This 74 year old gentleman attended the ED after phoning a friend because he ‘though he was having a stroke in both hands’. Paramedics had to gain entry to the house, which was in a state of disrepair, cold, and unclean. The patient was found on the floor, surrounded by vomit. I tend to do a VBG in situations like this because I get an acid/base status and other useful information back faster than formal bloods.
His observations were essentially normal, apart from his 3 lead which was a veritable soup of short lived atrial arrhythmias, and PVCs. He was also a bit cold 34 degrees C. What is your interpretation of this gas?
So starting from the top the patient is Alkalotic, with an elevated CO2. This means they have to have a metabolic alkalosis with respiratory compensation. Lets looks more closely at the metabolic component, the BE is 22, which means we have ‘22’ more bases than normal, we can also see that his bicarbonate is 40.8. (thats where they are coming from).
There are clues here. We know that bicarbonate takes time to respond to problems. This man must have a chronic problem causing his bicarbonate to go up. We can infer this is a chronic metabolic alkalosis with a degree of respiratory compensation which is probably new.
Lets examine the AG – the gap is 55.4! Which is the highest gap I have EVER seen. Remember that AG is calculated by adding the Na and K, and taking the chloride from bicarbonate. Where is the source of the gap. It’s predominantly from the Chloride. Look it’s 44! That’s less than HALF what it should be, I suspect that it’s not the only cause of the Gap here, as we’ve got a lactate of 20 , pushing in the other direction and perversely helping to correct the alkalosis.
If you fancy you can calculate his SID which is 96! High SID alkalosis is usually caused by gastric outlet obstruction, vomiting, excessive NG suctioning, diuretic mistakes, primary hypoaldosteronism, or volume depletion.
This man has pyloric stenosis from untreated chronic H pylori, and acute renal failure secondary to volume depletion. I think his gas shows a chronic metabolic alkalosis with respiratory compensation and a hyperlactaemia. I have never seen this pattern in an adult before!
Now we’ve covered the easy bits of acid base, the determinants of pH, and compensation. This week I’m covering all the weird and wonderful calculations you can do to help you potentially whittle down the diagnostic options. You don’t have to calculate SID, SIG, or OG for every gas, but its useful to know how to, so when you need to, you can.
So a metabolic acidosis is caused by a decrease in SID, making the buffer base contract to maintain the law electrochemical neutrality. This is usually because strong anions accumulate. Now we don’t normally measure the strong anions, so we find them by calculating ‘gaps’. Most of these ‘gaps’ were introduced before we started routinely measuring lactates, so I think it’s a point of contention whether lactate still counts as an ‘unmeasured’ anion, but for the purposes of what we’re talking about let’s call it a measured, unmeasured one!
Take your measured anions and subtract them from your measured cations.
Anion Gap = [Na+] + [K+] – [Cl-] – [HCO3–]
If the anion gap is really large (>30) you have a metabolic acidosis. Anion gaps get bigger if you have less HCO3–, Cl–. Now the ‘gap’ itself, is this unmeasured component. The anion gaps is affected by fluctuations in ATOT because A- which makes up ATOT is usually the biggest component of the gap in healthy individuals. This means that if someone’s albumin, and sometimes phosphate are low, the size of their anion gap may be less, or normal, when in fact they’re got a rip roaring metabolic acidosis.
We can correct for albumin;
AGc = AG + 2.5 (normal albumin in g/dl – observed albumin in g/dl)
The normal range for Anion Gap and AGc is between 10 and 18 mEq/L
Causes of a high anion gap metabolic acidosis can be remembered by the mnemonic MUDPILESCAT and causes of a normal anion gap acidosis can be remembered by USED CRAP
This is a quick by the bedside estimate. If you wanted to calculate a proper SID, you’d need a whole host of other measured values, but as you now know was makes up SID you can use this quick and dirty measure.
What you are doing here is taking your strong ion difference, and taking it away from the buffer base, if you do this. The normal SIG should be close to zero. The SIG is NOT the AG, and it is not the same as the Standard Base Excess.
