benefit. A study that randomized patients who are otherwise doing poorly, or whose resusc has "lost its way" to PA or no PA (without option for serial echo or other advanced hemodynamic monitoring) might show benefit. Just saying...
 
group_theory said:
Agreed. If they are truly in septic shock (to begin with) and you fluid overloaded them, then stop giving fluids and remove them (either via the beans or via machine). Pressors are your friends. And if that's not enough, sometimes they might need mechanical support. Sometimes you will need a Swan Ganz to help determine hemodynamics (although fallen out of favor, there is a role for them at times)
Click to expand...

While they have fallen out of favor with many who don't use them regularly, there is a reason why the advanced heart failure/transplant community uses swans frequently and also why many of the top CCUs also like them.

For run of the mill sepsis they arent needed. But for advanced cardiogenic shock or when the left sided filling pressures arent clear, swans are great IMO.
 
smartparts said:
if you have a central line you can get an SvO2 and calculate a Fick to see how youre doing. if its really low i think i would challenge with lasix.
Click to expand...

Low index isn't really going to influence the decision for a handful of reasons. Usually these guys who have true cardiomyopathy are going to live in the index of 2.1 range. Add sepsis with the inflammatory and cardiodepressant mileu and their index is going to be low.

Furthermore, decreasing filling pressure isn't going to impact your cardiac output much. You assume there is some interventricular dependence at very, very high filling pressures due to the constriction of the pericardium but the impact is minimal. You get much more bang for your buck by increasing filling pressures when they are underfilled.
 
sluggs said:
3. In select cases where you resusc is not going well, you need to assess whether cardiac dysfunction is impairing your resusc. Echo is a good way of doing this, with the caveat that it gives you a single point in time, and you can't see the result of your interventions unless you are doing serial echo
All I'd add is that in select cases a PA cath can really help with 1 and 3. If everything is going ok, you likely do not need a PA cath, so a study that randomizes patients to PA or no PA is not going to show a benefit. A study that randomized patients who are otherwise doing poorly, or whose resusc has "lost its way" to PA or no PA (without option for serial echo or other advanced hemodynamic monitoring) might show benefit. Just saying...
Click to expand...

To play devil's advocate, what are you following inthe echo? I find the hardest patients to manage with sepsis are those with low EFs to begin with. Add severe sepsis and you are going to have a lower EF. So the EF (or delta EF) isn't all that useful. How do you use that to change management?

You could follow stroke volume or SVI but if you've actually measured this by echo you know the measurements are difficult to make. I could take a given patient and take 5 measurementts of stroke volume and they'd all be vastly different. The LVOT diameter is tough to measure and it is squared in the calculation, squaring your error. I guess, the purest way would be to follow serial LVOT VTI measurements. But as you guys who do echos in the iCU know, when tubed and positioned wrong, it is tough to get good pictures.

I agree, PA catheters in selected patients are very useful but as others have alluded to, you need to be familiar with swan management based on numbers and unfortunately, most non-cardiac anesthesia and non-cardiologists don't use them often enough to be familiar. Plus if you don't know what you are doing, you can cause serious harm with swans (PA ruptures aren't pretty to witness).
 
Instatewaiter said:
While they have fallen out of favor with many who don't use them regularly, there is a reason why the advanced heart failure/transplant community uses swans frequently and also why many of the top CCUs also like them.
Click to expand...

They like them because it moves their patient up the transplant list. LOL. I have no doubt that some of them know what to do with the numbers. However, during residency I've had a cardiologist and cardiothoracic surgeon (at one of the private hospitals we rotated in) place Swans on all their pretansplant patients and then disregard the numbers afterward.

Going back to the original question, as a simple-minded hospitalist, I do give diuretics sometimes if the patient is grossly overloaded but is otherwise stable despite having sepsis. My mantra is that it's very easy to give the fluid back 😛. For patients with all the complicated discussion up there in this thread, that's when I know to call the intensivist (sometimes the cardiologist too, but in my experience it's better to just have one of them on board -- oftentimes there are two or so possible not incorrect ways to go treat the complicated patient and they would both insist [and ORDER!!!] what they have in mind to the exclusion of the other's opinions).
 
flipmd said:
They like them because it moves their patient up the transplant list. LOL. I have no doubt that some of them know what to do with the numbers. However, during residency I've had a cardiologist and cardiothoracic surgeon (at one of the private hospitals we rotated in) place Swans on all their pretansplant patients and then disregard the numbers afterward.

Going back to the original question, as a simple-minded hospitalist, I do give diuretics sometimes if the patient is grossly overloaded but is otherwise stable despite having sepsis. My mantra is that it's very easy to give the fluid back 😛. For patients with all the complicated discussion up there in this thread, that's when I know to call the intensivist (sometimes the cardiologist too, but in my experience it's better to just have one of them on board -- oftentimes there are two or so possible not incorrect ways to go treat the complicated patient and they would both insist [and ORDER!!!] what they have in mind to the exclusion of the other's opinions).
Click to expand...

**** cards

they're like the ortho of medicine
 
If they're truly fluid overloaded then yes diurese or ultrafiltrate, but you better be damn sure you're right.

LVOT VTI is a great option, the whole rant about squaring the calculation doesn't matter if you're using the same value for each measurement when calculating the change in stroke volume.

An easier and possible alternative is carotid artery VTI. Some evidence for it but still not ready for primetime.

Also lactate is not always a measure of oxygen debt in septic patients, but I agree it is a marker of severity of illness.
 
You must log in or register to reply here.

Similar threads

DrMetal
idiopathically elevated CK patient (paging Rheum?)
Replies
6
Views
2K
dozitgetchahi
D
scoopdaboop
The consultologists
Replies
19
Views
2K
Meerkatology
Meerkatology
W
The demise of hospital based work?
Replies
12
Views
2K
chessknt
chessknt
N
  • Question Question
Dual Applicant — Advice on Navigating the Cycle
Replies
1
Views
657
gutonc
gutonc
Green_Goose
Best community fellowships in the Midwest?
Replies
3
Views
1K
AussieAg
A
Top