Thursday, October 18, 2012

Is normal saline bad for the kidneys?



Answer: Yes.

This week, the two heavyweight medical journals, JAMA and the New England Journal of Medicine, featured papers describing the effect of certain intravenous fluids on the incidence of renal failure in critically ill ICU patients.

The JAMA paper compared normal saline (relative to human plasma, a high chloride-containing solution) administration to more physiologic, low chloride-containing IV fluids such as Hartmann’s solution (very similar to Ringer’s lactate) or Plasma-Lyte 148. It showed that using the low chloride intravenous infusions led to a statistically significant decrease in the incidence of acute kidney injury and the need for renal replacement therapy.

The NEJM paper compared the use of intravenous hydroxyethyl starch (HES) fluid resuscitation to normal saline and found that patients given HES had significantly more acute kidney injury and needed more renal replacement therapy. Bear in mind that HES is actually a solution of 6% HES in normal saline.

Neither study found a significant difference in mortality rates related to the various solutions used.

Both studies were performed in Australia during different time periods. The JAMA paper was based on research from a single hospital in Melbourne in 2008-2009 and was a before-and-after trial while the NEJM study was multi-institutional, randomized and prospective and took place from December 2009 to January 2012.

Is normal saline bad for the kidneys? Yes. If you compare high chloride normal saline to lower chloride solutions, normal saline causes more renal dysfunction and need for renal replacement therapy. Normal saline vs. HES really compared normal saline alone to 6% starch in normal saline, and showed that the starch is probably the factor causing renal injury.

So what is a clinician to do? Normal saline is not really “normal.” Solutions containing amounts of chloride closer to that of human plasma are the correct ones to use. As we surgeons have maintained all along, Ringer’s lactate should be the resuscitation fluid of choice in the U.S.

See the table below for the amounts of sodium, chloride and buffer in standard IV solutions.



17 comments:

Joel Topf said...

When lactate was used as the alkali in dialysate solutions it caused hypotension. Now we use acetate or bicarb. So plasma-lyte maybe a better choice.

Skeptical Scalpel said...

Thanks for commenting. Patients on dialysis represent a very special situation. Their kidneys have already failed. Dialysis can correct nearly all chemical abnormalities.

I'm glad you didn't point out that Ringer's lactate has 4 Meq of K per liter. Many people think you can't give it to oliguric patients because of that. They fail to understand that the average patient has a total body K of > 3000 Meq and 4 Meq is not going to harm anyone.

Matthew Weidman said...

LR and Hartmann's are low-sodium compared to plasma. Could that lead to hyponatremia in a significant fluid resuscitation effort? I know in studies comparing maintenance fluids, several studies have shown more hyponatremia with half-normal vs NS. LR and Hartmann's have significantly more sodium than half-NS, but if you're doing fluid resuscitation for sepsis and give 60 mL/kg or more ... ?

Skeptical Scalpel said...

Matthew, that's a very good question. Hyponatremia has not been a significant problem with LR. Most patients receive several liters of fluid during resuscitation but it would take quite a lot to dilute the intravascular volume since the difference in sodium is only 10 Meq/L. And since the kidneys are more likely to work properly with LR, they will sort it out.

Plasma-Lyte 148 costs about 3X as much as LR, ~$12 vs $4 respectively. That doesn't sound like much until you realize that 200 million liters of normal saline are used yearly in the US.

Rachel J. Kwon said...

Thanks for this review, especially your comment in the second-to-last paragraph about normal saline not being normal. I'm always surprised by my colleagues in medicine who order NS for EVERYBODY.

One of my SICU attendings taught me to treat IV fluids as medications -- both require consideration with dosing and both can have serious adverse effects. He would actually stop rounds to have us calculate the mEq of chloride in LR versus NS to show us that resuscitating patients with NS puts them into hyperchloremic acidosis. That was an exercise that really stuck with me and makes me think twice when ordering fluids, which are too often seen as completely benign, or all the same.

At my institution we use Hextend (essentially, HES in LR) so I'm curious if there are any similar large-scale trials looking at that...

Skeptical Scalpel said...

Rachel, I appreciate your comments. I agree about NS and hyperchloremic acidosis.

Somewhere deep in the DNA of all internists, is an aversion to Ringer's lactate. I think they may have been frightened by lactate when they were children.

I have never used Hextend and am not aware of any good studies about it and renal failure. Since it is 6% hetastarch in RL, I would think it's similar 6% hetastarch in NS. That is assuming the starch is the culprit. A discredited researcher named Joachim Boldt wrote about it but he's had about 100 papers retracted. I don't think he would be a good source.

Anonymous said...

Part of the aversion to LR is the mistaken belief that the Sodium Lactate contributes to Lactic Acidosis, as an ACLS instructor once told a class.

Skeptical Scalpel said...

I agree. That is a common and persistent misunderstanding.

Anonymous said...

Probably the right result. Now we just need an RCT to prove it....

Skeptical Scalpel said...

The New England Journal study was a randomized trial. It might be difficult to do an RCT comparing NS to RL. Who would fund it?

Anonymous said...

Yes, the NEJM trial was randomised but wasn't a chloride trial, which was what I meant we needed. I think such a trial is possible and the will from the ICM, anaesthetic & surgical community is there as recognition emerges that we really don't have a sound evidence base for very much of our fluid therapy at all. There have been 3 very good fluid trials (SAFE, CHEST, 6S) and they have proven these pragmatic trials can be done. I think the chloride trial will come....

Anonymous said...

Just in case any of the readers missed it, this is a fascinating study on the subject:

Ann Surg. 2012 Jul;256(1):18-24

Skeptical Scalpel said...

Anon, I misunderstood your comment calling for an RCT. Thanks for mentioning that study in Annals of Surgery. It was a good one.

Anonymous said...

We use Ringers acetate as standard fluid in the ICU and have been doing that for many years. Hyponatremia and hyperkalemia is not a problem. Hyperchloremic acidosis is a real problem with saline. Hard outcome data is however not available. Lactate is a problem since it may screw up your diagnostics in high volume resuscitation. (hard to evaluate lactate clearance when you are infusing lactate at the same time). HES is available in both saline and balanced solutions, but is intrinsically evil when used in severe sepsis.

Skeptical Scalpel said...

Good comments. Thanks. I agree with you. I think HES is on its way out.

Anonymous said...

SS: I saw your tweet today. I have been a general surgeon for about a decade; was at Mattox's trauma meeting in Vegas several years back, and he was blasting LR. (He was against the D-lactate isomer as being unphysiologic, I think.) He was favoring 1/2 NS or 1/4 NS as a maintenance fluid. He actually got me largely to stop using LR; now you are telling me that Mattox was wrong. I am now in a rural practice and find it much harder to follow the literature the way I used to.

Skeptical Scalpel said...

Right or wrong, Ken is always certain of his position.

Note that the papers were about critically ill patients receiving large volume resuscitation. I don't think it matters what you use for maintenance IV fluid in most patients.

Post a Comment