0
PART 1: Fluid selection for resuscitation of hypovolemic, hyperkalemic renal failure
0
Introduction with a case
0
An elderly man on an ACE inhibitor and NSAIDs presents to the emergency department after a few days of severe norovirus gastroenteritis with a creatinine of 5 mg/dL, a bicarbonate of 15 mEq/L, and a potassium of 6.5 mEq/L. He has been having diarrhea and hasn’t taken anything by mouth aside from medications for days. For the sake of simplicity let’s just say he’s 6 liters volume depleted. How should this man be fluid resuscitated?
0
I’m only going to discuss three fluids here: normal saline (NS), lactated ringers (LR), and isotonic bicarbonate (D5W with 3 ampules bicarbonate per liter).6 These fluids are available everywhere and they are all that you absolutely need. Other balanced crystalloids (especially Plasmalyte, with a Strong Ion Difference1 of 50mM) may be good choices as well but I’m not going to discuss them because I don’t have access to them at Genius General Hospital.2
0
Start with isotonic sodium bicarbonate
0
Isotonic bicarbonate is a reasonable fluid for the initial component of the resuscitation, as this may improve his acidosis and hyperkalemia. About two liters of isotonic bicarb will likely bring his bicarbonate up to a normal range. Calculating the bicarbonate deficit may provide a rough idea of the amount needed (e.g., use MDCalc to estimate bicarb deficit in mEq and then divide by 150mEq/L to estimate the number of liters of isotonic bicarbonate required).
0
Bicarbonate administration has a bad reputation in critical care, largely due to inappropriate use in lactic acidosis. A recent Cochrane review found no high-quality evidence regarding the use of bicarbonate for acute renal failure. However, bicarbonate is an accepted treatment for acidosis in acute renal failure. Given that acidosis is an indication for dialysis, treatment with bicarbonate may avert or delay the need for dialysis in some patients.
0
The use of bicarbonate for management of hyperkalemia is highly controversial and some feel that its utility is entirely a myth. I think it’s a bit more complicated than that based on two factors:
0
First, infusion of hypertonic solutions increases serum potassium. This occurs with both mannitol and hypertonic saline, and appears to be a consequence of the tonicity (rather than, for example, the effect of hypertonic saline on pH; Conte et al. 1990). Various explanations for this phenomenon include membrane damage with leakage of potassium out of cells or a “solvent drag” phenomenon. Regardless, ampules of 8.5% sodium bicarbonate are extremely concentrated (with an osmolality of 2,000 mOsm/liter), which likely acts to increaseserum potassium and counteract the beneficial effect of the bicarbonate. This may explain why studies have uniformly found 8.5% sodium bicarbonate to be ineffective in treatment of hyperkalemia.
0
Second, bicarbonate infusion appears to be ineffective in the absence of significant acidosis. Blumberg et al 1988demonstrated that even isotonic bicarbonate was ineffective when starting at an average bicarbonate concentration of 22 mM. The two most notable studies demonstrating efficacy of sodium bicarbonate for treatment of hyperkalemia both involved administering isotonic sodium bicarbonate to patients with significant acidosis (Fraley et al 1977and Blumberg et al 1992). Blumberg et al. used a regimen consisting of 8.5% bicarbonate for one hour followed by 1.4% bicarbonate infusion subsequently, and it was only after switching to the 1.4% bicarbonate infusion that potassium levels began to decrease.
0
The mechanisms by which bicarbonate reduces potassium are debatable. In patients with anuric renal failure the primary mechanism is likely shifting potassium out of the extracellular fluid, as well as some dilutional effect. In patients producing urine, bicarbonate may increase renal secretion of potassium.3 Studies have shown that isotonic bicarbonate lowers potassium over a period of hours in patients given slow infusions. It is possible that isotonic bicarbonate could work more rapidly if infused more quickly.
0
Bottom line? Giving isotonic bicarbonate to patients with significant acidosis probably decreases the potassium. This seems to require sufficient bicarbonate to raise the serum bicarbonate significantly (i.e., >5 mM), which will require large volumes of isotonic bicarbonate. For a patient with hyperkalemia and acidosis who needs volume resuscitation, isotonic bicarbonate is a sensible fluid choice which may improve the hyperkalemia.
