Normal saline causes acute renal failure? PulmCCM Central
Nov 292012
Does Normal Saline Cause Acute Renal Failure?

To internal medicine-trained physicians in the U.S., normal saline solution seems as harmless and healthy as mother's milk. Intensivists trained in anesthesia or surgery might more often mention normal saline's hypertonicity compared to blood, and its propensity to cause hyperchloremia, compared to lactated Ringer's or similar solutions. But who cares, really? Chloride is that number you can usually ignore in the chemistry panel -- almost always clinically irrelevant. Right?

Evidence suggests that in fact, excess chloride may be a little nastier than previously suspected. Normal saline infusion may worsen kidney perfusion, compared to PlasmaLyte, a crystalloid concoction that approximates human plasma in pH, osmolality, buffering capacity, etc. Normal saline also causes hyperchloremic metabolic acidosis, although any ill effects of that aren't clear.

So Nor’azim Mohd Yunos, Rinaldo Bellomo, Michael Bailey et al set out to see what would happen to rates of dialysis use at their Melbourne hospital if they replaced normal saline and all other high-chloride solutions with low-chloride ones.

See also: Resuscitation Fluids PulmCCM Review

What They Did

In this prospective, observational trial, the investigators first recorded lab values and other data on 760 patients admitted to a single ICU during a 6 month lead-in period in 2008, when normal saline flowed freely in usual critical care practice patterns.

They then abruptly cut off doctors' supply of normal saline and other high-chloride solutions. Physicians were forced to use Hartmann's solution (similar to lactated Ringer's), Plasma-Lyte, or low-chloride 20% albumin, unless they obtained attending specialist approval for normal saline. Patient data was again collected for 6 months.

What They Found
Patients received less chloride: an average of ~300 mmol less each, on average, in the intervention period (694-->496 mmol/patient). More interestingly, avoidance of normal saline and high-chloride solutions seemed to protect their kidneys:
  • Average serum creatinine rose by 0.25 per patient in the control period -- but only by 0.17 in the intervention period (p=.03).
  • The incidence of acute kidney injury (acute renal failure) was 14% in the normal saline period,  vs 8.4% with low-chloride solutions (p <.001).
  • 10% of patients needed renal replacement therapy during the normal saline epoch, but only 6% after normal saline was restricted (p = .005).
  • Patients receiving normal saline / high-chloride solutions had double the odds of RIFLE-defined acute kidney injury requiring dialysis, after adjustment for covariates (p = .004).

In-hospital mortality, length of stay in the hospital or ICU, and need for dialysis after hospital discharge were similar before and after normal saline restriction.

What It Means

Here's what all this stuff has in it:

Normal Saline vs Lactated Ringer's vs Plasma-Lyte

Normal Saline (0.9% NaCl) Lactated Ringer's Plasma-Lyte
Sodium (Na) 154 mmol/L 130 mmol/L 140 mmol/L
Chloride (Cl) 154 mmol/L 109 mmol/L 98 mmol/L
Potassium (K) none 4 mmol/L 5 mmol/L
Calcium (Ca) none 1.5 mmol/L none
Magnesium (Mg) none none 3 mmol/L
Lactate none 28 meq (28 mmol/L) none
Acetate none none 27 mmol/L
Gluconate none none 23 mmol/L
Tonicity Hypertonic (308 mOsm/L) Hypotonic (276 mOsm/L) Isotonic (294 mOsm/L)
Cost $2 / liter $4 / liter $12 / liter

Several intriguing small experiments in dogs and people suggest that chloride causes decreased kidney perfusion and may interfere with hemostasis. This interesting trial brings us a step closer to discovering any ill effects of normal saline and high chloride solutions. Administrators will cringe at the price difference between normal saline vs. lactated Ringer's and Plasma-Lyte, but this same group previously reported that overall fluid costs went down after restricting saline (most physicians switched to the Hartmann's solution). However, in that study -- but not here -- they also noted more episodes of alkalemia and slightly higher lactate levels in the no-normal-saline group.

Now, to better answer the question and eliminate the potential contributing secular, Hawthorne or "bundle" effects due to unblinding and the multiple simultaneous interventions in the present study, how about a randomized trial? Paging ANZICS and the Canadians!

Nor’azim Mohd Yunos, Rinaldo Bellomo, Michael Bailey, et al. Association Between a Chloride-Liberal vs Chloride-Restrictive Intravenous Fluid Administration Strategy and Kidney Injury in Critically Ill Adults. JAMA 2012; 308(15): 1566-1572.

Resuscitation Fluids in the ICU: PulmCCM Review

Liked this post? Get a weekly email update, and explore our library of clinical guidelines, practice updatesreview articles. and board review questions.

