Quantitative Approach To Acid Base
When I was initially taught acid base interpretation in paramedic school, we would start by drawing a tic-tac-toe grid.
As we read off the values of the ABG, an X would be made in the corresponding box. In this graph above, we see a primary metabolic acidosis with respiratory compensation. After you did this a few times in class, you could easily do it in your head during simulation. I actually still use this method when a nurse or physician is reading off the ABG values to me prior to a flight(minus the tic-tac-toe box). The net result however typically ends with you identifying that they are in a metabolic acidosis, their bicarb levels are below norm, and there is an inherit desire to replenish with exogenous bicarb. I mean wouldn't it only make sense?
I want to do my best to explain a technique called the "Stewart Approach" when interpreting acid base disturbances. I believe this to be a game changer when it comes to guiding fluid therapy and knowing when to pull the trigger on administering sodium bicarbonate.
The Down & Dirty Concept
Your body is always trying to maintain electrochemical neutrality between anions and cations. If you were to have an imbalance of one or the other you would probably get shocked every time you touched something :0. So in order to maintain this balance, your body has the ability to generate hydrogen (cation) and bicarbonate (anion). These perfectly equalize each other as seen below on the Gamble-Gram.
Obviously you can tell that sodium and chloride make up the majority of the strong ions. We call them strong ions because they will disassociate in water. If you drop table salt into a cup of water it will disassociate into sodium and chloride. Whats interesting though is that you always have to introduce a cation with an anion into a liquid. Otherwise you would literally get an exothermic reaction (that means explosion... I looked it up). That is why you always see sodium-chloride or sodium-bicarb, etc.
If you subtract the chloride level from the sodium you get something called the strong ion difference or SID. This SID will be made up of the things you see above (albumin,bicarb, and lactate). A normal SID (based on a sodium of 140 and a chloride of 102) is 38. If the SID is lower than 38, you have a metabolic acidosis. if your SID is higher than 38, you have a metabolic alkalosis.
So reset your brain, and recognize that its normal to wonder why this makes sense..
Now let's say we give someone a liter of 0.9% saline. There is 154 meq of sodium and 154 meq of chloride. This concentration of chloride is obviously higher than our endogenous chloride. This will cause our serum chloride to elevate (shrinking the SID). As mentioned above, as we exogenously increase the anions, the body will need to release cations to maintain neutrality. This is done by releasing hydrogen ions. Because pH is a measurement of hydrogen ions, this will decrease pH. So its not actually the chloride making us acidic, it's the bodies response to it.
On the other hand, if someone gets super sick and has explosive diarrhea (1 year and I finally get to use that word in a blog.. yes!) They can lose bicarb at rapid rates. The body can not replenish it as fast as they are losing it. This will cause the bicarb box to shrink and the chloride level will elevate to maintain neutrality. Make sense? cool!
If your lactate levels increase, your body will need to balance the additional anions with hydrogen ions to maintain a positive/negative match. This will cause your bicarbonate level to shrink to maintain neutrality. Your bicarb level is not low because it is depleted, it is reduced because the lactate is elevated. For some reason it reminds me of the Monroe-Kelly theory of ICP. The equilibrium of each anion adjusts to maintain neutralIty. It's worth noting that in the absence of renal failure, patients should produce endless amounts of bicarbonate. The correct way to reverse this is to reverse the cause of the elevated lactate. Not to tunnel vision on correcting the bicarb.
If the albumin (weak acid) elevates... blah blah blah it will need to be balanced by hydrogen ions to maintain neutrality. This will be its own separate blog soon.
The majority of carbon dioxide is transported as bicarbonate. An ABG will utilize a calculated bicarb based on the PaC02 and pH. The BMP will utilize a directly measured bicarb and will typically label it under C02. The more c02 you retain, the more bicarb you need for transport. This is why patients with COPD have higher than normal bicarb levels. This is also why you shouldn't try to normalize PaC02 levels in a patient with COPD. They will reach a normal PaC02, realize they don't need the bicarb, piss it out, and then have a hard time weaning off the vent.. rant over. Regardless, elevated bicarb levels due to elevated c02 will increase hydrogen concentration to maintain neutrality.
The tic-tac-toc method above also does not account for something I find extremely helpful... the base excess.
