Acute Care

Metabolic Acidosis

You are the medical FY2 doctor on call and a GP referral has just arrived in the Medical Assessment Unit (MAU). You are told the patient has been vomiting for 2 days with associated polyuria and polydipsia, they are tachycardic but have a stable blood pressure. The nurse is worried and the patient looks sick. The nurse inserts a cannula and takes blood samples. Just before she finishes you ask for a venous blood gas (VBG) because you want some answers immediately about how unwell the patient is.

The result you get is:

  • pH 7.10
  • HCO3 6 mmol/L
  • BE -14 mmol/L
  • K 7 mmol/L
  • Na 140 mmol/L
  • Cl 105 mmol/L
  • Lactate 6 mmol/L

What is causing this? How do we work it out? What do we do?

Dr Simon Ridler is a Consultant Anesthetist and Intensivist at Chester who has a keen interest in the problem solving of metabolic acidosis. In the episode, we discuss an alternative approach to metabolic acidosis to help guide therapy.

Find the podcast below:



Podbean (non-iPhone)

We use a case of DKA to help discussion. If you wish a recap on DKA and it’s management, check out our previous post on the subject here.



Metabolic Acidosis

Metabolic acidosis is the herald of something bad happening inside a patient’s body (1). The causes of this acidosis are multiple and the interpretation of the acidosis can be complex. Acidosis is worrying to us because it provides numerical representation of an underlying bad diagnosis for the patient. I remember medical school teaching me basic blood gas analysis and the anion gap in a way that never leant itself to use at the bedside.

Throughout the podcast, Simon references two papers (2, 3) on the subject which can be found by following links at the end of this discussion. They are worth reading in order to gain a deeper understanding of the underlying science behind our proposed interpretation method.

Metabolic acidosis can be divided into three severity categories (4):

  • Mild: pH 7.30 – 7.35
  • Moderate: pH 7.20 – 7.29
  • Severe: pH <7.20

Though these categories may not have clinical significance using such terminology can help orientate others when discussing cases with them.


Effects of Acidosis

A metabolic acidosis has a profound impact upon a patient and is associated with an increased mortality (4 – 7). The physiological effects are (4, 8):


  • Decreased cardiac output
  • Cardiovascular instability and predisposition to arhythmias
  • Decreased response to catecholamines
  • Arterial vasodilation and decreased systemic vascular resistance


  • Hyperventilation (Kussmaul breathing)
  • Right shift of oxygen-haemoglobin dissociation curve (decreased affinity of Hb for Oxygen)


  • Hyperkalaemia
  • Increased sympathetic activity but decreased responsiveness to it
  • Impaired ATP production
  • Decreased immune response and leukocyte function
  • Increased cellular apoptosis

The important factor when faced with a severe acidosis is finding the underlying cause and treating that definitively (1, 9).



Remember our blood gas results:

  • pH 7.10
  • HCO3 6 mmol/L
  • BE -14 mmol/L
  • K 7 mmol/L
  • Na 140 mmol/L
  • Cl 105 mmol/L
  • Lactate 6 mmol/L

This is a medical emergency and needs treating immediately. The patient is profoundly acidotic and hyperkalaemic.

Why are they acidotic?

With it being a VBG, we have no pCO2 to look for any respiratory component to the acidosis but it would likely be low to compensate for the metabolic acidosis. It may be venous as opposed to arterial, but especially when looking at metabolic components, there is good correlation between the two and an arterial sample may not be necessary (11).

The Base Excess is very negative, representing a metabolic acidosis.

The BE represents the amount of HCO3 that would need to be added to the sample in a standardised Hb to restore a normal pH (3). As will become relevant later, the units are fortunately mmol/L. BE is heavily influenced by haemoglobin and so the standard BE is used which is independent of pCO2.


Why are they acidotic?

We need to work out why the patient is acidotic in order to be able to treat them appropriately and our therapy will be different depending on the cause. When faced with acutely unwell patients, we want to be making quick decisions with what information we can get as quickly as possible. A blood gas is often quickly available and can yield a large amount of information when analysed.



Stewart’s Theory and the Quantitative Approach

In our case, we have an acutely unwell patient with a history of polyuria and polydipsia preceded by 2 days of vomiting. This suggests possible DKA as the underlying cause. Is it the only explanation? DKA is a good example because during their on-going treatment they may have more than one cause of acidosis and the management is very different for each.

Simon sums up Stewart’s Theory very well in that the [H+] is the end result of a series of equilibriums rather than the driving force for the entire process. According to Stewart, an acid increases [H+] relative to [OH-] and a base decreases [H+] relative to [OH-] (2).

