Should Normal saline be the norm?

Here is my observation of clinical practice in most places I have worked in Australia.

The doctors use normal (0.9%) saline as a default fluid for most situations.  It is recommended by all the resuscitation algorithms as the first line and whenever an intern charts some fluid – you can bet that 90% of the time it will be saline, or saline with K added if they are worried about a low K.

When I went to anaesthetic school, I changed my practice and started giving everyone Hartmann’s (compound Sodium lactate / Ringer’s).  Why?  Because that is what all my bosses did – they seemed like a smart bunch, and it was the path of least resistance – plus my charting CSL made me look like I was in the exclusive club of gas-docs who “know” a little more than the ward folks!  Here is the ugly truth – as a rule doctors have a relatively poor grasp on the effects IV fluids and their prescriptions have on patients physiology.

Over the last few years the use of various fluids in various situations has been bandied around the literature and in the online medical social media.  So here is where I am as of today with my thinking around IV fluids for electrolytes & acid-base control and hydration.  I have evolved a few “rules” in my own mind which I now produce for criticism and improvement.  Before we begin – here is a nifty Fluid make-up chart from Scott Weingart @ Emcrit for your reference. Open it in a new tab to make the rest of this post easier to understand.

Here we go, he evidence is sketchy and the rules are largely my opinion:

Saline is isotonic, but relatively hyperchloremic – which means it is on balance an acidic fluid in simple terms. It contains no other electrolytes or buffer. As a strategy for ongoing fluid replacement it is essentially diluting a lot of good stuff and offering no replacement. If you run a sick patient on saline alone – especially if their kidneys are struggling – they will become acidotic.

The indications for saline over CSL etc as 1st line are uncommon – here is the list a recent “twitter storm” produced:

1. Hyperkalemia (though isotonic bicarb is probably better)
2. Brain injury (CSL is hypotonic, isolyte/plasmalyte are ok)
3. Rehydration for vomiting – would have to be severe though
4. Fluids in a alkalemic  eg. pyloric stenosis, some toxic ingestions, maybe upper GI fistula?

CSL is slightly hypotonic, has a bit of each of the main electrolytes. It contains lactate – which is a good thing.

Why lactate, why not bicarb?  We cannot store bicarb in a plastic bag – it would just diffuse into the environment as CO2. So lactate is stable, and it gets converted to bicarb by the liver – so as long as your patient has a functioning liver you are essentially giving bicarb in a bag at physiological concentrations.

We give IV fluids all the time, but I don’t think many of us stop and think about it as a ‘prescription’. We never give it the same cognitive effort that we might give say antibiotic choice, or anti-hypertensives. We tailor our intervention to suit the patient. I think most of us use a default position and only change when it is really unavoidable – eg. severe metabolic derangement, drug incomaptibility…

So here is my rule – look at the patient, the acid-base status, the electrolytes and the plan.  Then consider which fluid is going to cause the least problems, or maybe provide a benefit.

The myth of Paediatric fluid therapy I was brought up in was that kids needed hypotonic fluids (eg. 0.45% saline +dextrose). I believe the idea was that their kidneys would not cope with high solute load? Not sure, it is just a myth after all!  There is a summary table of commonly used Paeds fluids from RCH here.

Now there are a few recent RCTs which show that the use of hypotonic fluids as maintenece runs a real risk of causing hyponatremia.  Check out this RCT from Pediatrics, Nov 2011.  CSL was also safer than hypotonics in this Aussie study.  After all Hartman was a Paediatrician! Cliff Reid @Resus.Me has a nice summary here.

In trauma or bleeding patients we are drilled to give saline by all the mainstream courses. The evidence for this is not good. There is a wealth of data coing out of the wars of recent years showing crystalloids are doing bad things in trauma.

My practice as of now – I use crystalloid if I really need to give volume (ie. MAP < 65) in the early stages of a resus in the absence of blood.  Some people would argue we should use hypertonic saline here – but this is not an option yet in most small hospitals.

My cut off is at the 2 liter mark.  If we are hanging a 3rd litre of crystalloid – then we need to make sure the blood is on its way up the hall!  And remember, just because it is a trauma – you do not have to squeeze in fluids at high rates if the numbers are OK.  I think this is a reflex we all learned in our ABC courses.  It is likely doing some harm if you do it indiscriminately.  For more on this see my post on Massive transfusion.

I was trained to use normal saline for the severe metabolic acidosis of DKA. As a template for acidosis resus, it is one we all see and adhere to the dogma. But why?

Saline is making metabolic acidosis worse right?  They need volume, but would it not be better to give a balanced fluid?

Here is an excerpt from the UK DKA guidelines 2010. “In theory replacement with glucose and compound sodium lactate (Hartmann’s solution) with potassium, would prevent hyperchloraemic metabolic acidosis, as well as allow appropriate potassium replacement. However, at present this is not readily available as a licensed infusion fluid.“

OK, in my place CSL is easily available. So I use it for DKA / metabolic acidosis. I monitor the electric lights closely. It works, so my plan is to continue until I see evidence to the contrary. No saline in DKA for me.

Now I am sure that this will upset a few readers – there are so many scenarios we could cover  – but the ones above are the main areas where I see saline used and I wonder if it is not a mistake.

Love to hear your comments.  Casey


Add a Comment

Your email address will not be published. Required fields are marked *