The base excess or BE is usually at the end of your ABG/VBG. It is calculated by pretending your PaC02 levels are normal, and looking to see how much base or acid would be needed to get back to a normal pH. If your BE was -10, you would know that you need 10 of base (loss of bicarb), or you have an additional 10 of acid. To get back to a physiological pH. If your BE was 5. You would know you need 5 of acid to return to physiological pH. A normal BE is +/- 2 based on a pH of 7.40. Obviously there is a parameter of 7.35 to 7.45, so the normal BE does not HAVE to be subtracted from your BE for this to work.
Now let's try this out in a case!
A patient presents with a lower GI bleed and diarrhea x 1 day. Initially presents with diminishing LOC and hypotension. She is administered a total of 4 liters of 0.9% saline in 2 hours. Now she appears tired but will look at you and answer questions appropriately.
The nurse hands you a bag of 0.9% saline medicated with 100 meq of sodium bicarb. She said the referring physician would like this hung on transport. You ask for a copy of the most recent labs and get the following:
Glucose 49 mg/dl
Your brain recognizes the acidosis, immediately calculates a SID (Sodium - Chloride) of 34, and notes that there is a component of hyperchloremic acidosis.Your thought is that there were +4 liters of 0.9% in a 60 kg elderly woman, and there is likely an iatrogenic hyperchloremic acidosis component.
However, there is elevated anion gap, which obviously the lactate is doing most the heavy lifting on. To break this down into a quantitative approach you take the base deficit of 12.9 and correlate 9.5 of that to lactate. That leaves you with 3.4 of acid left. A normal SID of 38 subtract the current SID of 34 leaves you with 4. You now have 0.6 remaining and +2 to -2 is considered a normal variable. This means you have accounted for the metabolic component of the acidosis.
So now back to the question of “should we hang this liter bag of 0.9% saline with 100 meq of bicarb in it?”
0.9% Saline Concentration
Lets start with recognizing what the sodium concentration of a 0.9% liter bag is. If you read the front label you will see there is 900mg of sodium in every 100 ml’s. This is why we say 0.9%. If you were to read a bag of 3% saline it would say 3 g of sodium for every 100 ml’s.
8.4% Sodium Bicarbonate Concentration
One amp of Sodium Bicarb contains 4.2 grams of sodium in 50ml’s. This breaks down to 84mg per ml. If you were to look at how much this is per 100ml’s it would be the equivalent to 4.2 grams per 100ml’s or 4.2% sodium. However it is in 50 ml's so it doubles the concentration.
What about when we mix them?
Imagine now you have a liter bag of saline that for every 100 ml’s has 900 mg of saline.
You now add 8.4 grams of sodium bicarbonate (two amps). 8400 grams divided by 1000ml’s gives you 8.4 mg per ml. This is an additional 840 mg per 100 ml’s.
Now lets figure out the concentration..
900mg + 840mg = 1.7 grams (1.7% sodium) per 100 ml’s
So will this combination of 0.9% saline and two amps of bicarb help the patients acid base disturbance?
900 mg per 100ml’s of chloride gives us the following:
154 Meq of Sodium
154 Meq of Chloride
This leaves us with a SID of zero which is obviously less than our bodies normal SID of 38. This means if we were to keep infusing 0.9% saline into a patient they would become profoundly acidotic.
But they added two amps of sodium bicarbonate.
The only thing we change in the above bicarb/0.9% combo is the concentration of sodium.
154 Meq of Chloride
254 Meq of Sodium
Each amp of bicarb will read 1 meq per ml. This added to our liter of saline should be an additional 100 meq.
This leaves us with a SID of 100. Technically this increase in the strong ion difference should make the patient alkalotic.
What happens when this combo hits the blood?
The SID of a fluid needs to be measured against the patients current SID. Our patient above has a SID of 34. As we increase the serum cation concentration with sodium we will need an anion to match and maintain neutrality. This will not be able to be completely matched by chloride, and the body will need to release bicarbonate (anion) to maintain neutrality. The release of bicarbonate will begin to alkalinize the serum. Because this concentration is hypertonic, it will draw fluid out from the cells and can increase serum potassium levels. This is known as a solvent drag and must be considered.
So does this patient need bicarb?