  • pH 7.10
  • HCO3 6 mmol/L
  • BE -14 mmol/L
  • K 7 mmol/L
  • Na 140 mmol/L
  • Cl 105 mmol/L
  • Lactate 6 mmol/L
  • Ketones 5.4 mmol/L

All of these results were obtained by point of care testing and we can start working out the reasons why the patient is acidotic by simple maths.

The BE = -14 mmol/L. We have 14 mmol/L to explain and gauge the magnitude of contributory components. There are three steps (3):

  1. Eliminate the “saline effect”
  2. Correct for albumin effect
  3. Calculate “unmeasured” anion effect (the effect from other anions, most commonly this would be ketones and lactate (12, 13) which can be measured at the bedside)

For use at the bedside, some anions will be measured such as lactate and ketones. Lactate and ketones are expressed as negatives in the sums because they reduce the base excess (ie produce acidosis).

To follow a more useable structure try this method:

  1. Eliminate the “saline effect”
    • (Na-Cl) – 38 = (140-105) – 38 = 35 – 38 = -3
    • -14 – (-3) = -11 mmol/L
  2. Subtract lactate -11 – (- 6) = 5 mmol/L
  3. Subtract ketones -5 – (- 5.4) = 0.4  mmol/L
  4. Is the BE completely explained? If not, correct for albumin:
    • (42 – serum alb) x 0.25
  5. Is it explained now? If not, consider other unmeasured anions produced by renal and liver failure or ischaemia. Keep an open mind that this may be a case of exogenous acid poisoning. Measure osmolality and compare expected versus measured via the osmolal gap if toxic alcohols or poisoning are suspected (4).

In our example case it is explained (and these are genuine numbers!). The total effect from the ketones, lactate and “saline effect” renders the left over BE within the normal limits of +2 to -2 mmol/L.


Endogenous vs Exogenous Acid

Endogenous acids, broadly speaking, are acids produced by the body when it is unwell. This category includes lactate, ketones, renal failure and unmeasured anions found in conditions such as in acute liver failure and ischaemia. For the inpatient population, one small study found that the risk factors for developing a metabolic acidosis were hepatic, renal and cardiac dysfunction and all cases of acidosis were related to an underlying illness (13).

Exogenous acids would be anything ingested that produces an acidosis. Examples include salicylates, glycols, methanol etc. Indeed, only a minority of cases will be due to exogenous acids (4).



There are different methods for interpreting a metabolic acidosis, Stewart’s Theory and the Quantitative Approach is just one. Does it matter what you use? Probably not (10, 14). What the quantitative approach does is inform what the contributing factors to the acidosis are and help us target therapy at the underlying cause (10). However, what we really want is a method that can be used at the bedside that focusses on the underlying pathophysiological cause and helps us target treatment accordingly.


Traditional teaching has focussed on the Anion Gap and differentiating the cause from this. The traditional mneumonic taught is MUDPILES:

  • Methanol
  • Uraemia
  • Diabetic ketoacidosis
  • Paracetamol/propylene glycol
  • Iron/infection/isoniazid
  • Lactate
  • Ethylene glycol
  • Salicylates


A more modern alternative taught is GOLDMARK:

  • Glycols (ethylene and propylene)
  • Oxoproline
  • L-lactate
  • Dlactate
  • Methanol
  • Aspirin
  • Renal failure (uraemia)
  • Ketoacidosis


In practice when you are faced with an acutely unwell patient, the history will often point you down the right road. You may not have a history ingestion or poisoning but the situation may be suggestive. In truth, especially when treating inpatients, the most likely causes are much narrower and can be thought of as something LURKing inside the body.

We propose a simplified mnemonic for the endogenous causes of a metabolic acidosis that we are most likely to be faced with:

  • Lactate
  • Unmeasured anions (this includes liver failure)
  • Renal failure
  • Ketoacidosis

Why use this? It may not be evidence based, but it is simple and includes what would be the more commonly occurring and immediately treatable causes for metabolic acidosis presenting to and in hospital. If these causes were in fact all normal then you would need to consider the exogenous causes mentioned previously but they can be ruled out easily. Alongside this, especially when reassessing response to IV fluids, consider the saline effect and hyperchloraemia.


Acidosis Treatment

In our initial example, we have explained the acidosis by adjusting the Base Excess. Our actions need to target the underlying problems.

First, we need to switch off the ketogenesis by providing insulin. Most hospitals will have a standardised DKA pathway to follow. Second, the body is under stress and producing lactate. In the case described, we likely have volume loss through vomiting and an osmotic diuresis driving this lactate and need to provide fluid to correct this. In the setting of DKA, this is often regimented by the pathway. If there is another reason such as sepsis then your fluid therapy would likely be in the form of boluses and reassessment.

However, our fluid therapy itself can skew the interpretation of the blood gas.


Saline Effect?!


