This is the 3rd part of a 4 part series on acid base.

You should listen to Acid-Base Part I first where you will learn about the underlying chemisty of acid base. Part II then delves into the underpinnings of the mathematics of acid base. In part III, we will go through two actual problems and show how the EMCrit method plays out. Part IV delves into the acid-base of solutions.

Before we get to the clinical stuff, I am giving three lectures at the 8th annual NY Symposium on Neurological Emergencies and Neurocritical Care. Should be a great conference. If you are free for some of the days between June 14-17, 2011; consider coming.

Ok back to acid base stuff.

For this podcast to be optimally effective, you need to print out my acid base sheet:

###### EMCrit Acid Base Method

### Here is the 1st problem from last podcast:

### Here is the same patient after we treated his DKA:

Mike asked if there was any literature to support the simplification I am using to make the incredible complex quantitative formula more approachable. The answer is yes and here is the pdf you want to read:

##### Story DA, Morimatsu H, Bellomo R. Strong ions, weak acids and base excess: a simplified Fencl-Stewart approach to clinical acid-base disorders. Br J Anaesth. 2004 Jan;92(1):54-60.

Want an incredible program that will do all of the work for you and teach you about the quantitative method at the same time? Look no further than this incredible site:

**Need an Audio Only Version?**

Acid Base Part III MP3 (Right Click and Choose Save as)

Podcast: Play in new window | Download (136.8MB) | Embed

Subscribe: iTunes | RSS | More Subscribe Options

Casey says

Hi Scott,

Loving this simple method, makes acid-base understandable rather than just rote!

For the Aussie readers it is useful to change the “albumin correction” component to:

(42 – [albumin])/4 as we always get this is g/L – all works the same, but it seems easier without the decimal point.

C

Casey says

Hi Scott – got a question. I see a lot of bad, non-compliant dialysis patients who walk around with K+ of 7 on a good day (unfun), then get septic on a bad day. If I am using this approach, and don’t wan to ignore the K+, do I just add in the K value to the Na – then subtract 42 instead of 38 to get the SID?

Let me know if this is right

Thanks

Casey

Scott Weingart says

Yes, Casey that is exactly right.

Greg Guldner says

Scott, in patients with significant hyperglycemia should the sodium be corrected for the hyperglycemia prior to the calculation of SID or does it make a difference? The acidbase.org calculator seems to do this.

Thanks.

Greg

emcrit says

For the acid base purposes, never correct the sodium, b/c the sodium level is real. Cl drops with the sodium to maintain SID. Correct sodium with hyperglycemia only to find out if the pt has hypo or hypernatremia.

David says

Hello Scott! Im a medical student in Sweden, where emergency physicians doesnt exist as a speciality. I really appreciate you going through acid base disturbances in this systematic way. However there is something i still dont get(maybe the main point).

For example, i saw a patient with chronic renal failure in the ED, who presented with weakness and tiredness thats been going on for about a week. The ABG came back with

pH 7,14

pCO2 3,0 kPa

BE -20

Na 131

K 7,6

Cl 115

Lactate 0,6

So the SID should be 131 – 115 = 16 right? SID – 38 becomes -22, which accounts for a little bit more than the whole BE. So there are no unknown component.

So my question is: Does the renal excretion of bicarb cause the SID to become low, or is the low SID in itself causing the acidosis? Or am i thinking in a completely wrong way?

Thank you for your podcasts, they are invaluable.

/David

Gina says

When calculating the SIG, particularly in DKA, is their any use of incorporating betahydroxybutarate in the calculation? If you are subtracting lactate in the setting of sepsis to take into account its presence in the “anion gap” it would seem logical that one could do the same with Bhob in DKA – does this line of reasoning play out in reality?

emcrit says

it does if you have a quantitative b-hydroxy. It comes right out of the SIG just like lactate as you say.

Fergus Thornton says

Anybody know the conversion for Alb in Australia? .25 perhaps?

Alb measured g/dL in US, g/L in Oz.

Thanks.

emcrit says

multiply US units by 10

Fergus Thornton says

I meant the “conversion” aspect i.e. multiplying by 2.5 in the US as Scott does above. Since the difference is 10, would the conversion in the formula be .25?

Fergus Thornton says

Never mind, I answered my own question: .25 X 40 = 10.