You probably have been wondering why we haven’t addressed the sky high C02 yet. The base excess equation pretends the c02 level is normal, and thus removes the respiratory component. However, we know that the three independent variables of acid base are SID, C02, and Albumin. I believe this patient needs to be placed on non invasive positive pressure to assist in correcting a respiratory acidosis. The patient had several periods of hypotension reported by the staff contributed to hypovolemia/GI bleed. I believe the elevated lactate could very easily be from the respiratory acidosis and hypotension. So my conclusion is that this patient probably does not need bicarbonate. They need respiratory assistance to blow off c02, follow up periods of hypotension to be treated with blood transfusion, and GI consult.
The only thing that would lead me towards giving bicarb is the elevated creat. No reported hx of renal failure, and this could be attributed to the period of hypotension.. but with the data from the Bicar-ICU trial I was tempted to but 50meq of bicarb in a 50 bag of D5w and run it in over the flight.
It is worth mentioning that the conversion of bicarbonate to c02 and water within the pulmonary microcirculation will only take place if the concentration of oxygen is higher than c02 within the alveolus region. In a patient with elevated carbon dioxide levels, bicarbonate will be converted back to hydrogen and water, thus causing an intracellular acidosis. This is explained very well in a recent guest blog by Jay Nance called "The Capped Volcano."
What if they needed bicarb?
If I decided the patient did need sodium bicarbonate, I would probably put 50 meq of sodium bicarbonate into a 50ml bag of D5W as recommended in Josh Farkas's recent pulmcrit post. This would give me 4.2 % (4.2 grams in 100ml's). Given the usual large excess of fluids our patients have received prior to our arrival, I believe this to probably be the most beneficial technique. If volume replenishment was needed, I would put three amps of sodium bicarb in a liter bag of D5w. This would allow me to avoid any excess chloride, and elevate the SID.
FOAMed Peer Review
Dr. Jason Bowman
I agree with the lower dose of bicarb if needed, but literally nobody in EM outside of people you can name know about the BICAR ICU data. So you’d be so far ahead of the curve. People would assume you were wrong and probably file a complaint on you. Your calculations all look correct, looks like a mixed acidosis. The respiratory component is the easiest to fix. The biggest difference in treatment I would make is if this patient truly has a GI bleed, CO2 in the 90’s and pH of <7 I would skip NIPPV and would’ve electively intubated immediately. Correct the acidosis, protect the airway and now the patient is ready for GI to scope emergently. Some of the confusing things to me though is that her Hgb is fine. This makes me think sepsis. Obviously acute bleed wouldn’t drop hgb that quick but GI bleeds are rarely acute and if they’ve given 4L of NS it should be lower. Sepsis gets confused for a GI bleed a lot, so it may be something like that. Also, I would give some dextrose, her sugar is too low, plus it should help the acidosis to stop ketosis. I feel like those labs aren’t convincing for GI bleed and the patient is obviously sick, so I’d be going back to square one on her.
Based on my quick reading of this I think your right on.
I think the lactate is only accounting for 7 of the BE however. (-+/2). Probably stress induced. Does it clear with treatment?
Your SID is correct. (34)
Anion gap of 14 doesn’t really do anything for me. If uncorrected that is slightly high. The K+ is low based on that pH of 6.92. A little concerning. But if you correct the pH with CO2, it corrects to 7.32 using a perfect (40) for PaCO2. I’d treat that.
If using a corrected anion gap it still doesn’t show much. Maybe 18-20. There’s no CO2 listed from chem 7 panel. So guessing.
I fully agree with no NS or bicarbonate. I’d fix respiratory first and reevaluate q 1 -2 hr
If lower GI bleed. The H&H is looking great. But that leads me to think that’s an older lab and secondary to hemo-concentration. I would also be curious if blood products had been given.
Dr. Josh Farkas
1) NS with amps of bicarb creates a hypertonic solution, this should basically never be given. If you want to give hypertonic bicarb, just use amps of hypertonic bicarb. The NS/bicarb solution is weird as the NS component will worsen the hyperchloremic acidosis so this will work against you.
2) agree pH derangement mostly respiratory, bicarb unlikely to change this much
3) agree giving 1-2 amps of bicarb would be reasonable (largely to offset the hyperchloremic acidosis due to NS), but this is unlikely to make a huge change in pH