High SIG (> than 2)
Low SIG (less than -<2)
Increase in unmeasured strong anions
Increase in unmeasured weak anions
Polygelinate (Gelspan etc)
Increase in unmeasured strong cations (lithium)Increase in unmeasured weak cations
I think SIG is quite difficult to use, one of its big limitations is quite small errors in the lab will affect the SIG. It’s also not really been studied enough to make it easy to work out routinely by the bedside. Whereas things like the AGc have been around for decades. Now some studies have shown that a high SIG is a very strong predictor of mortality [ROC 0.991 (95% I 0.972-0.998) ] but this was in penetrating trauma patients requiring vascular surgery (quite a narrow population).
[citation – Kaplan, Lewis J., and John A. Kellum. “Initial pH, base deficit, lactate, anion gap, strong ion difference, and strong ion gap predict outcome from major vascular injury*.” Critical care medicine 32.5 (2004): 1120-1124.]
Anion Gaps, SIDs, and SIGs can be useful for estimating charged particle concentrations within plasma. What about uncharged molecules? For this we need a different gap.
The osmolar gap is calculated by taking measured osmolality from the calculated osmolality. It should normally be less than 10. The osmolality is the measured. You can use an osmolar gap calculation to look for molecules dissolved in the plasma that do not have a charge.
Osmole: A unit of osmotic pressure, the amount of a solute that dissolves in solution to form one mole of particles.
Osmolality: number of osmoles dissolved in a kg of solvent [mass] (This is measured)
Osmolarity: number of osmoles dissolved in a litre of solvent [volume] (This is calculated)
Calculated Osmolarity = (2 X Na+) + Glucose + Urea
Osmolar Gap = Osmolality – Osmolarity
If we have a high osmolar gap, we can infer there is an excess of a weird dissolved molecule in the blood. If you work in a department that doesn’t routinely measure ETOH, an elevated osmolar gap metabolic acidosis might be your only blood result that tells you they are sloshed (you may have worked this out in other ways). Some calculators or equations add in alcohol to this equation too.
Causes of high osmolar gap:
Propylene glycol (used to suspend lorazapam and diazepam IV solutions)
Glycine (think TURP)
Three words of caution
There are 2 types of osmometers, one works by using vapour pressure, the other using freezing point depression. Only the freezing point depression method works accurately with volatile chemicals like methanol and ethanol. Your lab could use either. If they use vapour pressure osmometry you might get falsely normal results.
There are many different formulae for the OG, depending on local units of measurement. None of them are wrong, but all of them produce slightly different results.
An obtunded patient, with a high osmolar gap could have ingested more than one thing.
So in conclusion
We can use the Anion Gap to give us a better idea of the causes of metabolic acidosis. SID and SIG can give us some further clarity in certain situations but there utility is nowhere near as good as the anion gap. The Osmolar Gap gives us an indication of unmeasured uncharged molecules, but doesn’t tell us what they are.
I know last week I said I’d talk about SIDe and SIDa, but I needed to put this groundwork in first.As it came to over a 1000 words, we’ll save the gaps for next week.
The way we are taught to analyse blood gases comes from a miss-match of 2 slightly different schools of thought based on Henderson Hasselbach understanding of blood gas chemistry (Boston and Copenhagen). Though the stewart hypothesis helps us understand whats going on with acid-base its utility by the bedside is limited.
Old School methods, are based on observational data, and often fall apart or get confusing when things get complicated.
Most people’s mental algorithm for blood gas analysis looks like this:
Step 1: Look at the pH is it low or high?
Step 2: Look at the PCO2 is it low or high?
Step 3: Look at a Metabolic component, is it low or high?
So when we all look at a blood gas the first thing all of us do is look at the pH and then the CO2. If the pH is low and the CO2 is high we throw up our hands and go “it’s a respiratory acidosis” and move on. However if the patient has a low pH and a normal CO2 we claim it’s metabolic and move on.
Now if you are a Bostonian you look at the HCO3- and apply a series of rules of thumb to work out what kind of metabolic disturbance, if you are a Copenhaganite you look at the standard Base Excess.