Piece of cake!

Chris says

Hey Scott,

Absolutely love your blog, my favorite of all the medical blogs out there right now. Regarding the quantitative acid-base calculations, what do you suggest when we are doing calculations from blood gases, but do not yet have an albumin level (ie: in the initial phases of resuscitation)?

I learned the H-H way with the modification of adjusting the normal anion gap based on serum albumin, but that seems to be one of the very minor points when we were taught it as medical students, when in fact it is critically important.

Cheers,

Chris

emcrit says

nothing to do except to assume the albumin is normal

HASSAN ALMAATEEQ says

thank you for the new experience with Blood Gas stuff.. I know you made it very simple in this case.. what if I use the Na:Cl ratio (1.24-1.4) to look for Normal gap acidosis.. here it will guide me in the second blood gas (after correction of DKA), although it didn’t work well in the initial blood gas .. What do you think ?

emcrit says

ratios don’t hold up. Na 140 Cl 100 is VERY DIFFERENT acid-base wise than Na 70 Cl 50.

scott

HASSAN ALMAATEEQ says

Thank you for your response.. One more question: Will Winter’s formula (1.5XHCO3+8) give similar results of expected pCO2 here?

Adam Kessler says

Dr. Weingart,

As apart of our didactics today we went through your Acid-Base lectures 1-4 with the printed handouts. We then used your method and looked at a couple of random ICU patients’ labs and we tried to see if your method would have changed anything we did. In doing so we got a little confused.

In one patient who had a profoundly elevated lactate (~10), we found that our SIG came to 15. We think this means that there are still 15 unmeasured anions floating around somewhere that we were unable to account for. Is this the correct assumption?

The next patient was someone who had a pH of 7.43 and their SIG came out negative, but no other metabolic issues were present. They were however having respiratory issues. Does your method work with these patients.

Also, we found that either way it would not have changed our treatments. The residents just wanted some clarification and some guidance and thought you may have a comment.

Thanks for all your work, I love the podscast.

Adam

emcrit says

Adam,

Can you send me the numbers for both cases and then I’ll take a look.

Anand Senthi says

loving learning this quantitative approach to acid-base – thanks. Have a question:

in your podcast you reunited H-H with quant approach by showing how you can just correct for albumin/lacatate and consider the SID with H-H and you get the same results as quant approach. However am I right in thinking that H-H will still have an inaccuracy because it uses bicarb instead of base deficit?

The base deficit effectively takes out the effect of CO2 changes on bicarb. However using bicarb doesn’t. I’ve tried using both approaches with real patient gases and found them a bit disparate (after taking into account albumin/lactate) and I believe it is due to the bicarb and base deficit difference. I even tried to calculate an expected bicarb by using a formula to take out the effect of the CO2 change but this didn’t solve the disparity.

Thoughts?

emcrit says

the results should almost always be identical. next time you have some real lab values that are drawn simultaneously (vbg, chem, and alb) check it out and comment here if they are disparate.

Anand Senthi says

ok here are the numbers I have from the Australian branch of Janus General. 3 sets of results at different times on the same patient. Albumin was 45 (4.5 in your units) at the first draw. After that not measured but shouldn’t matter as the correction for albumin the same in each method.

Time: 21:00 18:50 17:28

Units Ref Range

pH: 7.18 7.20 7.29 (7.35 – 7.45)

pCO2: 43 29 26 mmHg (36 – 44)

pO2: 30 53 42 mmHg (80 – 95)

HCO3: 16 11 12 mmol/L (20 – 26)

Base XS: -12 -16 -13 mmol/L (-2 – 4)

Saturation: 46 78 71 %

Sample Type:Venous Venous Venous

Sodium: 134 142 137 mmol/L (136 – 145)

Potassium: 13.6 5.4 6.1 mmol/L (3.4 – 4.5)

Chloride: 109 112 106 mmol/L (98 – 107)

Glucose: 25.0 27.0 33.0 mmol/L (3.0 – 8.0)

Lactate: 5.5 4.5 4.9 mmol/L (< 2.0)

Anand Senthi says

So 17:28:

Quant method: 13+ (31-38) + 0.25 (42-45) – 4.9 = 13 – 7 – 0.75 – 4.9 = 0.35

H-H Method: 137-106- 12 + 0.25 (42-45) – 4.9 = 13.35 which is 1.35 above a normal AG rather than 0.35. Or I suppose if you say a normal AG is 11.5 by taking out lactate of 0.5 it is 1.85 above this. Depends what you call the real “normal” AG.