PulmCCM is an independent publication not affiliated with or endorsed by any other organization, society or journal referenced on the website. (Terms of Use)

Authors: contribute your work in a guest post.

  15 Responses to “Normal saline: toxic to kidneys? Chloride solutions may cause renal failure”

  1. Just a note that normal saline is barely hypertonic as compared to blood as you said in the direct paragraph. I don’t think that the osmolality of normal saline is in anyway problematic.

  2. I don’t think it was the hypertonicity that they were casting blame on but rather the excess chloride in NaCl 0.9%.

  3. If large amount of 0.9% NaCl is transfused,then only chances of normal anion-gap acidosis arise.

  4. How can those costs be correct? I can find a site that will sell me a bag of 1 L NS for $2.43.

    • Hi Carol. The costs in the table are the retail prices from Baxter’s online catalog. As you know, hospitals negotiate much lower prices than these, but probably not as low as $2.43. That was probably sterile irrigation saline (packaged differently from infusion saline) or saline for veterinarians. Thanks for writing.

  5. What about all the other electrolytes that are different? Surely potassium might make the difference. RCT definitely needed.

  6. The control period wasn’t conducted in secret, as you suggest. This was a prospective, open-label study.

  7. $80/bag is too expensive, you may be quoting the per carton price. Anyway, Plasmalyte is, as identified, about 2.5x more expensive than NS.

  8. Wait! Since when is NS considered Hypertonic? Please educate me on this one.

  9. Always agree, I in our institute switched over to half normal saline and use RL and have definitely found a decrease in renal failure. Chloride does something that we don’t know

  10. means balanced crystalloids are to used? like plasmalyte and stereofundin( b braun)

  11. […] Enough with the “Normal” Saline! So its been about a year since a JAMA article ( finally showed that the downside of 0.9% saline isn’t just theoretical, but has some associated clinical morbidity (bad for the kidneys!).  Sadly enough, it still seems to be the routine fluid used for boluses. Whether the ER, hospitalist or intensivist, residents, students…it seems people are reluctant to let go. Today, rounding in the ICU, I was changing an order for a bolus from another doc from NS to RL, and a nurse asked me why.  I gave her a capsule summary and she was in disbelief.  ”Come on Phil, they wouldn’t call it normal saline if it wasn’t!” I’m an internist by training, so naturally I grew up using NS, since that’s what all the attendings and residents used around me.  Ringer‘s was the stuff the surgeons used, so well, I guess it had to be wrong…no? So forward to 2001 and John Kellum‘s lecture on acid-base I’ve previously mentioned, and my exploring Stewart’s Physicochemical Approach, and wait, I look at the back of a bag of NS, and find out, much to my dismay, that the stuff I’ve been using like holy water has a pH of 5.6.  And who have I been giving liters and liters of this stuff to?  Yup, mostly patients with acidosis. Hmmm. Interesting. So although I don’t necessarily advocate correcting metabolic acidosis for the sake of doing so (see my previous post on bicarb), I’m not a proponent of worsening acidosis either, even if by another mechanism. I think there are a number of factors that have resulted in this situation.  For starters, there is the issue of false advertising – the “normal saline” monicker has been influencing subliminal thought for decades (think Malcolm Gladwell thin-slicing), making physicians feel they are giving and inherently “good” substance.  Then there’s the whole tribalism thing with the surgeons vs non-surgeons making all the non-surgeons polarize away from RL (not that RL is perfect, just a bit better, and certainly closer to “normal”). Finally, there’s this sad, sad factor that makes people, even (or maybe even more) smart people reluctant to accept that they have been doing something wrong (or, for those who are offended right now, not ideal) for a long time (I sure was) and prefer to fight it and rationalize it for a few more years until, eventually, the evidence becomes overwhelming or the changing of the guard has fully taken place. I think what we should be hanging on to is not a drug or a fluid but rather what we learned in the first couple of years of med school: physiology.  Now mind you, at that point we (or most of us) didn’t have a clue how to use it for anything more that answering multiple choice questions, but at some point, we have to go back to it and realize that is what we should be basing our assessment of our therapeutic acts and decisions. So…if I have a situation where I am low on chloride, I might want to use NS. But otherwise, let try to give something whose composition is a bit closer to our own than NS is.  So, for my students and residents, don’t let me see you prescribing boluses of NS.  If you really, really need to, wait until your next rotation please. thanks! Philippe ps for a great review of the original aritcle, please see Matt’s on PulmCCM at :; […]

  12. we have learnt about hyperchloremic acidosi is caused by chloride shift. Is it not prudent to use the so called normal saline excessively in shock I do agree with the author in this context