The underlying science for this effect centre around the principles of Stewart’s Theory. It is best to read about this for yourself. Simply put, our body desires a state of electrical neutrality where the total cations equal the total anions and there is no overall ionic charge. Cations are tightly regulated whereas anions can have greater fluctuation (10).

anions vs cations

The major cation in serum is Sodium (Na). The major anion is Chloride (Cl). These are known as strong ions, ie they fully dissociate when in solution. HCO3- and the A- component (comprised of mainly albumin) are both weak ions ie they do not fully dissociate when in solution and are in equilibrium with their undissociated states. Because they exist in equilibrium, the balance can shift in either direction in the presence of pH disturbances or electrochemical disturbances.

bicarbonate_equationsImage result for weak acid dissociation equationpH is a logarithmic derivative of [H+]. [H+] is measured in NANOmoles/L and the other substances we are interested in are measured in MILImoles/L. Therefore, only minute alterations in [H+] are needed to affect pH (2). The slight [H+] alterations induced by fluctuations in the difference between Na and Cl could, in theory, cause derangements of pH and lead to acidosis.


Hyperchloraemic Acidosis


Often through resuscitation, we administer large quantities of Sodium and Chloride via “Normal Saline”. One litre contains 154 mmol/L of both Na and Cl and has a pH of 5.3. Remember though that our normal Cl is nearly 40 mmol/L less than our normal Na. Therefore, we add proportionately more Cl to every litre of serum than we do Na. In doing so, we narrow the difference between the two ie a reduced strong ion difference (SID). It is this narrowing that results in hyperchloraemic acidosis by its effect on [H+].

In this situation, you may find that some misdiagnose this as persistent DKA despite resolution of the ketosis and adjust the insulin. You may not be able to change your IV fluid due to the need to administer potassium if treating a DKA. If using Saline for resuscitation then you could switch to a balanced solution such as Hartmann’s or Ringer’s. It is important to recognise this type of acidosis to avoid unnecessary treatments that seek to correct this iatrogenically created state.

For example, 6 hours into our patient’s treatment:

  • pH 7.27
  • BE -15
  • HCO3 14
  • Na 150
  • Cl 122
  • Lactate 0.6
  • Ketones 1.6
  • Albumin 33
  1. Subtract “saline” effect
    • (150 – 122) – 38 = -15 – (-10) = -5
  2. Eliminate known anions (ketones & lactate)
    • -5 + 0.6 + 1.6 = -2.8 mmol/L
  3. Correct for albumin
    • -2.8 – 0.25(42-33) = -5.05 mmol/L
  4. Is the BE corrected?

[CORRECTION] In this example no. If you had only the point of care tests available to you, you would reach a corrected BE of -2.8 mmol/L and know the majority of the acidosis is contributed to by the saline effect rather than other anions. Remember when correcting for albumin, it is a weak acid within the body and often decreases in acute illness and the chronically unwell. A loss of this weak acid masks some of the acidosis in the unwell patient. This should alert you to the possibility of other causes of acidosis such as renal failure, especially in the acutely unwell inpatient population, or an ischaemic event. The point of this exercise is not as a diagnostic modality but to demonstrate the factors contributing to the acidosis to direct treatment against them.

Remember: Acidosis occurs as a result of an underlying pathological process, interpretation helps us differentiate these processes and target therapy accordingly.

There is some evidence that excess chloride administration can have negative implications for the patient (16, 17). However, from the perspective of a junior doctor, the underlying cause of an acidosis needs treating and will often involve IV fluids. Use what is immediately available (as long as it is not Glucose) for resuscitation and adjust later based on results when available.


Renal Failure

Unfortunately, we cannot use a quantitative approach to demonstrate the contribution of renal failure to the acidosis. However, your history and examination may well allude to its presence.

If you find a patient who is acutely unwell who has risk factors for an Acute Kidney Injury (eg older age, sepsis, dehydration, medication) and despite simple calculations there remains a acid component unexplained then you should consider renal failure.

Renal failure prevents the kidneys from generating HCO3- to act as a buffer for the acidosis and decreases the excretion of organic acid anions (4).

Whatever the underlying pathology in this situation, your first treatment is likely going to be fluid resuscitation and/or discussion with critical care/nephrology if indicated. But you can identify these patients sooner by use of a blood gas and detect complications such as hyperkalaemia.


Sodium Bicarbonate

It’s a common question that juniors doctors do ask: when do you give sodium bicarb? There isn’t a definite answer to this.

The literature has not found good evidence of benefit from bicarb therapy in situation where it is the underlying condition that has led to the acidosis. These situations include sepsis, hyperlactataemia and DKA (1, 9).