In any case there is a slight discrepancy.

When I calculate it for the the 2nd set of bloods at 18:50 I get SIG of 2.75 v’s AG of 13.75.

When I calculate it for the 3rd set of bloods at 2100 I get SIG -7.25 and AG of 3.75.

Working off an assumption that normal AG is 11.5 (factoring in removed lactate of 0.5) then the discrepancies are between measured SIG and AG difference from 11.5 is -1.5 (0.35 v 1.85), 0.5 (2.75 v 2.25) and -0.5 for the 3 sets of blood respectively.

For those calcs for 2nd and 3rd bloods I have assumed the same albumin which should be fine as the correction for albumin is the same in each calculation.

Also would it be better just to subtract a 45 x 0.25 for albumin (rather than (45-42) x 0.25) to work out the actual AG (i.e working off the premise that there is no “normal” AG) and compare to SIG? If I do this I get 2.85, 3.25 and -6.75 for the AG’s for the 1st, 2nd and 3rd set of bloods respectively. This gives SIG v AG disparities of -2.5, -0.5, -1 respectively.

Thoughts on these discrepancies? I hope I’m working all this out correctly.

Anand Senthi says

Another question re the normal anion gap. I was confused re your podcast and other sources showing normal AG as 12 (+/-4) yet other sources were stating about 8 (+/-4). I now realise this is due to the modern day machines using ion selective electrodes instead of the older flame photometry resulting in chloride being measured about 4 mmol/L higher than previously as outlined in this recent article

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3681403/

This raises a couple questions:

1. When you explained the normal AG in the podcast with albumin occupying about 10.5, 0.5 for lactate and about 1 for other anions this calculation no longer makes sense as it equates to 12 not 8. I note that the conversion factor for albumin is probably 2.3 more accurately rather than 2.5 but that still won’t make the calculation measure closer to the AG of 8. So how can we explain the newer AG of 8 with albumin/lactate?

2. Does the fact that modern daymachines are measuring cloride 4 higher mean when using the quantitative approach and calculating the SID we should be using 34 instead of 38?

Thanks Scott

J Kim says

Ok, i just spent enormous amount of brain energy for days looking at all sorts of acid-base papers to understand this physiochemical Stewart approach of acid-base thanks to your lecture, and thank you for that so much. I’ve finally come to grasp this and I think I found better references for better way of looking at this.

Although I think your approach is fine, I think the approach suggested by Gunnerson (Clinical Review: the meaning of acid-base abnormalities in ICU, Critical Care October 2005) is much better. They basically use the simplified calculation of SIG, derived from Dr. Kellum’s paper (Determinants of blood pH in health and disease, Crit Care 2000), using the equation (Na+K-Cl-HCO3-2.5(albumin)-lactate). This equation is slightly more simpler than your equation and conceptually, it overall represents the true absolute value of SIG (the unmeasured anions) much better in my brain than your equation because it takes out everything in the anion side of bar you showed us except the unmeasured anions. I saw the reference you gave where you derived the equation and I’ve looked at other papers, and I’ve come to the conclusion that the Kellum equation is superior. The equation had excellent correlation to true SIG values in their paper.

Not only is it simpler to calculate but they actually gave me this framework of looking at metabolic acidosis with 3 classifications based on quick bedside calculations; the lactic acidosis, SIG acidosis, and hyperchloremic acidosis, based on which part contributes >50% to base excess. If none contributes to >50% to base excess, it’s labeled as ‘mixed’ acidosis, again, much simpler and intuitive.

Anyway, I would like you to check out the mentioned articles (I think you will love them if you haven’t read them) and, as always, thank you so much for introducing me this whole concept and giving me the impetus to do my own extensive research on this area.

emcrit says

The formulae are exactly the same. If you understand why, then you are really starting to get this stuff. Of course, I’ve read Kyle’s paper. If that version works better for your mind, go for it. It is what I call the corrected Henderson Hasselbach method, which I mentioned during the podcasts–anion gap corrected for weak acids. This method won’t help unless you add in some analysis of non-sig acidosis as well.