You can do either. I’m not judging you, but I find the Boston approach hard to reconcile as it seems you need to have the answer before you start looking at the blood gas, some people find this approach reasonable as you always have a clinical context to interpret the ABG in. The other issue with this approach is that it uses the HCO3– which we know from last week is a dependent variable.
Boston Rules of Thumb
NB: The Boston rules are AMERICAN, which means they use partial pressures in mmHg To convert to a KPa you need to divide the American value by 7.6 (or ask google to convert).
1 for 10 in Acute Respiratory Acidosis
The HCO3– will increase by 1 mmol/l for every 10mmHg in pCO2 over 40(mmhg)
Expected HCO3– = 24 + [actual pCO2 – 40]/10
Effectively the higher CO2 shifts the equilibrium towards the production of more HCO3–
4 for 10 in Chronic Respiratory Acidosis
The HCO3– will increase by 4 mmol/l for every 10mmHg in pCO2 over 40
Expected HCO3–= 24 + 4 [actual pCO2 – 40]/10
Renal compensation occurs over a few days.
2 for 10 in Acute Respiratory Alkalosis
The HCO3–will decrease by 2mmol/l for every 10mmgHg in pCO2 below 40
Expected HCO3–= 24 -2 (40- Actual pCO2)/10
However you can normally not get a HCO3–less than 18mmol/l because you cannot have negative values of PCO2. So if your number here is less than 18, it suggests a co-existing metabolic acidosis.
The 5 for 10 Rule for a Chronic Respiratory Alkalosis
HCO3– will reduce by 5mmol/l for every 10mmHg decrease in pCO2 below 40mmHg.
Expected HCO3– = 24 -5 [40-Actual pCO2]/10
The limit of compensation is about 12-15 mmol/l
Your answer can be +/-2
The One and Half pluse 8 Rule for a metabolic acidosis
Expected PCO2 = 1.5 x [HCO3-] + 8
The limit of PCO2 is about 10 mmHg Your answer can be +/-2
The point Seven plus Twenty Rule for a metabolic alkalosis
Expected pCO2 = 0.7 x [HCO3–] + 20
Your answer can be +/-5
There are certainly some pretty big limitations to this approach, it tends to fall down in a big heap if you have someone with a combination of acid base disorders, or has 2 of the same type (for example 2 causes of metabolic acidosis). It’s a lot to remember for little benefit over what we were taught at medical school but here are some examples where they might be useful. It also fails to take account of ATOT and SID in explaining acid base disturbance.
You’ve got Diedre, an end stage COPD patient who claims to be more short of breath than normal. She has a PCO2 of 70 and an HCO3– of 36. The expected HCO3 for her (rule 2) is 24 + 12 = 36. We can see that the actual measured value is the same as the expected value so we can be pretty sure there is no evidence of another acid base disorder lying in wait for us.
You’ve got Seth, a 27 year old space cadet who has been found unconscious by police with various white powders about his person. His GCS is E1V2M2 and his BSL is 6.9. His gas shows a pH of 7.2 with pCO2 of 70 and a HCO3– of 14. His expected PCO2 [rule 5] is 1.5 x 14 + 8 = 29mmHg, as his actual CO2 is way higher than his expected we can infer that there is a respiratory component to his acidosis as well, that he’ll probably need intubating.
To be honest I think rule 2 and rule 5 are the most useful, and they are the ones I try to remember.
Copenhagen approach – Base Excess, Standard Base Excess
Base Excess was a term coined in the 1960’s by Siggaard-Anderson. It is 0 when the pH =7.4 pCO2 = 40 mmHg and the temperature is 37°C. If either the pH or the pCO2 are note 7.4 or 40mmHg, BE becomes the amount of Hydrogen required to bring the pH back to 7.4, while maintaining a pCO2 of 40mmHg.
If the BE is +VE it reflects the amount of Hydrogen you need to add to the solution to make it neutral that means the solution is going to be alkali.
If the BE is –VE it reflects the amount of Hydrogen you need to take out of the solution to make it neutral, that means the solution is going to be acidic.
Okay that’s the Base Excess, we tend not to use this because of a couple of things, firstly because it fails to take into account the movement of CO2 due to Gibbs-Donnan forces, and it fails to model accurately enough the effect of ATOT (specifically the effect of Haemoglobin inside red blood cells). Standard base excess is base excess calculated at a Hb concentration of 50g/L.