0
Complete the volume resuscitation with lactated ringers
0
Following the isotonic bicarbonate, the patient’s bicarb has increased to a normal level and his potassium has also improved. However, he’s still volume depleted and not producing much urine. He needs additional crystalloid volume resuscitation. Continuing to use isotonic bicarbonate would cause a metabolic alkalosis.
0
The traditional dogma is that NS (rather than LR) should be used because the LR will cause hyperkalemia. This is a myth, and it’s completely backwards – the truth is that NS should be avoided because NS will increase the potassium!
0
The clearest evidence that NS is unsafe in hyperkalemic renal failure comes from a randomized controlled trials comparing NS to LR in the operating room during renal transplant surgery (O’Malley et al 2005, Modi et al 2012). Much to the authors surprise, patients receiving NS had higherrates of hyperkalemia requiring rescue therapy (insulin, etc.). The explanation for this is probably that NS causes a hyperchloremic acidosis which shifts potassium out of the cells, increasing the serum potassium level.
0
The amount of potassium in LR (4 mEq/L) is low and will not cause the patient to be hyperkalemic. By the law of averages, this potassium concentration will tend to bring the patient’s potassium closer to normal (i.e., if the patient’s potassium is 6.5 mEq/L and you’re infusing a solution with 4 mEq/L, you will decrease the patient’s potassium concentration). More importantly, LR won’t induce an acidosis which shifts potassium out of the cells. The intracellular potassium concentration is ~140mEq/L, so even shifting a tiny fraction of intracellular potassium into the blood can have substantial effects.
0
In addition to avoiding acidosis and potassium shifts, LR is a good choice here given accumulating evidence that balanced crystalloids improve urine output and reduce the risk of renal failure compared to normal saline. I don’t have space here to go into a full debate of NS vs. LR but I’ll simply state my opinion that NS is an acidotic, aphysiologic fluid which should be avoided whenever possible.4 Unless of course you’re intentionally using it to treat a metabolic alkalosis, in which case NS actually makes sense.
0
Case resolution
0
Following two liters of isotonic bicarbonate and four liters of LR, the patient’s volume status, bicarbonate level, and potassium have improved. He hasn’t produced much urine, but his urine output is beginning to pick up and he feels much better. At this point the patient has no indications for dialysis, and you’ve bought his kidneys some time to recover. If the patient had been resuscitated with six liters of NS, his acidosis and hyperkalemia would likely be exacerbated by the NS and he would need emergent dialysis, ICU admission, and probably spend about a week in the hospital. Instead he spends a few days on the medicine ward, his kidneys gradually improve, and he goes home sooner.
0
PART 2: pH-guided resuscitation
0
This case illustrates the general principle of pH-guided resuscitation. Large-volume resuscitation represents an opportunity to treat certain acid-base problems, specifically metabolic alkalosis, non-anion-gap metabolic acidosis, and acidosis from renal failure.5 Certainly not all acid-base disorders can be treated directly with fluid resuscitation. Most anion-gap metabolic acidoses (i.e., diabetic ketoacidosis and lactic acidosis) require specific treatment of the underlying disorder.
0
0
There are generally two phases to a pH-guided fluid resuscitation (see above). First, any acid-base disorder which may be treated with fluids should be treated (e.g. metabolic alkalosis will respond to normal saline, non-anion-gap metabolic acidosis will respond to isotonic bicarbonate). Second, once any treatable processes have been corrected, a balanced crystalloid such as LR should be used to avoid inducing any acid-base imbalance with the fluid resuscitation. During a large-volume resuscitation electrolytes may be followed to ensure that the resuscitation is on target.
0
Resuscitation is a uniqueopportunity to treat acid-base problems, because such treatment generally requires large volumes of fluid (e.g., two liters of isotonic bicarbonate above). Once the patient is completely volume resuscitated, this opportunity is lost. In the above example, if the patient had been resuscitated with six liters of normal saline it would worsen his acidosis (increasing his bicarbonate deficiency to threeliters of isotonic bicarb)… but at that point it would be impossible to treat his acidosis with isotonic bicarbonate without precipitating volume overload.
0
It’s very frustrating (yet very common) to see patients with acute kidney injury resuscitated from septic or hypovolemic shock with large volumes of normal saline. Such patients are often acidotic to begin with, which normal saline worsens by causing a hyperchloremic metabolic acidosis. A large-volume resuscitation with normal saline often aggravates the acidosis to a point where dialysis may be needed to manage the acid-base disorder. Many such patients would gradually resolve their kidney failure and escape dialysis if not resuscitated with normal saline. It’s possible that avoidance of this normal-saline-induced dialysis explains the reduced need for renal replacement therapy observed in the Yunos et al 2012 study when patients were resuscitated with balanced crystalloids.
