Search our site 
Advanced Search
Home | Exam dates | Contact us | About us | Testimonials |

You are in Home >> Intensive Care >> Metabolic >> Metabolic - articles and literature

Acid base: Stewart hypothesis and hyperchloraemic acidosis

Created: 23/5/2007
Updated: 26/7/2021

The Stewart hypothesis of acid-base balance

Dr Pete Watkinson DICM FRCA MRCP
Advanced Trainee Intensive Care Medicine

Focus on definitions of acid and base

In order to understand the differences between the classic approach ("Siggaard-Anderson") to acid-base analysis and the Stewart approach, it’s helpful to define a few points from which to work.

The first important point is how to define an acid? This is key, because from the definition comes the term "acidotic". Most people would think of a patient as being acidotic when the pH is below 7.35. As pH is purely a function of the theoretical hydrogen ion concentration of a solution (-log10[H+]), an acid is being defined as a substance that can donate a proton. This simple definition is in line with modern theory.

There are, however, various other ways of defining an acid. An acid can also be defined as any substance that produces an increased concentration of hydrogen ions when dissolved in water (the ‘Arrhenius definition’). The Stewart approach approximates this definition. Unsurprisingly, the attachment of different meanings to the same terminology has led to considerable confusion throughout the history of acid base physiology.


The second important point to note is that both the Siggard-Anderson and Stewart approaches to acid base analysis are mathematical models. Both approaches do not explain the mechanism by which a person has become biochemically acidotic/alkalotic, nor do they claim to do this, but both models merely attempt to illustrate the area the acid-base disturbance lies.

Focus on the Siggaard-Anderson approach


In the Siggaard-Anderson model, an acid-base disturbance is looked at using a combination of 3 factors:

  • The Henderson-Hasselbach equation
  • The base excess
  • The anion gap.

The Henderson-Hasselbach equation comes from the dissociation equation for carbonic acid (see Equation box), and its use is based on the premise that in normal plasma, bicarbonate is the most important buffer.

From this equation, the only two factors affecting pH are:

  • bicarbonate concentration
  • pCO2
The Henderson-Hasselbach equation provides an approximate relationship between the respiratory variable (pCO2), the metabolic variable [HCO3-], and the resultant pH. A flaw with this approach is that other important buffers exist and play an important role in acid base physiology (e.g. haemoglobin and albumin). HCO3- and pCO2 are not therefore independent. In the simplified dissociation equation below, a rise in pCO2 will cause the dissociation to shift to the right as a result of the law of mass action:

CO2 + H2O ↔ H+ + HCO3-

Protons will be buffered by haemoglobin and albumin, and the bicarbonate levels will rise. So a rise in pCO2 has resulted in a rise in [HCO3-]. The rise in [HCO3-] could easily be mistaken as a metabolic alkalosis, when in fact the true cause was a respiratory acidosis. The base excess concept was evolved to address this problem. It is a method of measuring the metabolic component. The base excess concept works by resetting the sample to a normal pCO2 (5.33kPa) by equilibration, and then titrating it to pH 7.4 using molar acid (now calculated from normograms). The number of mmoll-1 required equals the base excess, and is therefore a measure of how acidotic or alkalotic the sample is without any contribution of carbon dioxide. Finally, calculation of the anion gap (see Equation box) allows classification of a metabolic acidosis into those with a normal or increased anion gap. The anion gap is a measure of the concentration of unmeasured anions (e.g. plasma proteins) and is based on the theory of electrical neutrality (the sum of the positive ions must equal the sum of the negative ions). An increased anion gap suggests the presence of unmeasured organic acid, whereas a normal anion gap implies that the decrease in bicarbonate has been counteracted by an increase in chloride concentration (Table 1).

Table 1: Causes of metabolic acidosis - Siggard-Anderson approach

Increased anion gap

(usually decreased [Cl-]

Normal anion gap

(usually increased [Cl-])





Hyperosmolar nonketotic coma

Lactic acidosis

Uraemic acidosis


Ethylene glycol




Parenteral nutrition

Carbonic anhydrase inhibitors

Dilutional acidosis

Ingestion of HCl or other acid

Renal tubular acidosis



Pancreatic fistula
















Equation Box:


Focus on the Stewart approach


Stewart took the same system, but looked at from a slightly different angle. He concluded that one might model acid-base disturbances, based on the three conceptual contributors described in the Equation box above:

  1. The Strong Ion difference (SID)
  2. Weak Acids in Plasma (A total)
  3. PCO2


The law of electrical neutrality means that:


[Na+] + [K+] + [H+] = [Cl-] + [lactate-] + [HCO3-] + [A-] + [CO32-]


 Ignoring the minimal contribution of [H+], [HCO3-] and [CO32-], and substituting the strong ion difference shows:


[SID] = [HCO3-] + [A-]


Stewart puts the three variables together in the Stewart Equation described in the equation box. It is interesting to note that if you ignore the contribution of albumin in this equation, it simplifies to the Henderson-Hasselbalch equation. Thus, albumin is the major variable that Stewart has added in, left out by Siggaard-Andersson for reasons of simplicity.