SBE = 0.93 x ([HCO3–] + 14.84 x (pH – 7.4) – 24.4)
We can use either for day to day analysis, but remember that normal BE becomes less accurate with big swings in pCO2 and Hb.
If the BE is -VE the solution is going to need LESS hydrogen to make it neutral, this means that there is either a primary metabolic acidosis or a compensated respiratory alkalosis (you’ll know immediately depending on if the CO2 is high or low). This method isn’t perfect either, it still doesn’t tell you if the acid base disturbance is primarily metabolic or because of respiratory compensation.
So we can use our rules of thumb, or our BE to split our acid-base disorder into 1 of 4 diagnoses. These are the primary acid base disorders, the ‘osis’ part dictates the direction of acid base disorder, not the actual pH of the solution. So you can, in theory, have a respiratory alkalosis with a metabolic acidaemia.
Are you compensating for something?
Compensation is the response to acidosis. The body has 2 mechanisms for dealing with it depending on if the primary acidosis has a metabolic or respiratory origin.
Metabolic compensation occurs in response to a respiratory acid-base disturbance. The kidneys excrete either more or less Hydrogen, they do this by altering the extracellular SID (by changing the urinary SID, by altering their Chloride excretion). This is very effective and can compensate for swings of pCO2 from 25-80mmHg. This process takes time, up to 5 days, this is why we can have incomplete compensation, or partial compensation for respiratory acid-base problems.
Decrease urine SID, increase extracellular SID
Increase urine SID, decrease extracellular SID
Respiratory compensation is fast, but much less effective than metabolic compensation. A normal pH is not normally achieved. The increase or decrease in pCO2 is driven primarily by breathing faster or slower. This is driven by central and peripheral chemoreceptors. The central ones tend to kick in later, as it takes longer for the acidosis to equilibrate in the CSF.
A normal pH combined with an abnormal pCO2 can either mean that there are 2 opposing acid base disorders, or that there is a compensated respiratory acidosis. A normal PCO2 combined with an abnormal pH always represents 2 primary acid base disorders.
Acid-base homeostasis is vital to life. Disruptions in the balance of Hydrogen ions within the body, and within the cell cause problems that are often the symptoms of disease. Understanding how these mechanisms work gives us two valuable insights, firstly it aids in diagnosis, and second it can guide treatment.
Everything begins with water.
We are all approximately 60% water (though I have evidence that some ED Registrars are made entirely from gin). Water dissociates as follows
H2O <–> H+ + OH–
The dissociation of water is constant, and is dependent on temperature, at high temperatures you get more dissociation, this means you get more liberated Hydrogen ions, and so the acidity goes up.
If we are 60% water, that’s a lot of available, easily liberated hydrogen. We don’t measure hydrogen ion concentration, we measure the negative logarithm of proton concentration (because some bright spark thought that was easier).
Normal mean INTRACELLULAR pH = 6.8
ECF [pH usually aprx > 7.3] – Contains cells, particles, dissolved gases, fully and partially dissociated ions
Normal mean ARTERIAL pH = 7.4
So it’s important to realise that what we are measuring and what we really care about, are again, different things. We should care about the intracellular pH, but we can’t measure that, so we measure the arterial (or venous) pH of blood, and use that as a surrogate for extracellular fluid (ECF) which is, in itself, a surrogate for intracellular pH.
Determinants of pH
Numero uno; CO2
We make a lot of CO2 each day from aerobic metabolism and the trusty Krebs cycle (12 500 mEq/day). It goes from the cell to the ECF, to the blood, to the lungs where it follows it’s partial pressure gradient out into the atmosphere. For us to understand CO2’s effect on plasma pH we have to plug it into our formula.
CO2 + H2O <–> H2CO3 <–> H+ + HCO3–
Using this, and the formula for pH we can derive the following formula
pH = 6.1 + log10([HCO3–]/αPCO2)
This equation is the Henderson Hasselbach equation (not the top one!) and it is how we derive the bicarbonate in a blood gas readout. 6.1 is the pKa (another time), and α is the solubility coefficient of CO2. (which is 0.3, but we all knew that right?)