0
Conclusions
- Isotonic bicarbonate (D5W with 3 amps bicarb per liter) is a good choice for initial resuscitation of a renal failure patient with acidosis and hyperkalemia, as it may improve both.
- Normal saline has been proven to worsen hyperkalemia in renal failure and should be avoided in these patients. Contrary to prevailing mythology, LR is safe in hyperkalemic renal failure.
- The only potential contraindications to LR are elevated intracranial pressure or profound liver failure.4 LR is a reasonable choice of resuscitation fluid for most patients.
- Large-volume crystalloid resuscitation represents a unique opportunity to correct certain acid-base disorders using pH-guided resuscitation.
- Fluids are some of the resuscitationist’s best drugs: correct choice and dosage matter.
Notes:
(1) “Strong ion difference”(SID) is the most precise way to predict what a fluid will do to a patient’s pH. The pH of the fluid outside the body doesn’t mean much because it is based on both the bicarbonate and pCO2 concentrations, whereas once the fluid is infused, the body will quickly adjust the pCO2. SID is basically equal to the amount of bicarbonate or “bicarbonate equivalents” which the fluid will produce when given to the patient. For example, LR has a SID of 28 mM. Although it has no bicarbonate in it, it contains 28 mM of sodium lactate which is promptly converted into sodium bicarbonate by the liver. So giving LR has the same effect on the pH as infusing a solution of water with 28 mM sodium bicarbonate added.
Fluids with SID below 24-28 mM (such as normal saline with a SID of zero), will cause a hyperchloremic metabolic acidosis with a decrease in the bicarbonate level. Fluids with a SID close to 24-28 mM (i.e. LR, with SID 28 mM) will gently pull the patient’s bicarbonate level towards normal. Fluids with SID >> 24-28 mM (i.e., isotonic bicarbonate, with SID 150 mM) will have cause a metabolic alkalosis. Plasmalyte has a SID of 50 mM, so it has a more gentle alkalinizing effect than isotonic bicarbonate.
This is a gross simplification of the concept of SID but its adequate for clinical use. If you’re a physical chemist then it won’t work for you.
(2) Genius General Hospital is an imaginary hospital where I will treat patients on this blog (yeah, this is a takeoff on Janus General Hospital). It is a composite of various hospitals where I’ve worked and trained over the years. Everything that occurs at this hospital is not genius (!), but when mistakes occurs at Genius General Hospital do not assume that I’m referring to an event which occurred at my current institution. I may be referring to events which occurred years ago during my training in various locations, events which occurred at a various referring hospitals, or events which I’ve heard of through other means.
(4) The only potential contraindication to LR that I’m aware of is elevated intracranial pressure (noting that LR is slightly hypotonic). This probably won’t make much difference to the patient unless you’re bolusing enormous amounts of LR, but the neurosurgeon will yell at you. Patients with profound liver failure may accumulate sodium lactate – this will not cause an acidosis but may drive up the lactate level a bit (note that the “lactate” in LR is in the form of sodium lactate, notlactic acid). I’m not talking about patients with abnormal LFTs here, but rather patients who are dying of liver failure.
(5) In practice its rather rare to have an alkalosis that merits treatment. Often metabolic alkalosis is due to chronic compensation in response to respiratory acidosis in the setting of COPD, and these chronic compensatory acidoses should not be treated but rather must be maintained.
(6) I just learned that ampules of bicarbonate may be available in different sizes. The ampules which I am referring to here are 50ml of bicarbonate in a concentration of 8.4% by weight, which is equal to 1 mEq/ml. Adding three of these ampules (150cc of 8.4% bicarbonate) to a liter of D5W creates 1,150 ml of fluid with a concentration of bicarbonate of 130 mEq/L, which is close to isotonic. When ordering a “liter” of this solution, the patient actually receives 1,150 ml of fluid containing 150 mEq of sodium bicarbonate.
