Introduction
You can solve triple acid-base problems with basic arithmetic and straightforward logic. As an example, let’s work through the following rather classic triple acid-base problem:
An afebrile, atraumatic 26-year-old male with no past medical history is brought to the emergency department because of a three hour history of altered mental status, vomiting and vertigo. The physical examination is positive only for somnolence and tachypnea.
Relevant serum chemistries are:
- Na+: 143 mEq/L
- Cl–: 100 mEq/L
- Serum bicarbonate [HCO3–]: 16 mEq/L (normal = 24 mEq/L)
The arterial blood gas (ABG) reveals:
- pH: 7.50 (normal = 7.40)
- pCO2: 20 mm Hg (normal = 40 mm Hg)
- HCO3–: 15 mEq/L (normal = 24 mEq/L)
What are this patient’s acid-base disorders?
Step 1: calculate the anion gap
When solving acid-base problems, the first step is always to calculate the anion gap:
Anion gap = [Na+] – [Cl–] – [HCO3–]
Note that you don’t need an ABG to calculate the anion gap. In fact, an ABG can’t tell you whether there is a high anion gap metabolic acidosis. Only the serum electrolytes can tell you that.
Before proceeding to the next step, however, please pay attention to how high above normal this patient’s anion gap actually is because you will need this number for the steps that follow. This patient’s anion gap is too high by 16 points. It’s supposed to be 11, but it is 27, so it is too high by 16 points (27-11 = 16).
Remember the number 16 because you will need it for the next two steps.
Step 2: see if the serum bicarbonate (HCO3–) is dropped in proportion to the degree to which the anion gap rose
Recall from the previous step that, in this case, the anion gap is too high by 16 points (27-11 = 16). Therefore, you should expect the serum bicarbonate to be similarly 16 points below normal.
A normal serum bicarbonate is 24 mEq/L. (That’s just something you absolutely need to have committed to memory!). In any event, since the patient’s anion gap was too high by 16 points, the patient’s bicarbonate should be similarly low by 16 points. So you’d expect the serum bicarbonate to be 8 mEq/L (24-16 = 8).
Here, however, the bicarbonate is 16 mEq/L, which is 8 points higher than the expected 8. A higher than expected bicarbonate in a patient with no past medical history suggests a metabolic alkalosis. So this patient also has a metabolic alkalosis on top of his high anion gap metabolic acidosis.
Congratulations! So far, you’ve solved two thirds of this patient’s triple acid-base disorder. You know, again without even looking at the ABG (!), that this patient has:
(a) a high anion gap metabolic acidosis (from Step 1); and,
(b) a metabolic alkalosis (from Step 2); and,
(c) another yet-to-be-identified acid-base disturbance (because I told you it was going to be a triple acid-base disorder!).
Let’s go on to Step 3.
Step 3: look at the arterial blood gas and see if the partial pressure of carbon dioxide (pCO2) is dropping in proportion to the drop in the bicarbonate level
Look back at the ABG above and focus on the pCO2. Notice that the pCO2 is too low by 20 points. (A normal pCO2 is 40 mm Hg, and this patient’s pCO2 is 20 mm Hg. 40-20 = 20.)
Had the respiratory compensation been appropriate, the pCO2 should have dropped by the same number of points as the bicarbonate. Why? That’s just how it is. (If you’re not convinced, just plug the numbers into Winter’s formula and you’ll get exactly the same result.) In any event, recall that this patient’s serum bicarbonate dropped by only 8 points (it is 16, with normal being 24), so his pCO2 should have dropped by only 8 points as well. Therefore, his pCO2 should have been 32 mm Hg (40 – 8 = 32). However, this patient’s pCO2 is lower than the expected 32. It is 20. A too-low pCO2 means that there is a respiratory alkalosis going on here as well.
We therefore now know that this patient’s triple acid-base disorder is:
(a) high anion gap metabolic acidosis (from Step 1),
(b) metabolic alkalosis (from Step 2); and,
(c) respiratory alkalosis (Step 3).