Biochemically, sodium bicarb works by providing chloride-free sodium, thus raising the strong ion difference and alleviating the saline effect. The bicarb administered enters equilibrium and is rapidly removed as CO2 via the lungs (9). It may well correct the acidosis in some situations, but the improvements in numbers has not correlated with improved outcomes.

As we state in the podcast, we treat patients, not numbers so please don’t become distracted by therapies that target numbers rather than the underlying pathology.


Practical Application

In practice, I use this quantitative method frequently, accepting it does not always give me the exact answers. But the educated guess you can then make at a likely underlying diagnosis allows prompter and more targeted therapy. In does take practice though and some understanding of the basic principles. I have not gone into depth with the basic science here because others have done it for me far better than I could (2, 3).

The best way to become proficient at analysing blood gases is just to try and use the methods presented each time you look at one, even if you already know the reason why they are acidotic.

A blood gas, whether arterial or venous can give you a good amount of information about a patient and allow more immediate treatment based on the results. Use them liberally and become comfortable acting on the results you get when dealing with unwell patients.

Once you take one, make sure you see the results. Where possible take or ask someone to take the sample to a machine for a printout result rather than sending to a pathology lab. If you find a result that gives you concern, escalate it to your senior.


Please read the underlying literature to gain a better understanding of the topic.

Speak soon



An excellent resource on this topic is the EMcrit podcast. Scott Weingart has a series looking at particular aspects. It’s what piqued my interest in the first place.

GET IN TOUCH! Either through the comments section or email if there are any comments or corrections or if you want to be involved with future topics.

RATE US on iTunes to help others find the resource more easily!

SHARE US with others you think would benefit or even your training leads to help us distribute the resource further!



  1. Adeva-Andany MM et al. Sodium bicarbonate therapy in patients with metabolic acidosis. Scientific World Journal. 2014:627-673

  2. Chawla & Drummond. Water, Strong Ions and Weak Ions. Contin Educ Anaesth Crit Care Pain (2008) 8 (3): 108-112.
  3. Badr & Nightingale. An alternative approach to acid–base abnormalities in critically ill patients. Contin Educ Anaesth Crit Care Pain (2007) 7 (4): 107-111.
  4. Kraut JA, Madias NE. Treatment of acute metabolic acidosis: a pathophysiologic approach. Nat Rev Nephrol. 2012 Oct;8(10):589-601
  5. Kraut JA, Madias NE. Metabolic acidosis: pathophysiology, diagnosis and management. Nat Rev Nephrol. 2010 May;6(5):274-85

  6. Lee SW et al. Lactic acidosis not hyperlactatemia as a predictor of in hospital mortality in septic emergency patients. Emerg Med J. 2008 Oct;25(10):659-65

  7. Gunnerson KJ et al. Lactate versus non-lactate metabolic acidosis: a retrospective outcome evaluation of critically ill patients. Crit Care. 2006 Feb;10(1):R22.
  8. Libório AB. AKI complications in critically ill patients: association with mortality rates and RRT. Clin J Am Soc Nephrol. 2015 Jan 7;10(1):21-8

  9. Kimmoun A. Hemodynamic consequences of severe lactic acidosis in shock states: from bench to bedside. Crit Care. 2015 Apr 9;19:175

  10. Velissaris D et al. The Use of Sodium Bicarbonate in the Treatment of Acidosis in Sepsis: A Literature Update on a Long Term Debate. Crit Care Res Pract. 2015;2015:605830. Epub 2015 Jul 30.
  11. Krutz et al. Acid-base analysis: a critique of the Stewart an bicarbonate-centred approaches. Am J Physiol Renal Physiol (2008); 294(5): F1009-F1031
  12. Patricia A. Diagnostic Importance of an Increased Serum Anion Gap. N Engl J Med 1980; 303:854-858
  13. Scott KA et al. Acidosis in the hospital setting: is metformin a common precipitant? Intern Med J. 2010 May;40(5):342-6

  14. Kurtz I et al. Acid-base analysis: a critique of the Stewart and bicarbonate-centered approaches. Am J Physiol Renal Physiol. 2008 May;294(5):F1009-31. Epub 2008 Jan 9.

  15. Dubin et al. Comparison of three different methods of evaluation of metabolic acid-base disorders. Crit Care Med. 2007 May;35(5):1264-70.

  16. Neyra JA et al. Association of Hyperchloremia With Hospital Mortality in Critically Ill Septic Patients. Crit Care Med. 2015 Sep;43(9):1938-44
  17. Association between intravenous chloride load during resuscitation and in-hospital mortality among patients with SIRS. Intensive Care Med. 2014 Dec;40(12):1897-905. Epub 2014 Oct 8.
  18. Brill et al. Base deficit does not predict mortality when secondary to hyperchloremic acidosis. Shock. 2002 Jun;17(6):459-62.

One reply on “Metabolic Acidosis”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.