(6) I just learned that ampules of bicarbonate may be available in different sizes. The ampules which I am referring to here are 50ml of bicarbonate in a concentration of 8.4% by weight, which is equal to 1 mEq/ml. Adding three of these ampules (150cc of 8.4% bicarbonate) to a liter of D5W creates 1,150 ml of fluid with a concentration of bicarbonate of 130 mEq/L, which is close to isotonic. When ordering a “liter” of this solution, the patient actually receives 1,150 ml of fluid containing 150 mEq of sodium bicarbonate.
Latest posts by Josh Farkas (see all)
- PulmCrit Blogitorial – Use of ECGs for management of (sub)massive PE - March 24, 2024
- PulmCrit Wee: Propofol induced eyelid opening apraxia – the struggle is real - March 20, 2024
- PulmCrit wee: Why I like central lines for GI bleed resuscitation - March 13, 2024
Why not use sterile water instead of D5W which give you a isotonic (300 mosm/l) instead of hypertonic (552 mosm/l)?
Good questions. #1 – Genius General Hospital does not have liter bags of IV water, given that this is hypotonic and could not be infused alone (i.e., would rarely be used). If sterile water packaged for IV infusion was available, I completely agree that sterile water plus bicarbonate would be superior to using D5W with bicarbonate. #2 – D5W with 150 mM of sodium bicarbonate has an osmolality of >500 mOsm in the bag. However, once infused into the patient the dextrose is metabolized and enters cells, and is not an “effective” osmole in the body. For example, D5W has… Read more »
Hello Dr. Farkas, I once heard an opinion that to increase absorption of water from the GI tract it is better to give some sugar in the same time, so when one molecule of glucose get absorbed it takes two molecules of water with it. If that holds true, is it possible that the same happening at the level of cellular membranes inside other organs? If that hold true, could it possibly make more sense to give D5W to treat mild dehydration and restore intracellular and extravascular compartments, when intravascular compartment is not comprised ((nl BP (of course when PO… Read more »
Hey Josh,
Phenomenal post, as usual!
Just a quick question for you- when giving the initial 1-2 L of Isotonic Bicarb when resuscitating the volume-depleted patient with non-anion gap metabolic acidosis: at what rate are you infusing this? (Are you just rapid bolusing in the same way you would LR?)
Thanks so much,
Sam
Thanks, Sam. My practice is to infuse isotonic bicarb at a rate up to 1000 ml/hour. Theoretically the isotonic bicarbonate includes dissolved carbon dioxide gas, so if you infused it extremely rapidly you could render the patient more acidotic. In reality, the patient will breathe off the carbon dioxide pretty rapidly, so I don’t think this is a real problem (even if they can’t increase their respiratory rate, the increased tension of carbon dioxide in their blood will increase the gradient driving CO2 excretion, causing them to increase the CO2 clearance from their lungs).
Hi Josh. Thanks for you blog, I’m just discovering your posts one by one and its forcing me to reconsider physiology from the start. Again. I have a question derived from your footnote on SID. I agree that the SID of LR is 28 (131+5+2+1-111-29 = 0 outside the body but when the lactate is converted to bicarbonate by the liver you loose the 29, thus 28 in vivo). But I thought that this was because 28mM of Na is unmatched by a strong anion after infusion. My confusion is that giving pure water with 28mM sodium bicarbonate added is… Read more »
Hi Josh,
Also just a quick question for you regarding the use of LR in patients with profound liver failure:
Although sodium lactate shouldn’t cause lactic acidosis, would the inability of the liver to convert the sodium lactate to sodium bicarbonate not render the SID ineffective, thus still contributing to acidosis in these patients?
Thanks!
Sam
Yes, I believe so. Fortunately, there are very few patients with livers this sick (my guess is that this phenomenon is mostly encountered during liver transplantation in the anhepatic phase).
Quick question about isotonic bicarbonate – how do you make it dextrose 5% and what is 3 ampules of bicarbonate (from Scotland, this maybe a stupid but just wanted to get some clarification… thanks
Sorry found it where you explained.
Hi Josh,
Thank you for all your work here, you are helping people around the world (I’m from Chile).
I want to ask you for more information about the effect of LR in the intracranial pressure. Could you recommend me some reading?.
Kind regards
Is the ringer lactate a good option for fluid resuscitation in a child with mitochondrial metabolic disease with raised liver enzymes 560 and raised lactates of 23
Excellent exploration and dissection of this important topic and the caveats to be considered.
Very informative