From this comes a practical utilisation of Stewart theory in that it can be used to define the Strong Ion Gap (see equation box).

Which allows metabolic acidosis to be classified (Table 2).

Table 2: Causes of metabolic acidosis – Stewart approach


Low SID and high SIG

Low SID and low/normal SIG





Lactic acidosis


Ethylene glycol



Parenteral nutrition

Carbonic anhydrase inhibitors


Renal tubular acidosis



Pancreatic fistula



Thus, comparing Tables 1 and 2 illustrate that the two major classifications are broadly similar.

Focus on a real life example of a metabolic acidosis

The Sigaard Anderson and Stewart models of acid base disturbance can be illustrated by the following case, commonly encountered on the ICU. Table 3 shows the blood gas analysis of an elderly patient admitted to the ICU from theatre after a prolonged laparotomy for which he was given 6 litres of saline over the case:



 Table 3: Commonly encountered post-operative ABG



Admission to ICU

Discharge from ICU













Base excess



Sodium mmoll-1



Potassium mmoll-1



HCO3- mmoll-1



Chloride mmoll-1



Lactate mmoll-1



Glucose mmoll-1















In Siggaard-Anderson’s model, the patient has a metabolic acidosis with a normal anion gap. Given the history, a diagnosis of "dilutional", due to saline is the likely explanation.


In the Stewart model, calculation of the SID shows it to be low (normal ~40mmoll-1), and the SIG (normally ~0) is approximately normal, (assuming an albumin of 45gl-1 and a phosphate of 1 mmoll-1):

[A-] = 0.28x(albumin) + 2.14(phosphate) = 12.6 + 2.14 = 14.74


[SID] = [HCO3-] + [A-] = 19 + 14.74 = 33.74


Anion gap = [Na+] + [K+] – [Cl-] – [HCO3-] = 145 + 4.3 – 115 – 19 = 15.3


SIG = AG – [A-] = 15.3 – 14.74


Again, given the history, we draw the same conclusion. Neither of these models provide the mechanism by which saline has caused a significant acidosis. So in this case, the patient is in the group with either a decreased strong ion difference, or normal anion gap, depending on whose model you are using (tables 1 and 2). In either case the patient is hyperchloraemic, as the serum chloride falls outside the normal range (115 mmoll-1). Neither model explains why or even if hyperchloridaemia is causatory in acidosis.



Focus on the cause of hyperchloraemic acidosis


The cause of the hyperchloraemic acidosis is thought to be due to a combination of factors. Firstly, normal saline (pH 5-6) possesses little buffering capacity and is being used to replace blood (pH of 7.4) characterised by extensive buffering capacity. The acidic pH of normal saline is due to a combination of dissolved CO2 and an effect known as the ‘Grotthus mechanism’ whereby dissolving strongly dissociating ions in water causes disruption of water’s ionic bonding, leading to greater dissociation and generation of [H+]. Secondly, volume expansion causes plasma bicarbonate dilution, and renal bicarbonate wasting.


Key Learning Points


  • The two main approaches to acid base analysis are the Siggard-Anderson approach (the ‘classic’) and the Stewart hypothesis



  • The 3 main determinants of acid base status according to Stewart hypothesis are the Strong Ion Difference (SID), bicarbonate (HCO3-)and albumin (A-).



  • Albumin is the major additional variable of acid base analysis that Stewart has considered



  • Stewart’s consideration of plasma protein levels acid base analysis is important for ICU patients as albumin levels are often disturbed



  • Avoidance of such large quantities of saline-containing fluids may help prevent the incidence of hyperchloraemic acidosis on the ICU



Key References



Siggaard-Andersen O, Fogh-Andersen N.
Base excess or buffer base (strong ion difference) as a measure of a non-respiratory acid-base disturbance.
Acta Anaesthesiol Scand 1995; 39 S106:123-128


Corey HE.
Stewart and beyond: New models of acid-base balance.
Kidney International 2003; 64: 777-87


Kitching AJ, Edge CJ.
Acid-base balance: a review of normal physiology.
BJA (CEPD reviews) 2002; 2: 3-6


Waters JH, Miller LR et al.
Cause of metabolic acidosis in prolonged surgery.
Crit Care Med 1999; 27: 2142-2146


Waters JH, Gottleib A et al.
Normal saline versus Lactated Ringer’s solution for intraoperative fluid management in patients undergoing abdominal aortic aneurysm repair: An outcome study.
Anesth Analg 2001; 93: 817-22


SiteSection: Article
  Posting rules

     To view or add comments you must be a registered user and login  

Login Status  

You are not currently logged in.
UK/Ireland Registration
Overseas Registration

  Forgot your password?

All rights reserved © 2022. Designed by AnaesthesiaUK.

{Site map} {Site disclaimer} {Privacy Policy} {Terms and conditions}

 Like us on Facebook