Summary
Now that the (rather straightforward!) arithmetic is done, it’s time to look back at the patient history and physical examination to see if your laboratory-based acid-base diagnosis fits the clinical scenario. Recall that this was an afebrile, atraumatic 26-year-old male with no past medical history who presented with altered mental status. This history alone suggests a likely toxic ingestion and therefore the possibility of a high anion gap metabolic acidosis. He had been vomiting. That fits with metabolic alkalosis. He was tachypneic on examination. That fits with respiratory alkalosis.
And finally, he was altered and vertiginous. That fits with a final clinical diagnosis of salicylate poisoning with a characteristic triple acid-base disturbance of (1) high anion gap metabolic acidosis, (2) metabolic alkalosis, and (3) respiratory alkalosis.
Reference
- Kurtz, Ira, MD, Acid-Base Case Studies (2004, page 96, Case No. 7)

Excellent Mark!!
Now how to solve ‘Triple Acid-Base Problem’ when the respiratory component is acidosis.
Here is an example:
pH 7. 21 (7.35-7.45)
Base Excess -11.3 (<+3- -3)
pCO2 54.2 mmHg (35-45)
pO2 74 mmHg (80-100 mmHg)
Bicarbonate 33.1 mmol/L (22-28)
Na 146 mmol/L (138-145 mmo/L)
K 5.4 mmol/L (3.5-5 mmo/L)
Chloride 96 mmol/l (95-105 mmol/l)
Urea 36.2 mmol/l (2.5-7.7 mmol/l)
Creatinine 788 μmol/l (50-120 μmol/l)
In this patient HCO3 is higher than the expected compensatory rise due to Chronic Resp Acidosis. So 'Respiratory Acidosis is accompanied by Metabolic Alkalosis. High anion gap (146-96+33 = 17) indicates a co-existing metabolic acidosis.
Very nice, Aamir! Thank you very much for your contribution!
Thank you for the algorithm Dr. Yoffe.
I am a bit puzzled with the first step. The pH at presentation is in alkalotic range (pH: 7.50), but you conclude it is metabolic acidosis. Is just the fact of anion gap brings to the conclusion?
Thanks
Hi Alac,
Strictly speaking, acidosis or alkalosis are processes, while acidemia and alkalemia are states. In other words, the biochemical processes of acidosis and alkalosis drive the equilibrium toward a state academia or alkalemia.
I am just trying to understand if the fact that there is anion gap is sufficient to conclude that it is metabolic acidosis (as algorithm concludes)?
What if there is no anion gap and ph is 7.5 what will be the conclusion for the first step: acidosis/alcalosis?
Yes, the presence of a high anion gap is sufficient proof that there is a metabolic acidosis.
To Alec’s comment above, I looked at this problem this way:
1st look at pH to determine acidosis or alkalosis. pH is high, therefore this is Alkalosis.
2nd look at PaCO2. When PaCO2 and pH “move” in the same direction, the primary process is metabolic. When they “move” in opposite directions the primary process is respiratory. Therefore I would conclude the primary disorder here is a Respiratory Alkalosis.
3. Determine appropriate compensation. Since it is a respiratory Alkalosis, determine if HCO3 is appropriately changing. For an acute Respiratory Alkalosis we would expect HCO3 to change by 2 mEq for each change of 10 in PCO2. So, the HCO3 here should be 20. But it is 16. There is a concomitant metabolic acidosis (which happens to be an AGMA). It is also easy to calculate AG here and determine there is an AGMA.
4. Determine the delta/delta: The Change in AG is 27-12 = 15, and the change in HCO3 is 24-16 = 8. The Delta/delta is 15/8 >2. A Delta/Delta greater than 2 means the anion gap is changing to a greater degree than the Bicarb is changing, indicating an AGMA + Metabolic Alkalosis.
So, we have a triple acid-base disorder. Primary is Respiratory Alkalosis, with AGMA and a concomitant metabolic alkalosis
But, what if the pH is say, 7.39 (I know that they say that compensatory mechanisms never bring back pH to NORMAL but I have come across triple acid base balance disorders with pH in the range of 7.35-7.45. How do we determine the primary disorder in that case?
Respected anonymous, let us calculate the percentage variance in PCO2 and HCO3-. Greater variance hints at the associated primary disorder. Further, normal pH is towards higher side or lower side of the range of normaly, will again hint at the primary culprit. – SodaAd Adsoda, MD