Important point: HCO3– is a derived value. It is not a measurement.
Numero Dos; Non-volatile weak acid dissociation
Fluid compartments have a soup of molecules that don’t generate CO2. Some of these have acidic properties. They predominantly have a negative charge, and this charge alters in parallel with pH. The main constituents are albumin, and inorganic phosphate (haemoglobin is also a weak acid and does the same thing inside the red cell). This gets expressed with the incredibly complex formula
HA <–> H+ + A–
Now one of the laws of physics comes into play here. That law is conservation of mass
We can define the total amount of non-volatile weak acids in a fluid compartment ATOT,.
ATOT = HA + A–
KEY POINT: ATOT is a constant, it cannot vary with pH, because you would have to convert some of the mass to energy, which would cause your fluid compartment to EXPLODE [e=mc2]. A change in pH signals a shift in the balance between HA and A–.
Numero trois; Electrical Neutrality – and strong ions.
The balance of charge of cations and anions must equal eachother in a dissolved fluid. This is another one of the laws of physics, takes precedence over any desire for pH neutrality.
K+, Na+, Ca2+, Mg2+, Cl– exist as ions dissolved in fluid. We also have lactate, sulphate, and β-hydroxybutyrate acting as strong cations in the plasma compartment.
In human plasma we do not have a balance of cations and anions, this “gap” or “difference” is the Strong Ion Difference [SID], which is expressed mEq/L (because it’s an expression of CHARGE, not concentration).
SID = [strong cations] – [strong anions]
And because of the law of electrical neutrality we can re-arrange that formula to state:
SID + [H+] –[HCO3–] – [CO2-3] – [A–] – [OH–] = 0
Number “whats spanish for four?” – crazy voodoo – aka Gibbs Donnon forces.
This is the name given to the forces that compel dissolved ions to alter their equilibria across semi-permeable membranes (if you want to draw some circles with a dotted line across the middle and talk about concentration gradients…you can, I won’t stop you). The 3 fluid compartments that we are concerned about in this model are the ones inside red blood cells, as well in the plasma, and the ECF. Each compartment contains big molecules which have a charge, and will not diffuse (haemoglobin and albumin I’m looking at you). Red blood cells harbour a lot of negative charge (as they’re packed with Hb) so they attract Na+ and K+, if it wasn’t for the Na/K+ ATPase working against that electrochemical gradient, they’d swell up, pop, and we would all die. Chloride is the major anion, and it is shuttled around by this affect a lot.
Putting this all together.
There are 6 equations that govern, and predict how much H+ there is in the plasma. They are as follows.
Water dissociation equilibrium [H+] x [OH–] = Kw
Weak acid dissociation equilibrium [H+] x [A–] = Ka x [HA]
Conservation of mass for weak acids [HA] + [A–] = [ATOT]
Bicarbonate Ion formation equilibrium [H+] x [HCO3–] = Kc x PCO2
Carbonate ion formation equilibrium [H+] x [CO32-] = K3 x [HCO3–]
Stewart proved (using sexy maths) that 6 of the variables in these equations (HA, A, HCO3–, CO32-, OH– and H+) are reliant on just 3 independent variables: SID, ATOT and PCO2
The Strong Ion Difference.
SID is the gap between anions and cations on this bar chart. It is a ‘charge space’ – that is filled by weak ions dissociating in varying combinations to keep the charge in solution equal. The majority of this work is completed by just 2 components, the weak acids (A–) and the HCO3– . If we add more cations to the mix, the buffer base has to get larger to obey the law of electrical neutrality. If we add more anions, the buffer base has to get smaller to obey the same law.
The pH of human plasma is based on 3 things:
ATOT – the concentration of non volatile weak acids
PCO2 – the partial pressure of CO2
SID – the charge space between cations and anions dissolved in the plasma.
Next week – we’ll talk about “apparent SID” Vs “effective SID”, anion gaps, and what clinical applications all this has!
For more on SID – have a look at this youtube lecture, it really helped me understand this.