ABG or VBG?

by Bob Stuntz


I got a good question recently via “Ask me anything”:

Is the ABG back? I’ve been preaching to our residents that the ABG is a useless ED test unless you’ve been holding that intubated ICU patient in the ED for 12 hours and their sat is still 70%. Stuart Swadron did a great interview on EMRAP this month about the danger of hyperoxia. Do we need to go back to measuring ABGs on our sick patient’s to ensure that their PAO2 is not too high?

This is obviously a popular topic.  I have heard people get very fired up one way or the other about this.  The VBG is usually obtained faster, less painful, and with less potential complications, so it is an attractive test.  But can we really throw out an ABG?

I have referenced what I think are some of the relevant studies below.  Here is my take home and personal opinion about this:

1. If you need to measure pH, bicarb, or base excess, especially in hemodynamically stable patients, a VBG is equivalent and there is no need for an ABG.  On average for pH the difference is about 0.02 - 0.03.  So for your average DKA patient, no ABG is needed.  Hypotension and hypoperfusion does make the correlation between the two for these values somewhat worse, but I don’t believe would throw you way off.   

2. A VBG is a good screening tool for hypercarbia, with a venous pCO2 < 45 mmHg having about 100% sensitivity.  However, VBG cannot tell you the degree of hypercarbia reliably.  If you need an absolute number, ABG is the way to go.  In my opinion, if you have a high VBG pCO2 and that fits the clinical picture, who care what the number is, just start ventilating them.  Again, shock does seem to lead to a worse correlation between the two.    

3. If you are concerned for hypoxia, or your workup requires calculations of oxygen tension, you need to do an ABG (hypoxic arrest, calculating A-a gradients)

4. Studies are raising the concern for increased mortality in patients who have hyperoxia after return of circulation after cardiac arrest.  These patients need ABG’s to guide your vent settings.  The increased mortality in these patients has even been seen within the first hour, so it is important not to punt this to the ICU.  We need to be the ones jumping on this.  

5.  If you are managing your vent settings in the ED post intubation, you will likely need an ABG to manage your pCO2 and pAO2.  

Having talked to colleagues, it seems many have a protocol where they get an ABG about 30 minutes after intubation to guide vent settings.  

Bringing up a side point, while we’re talking pCO2: please be careful with your COPD patients.  Remember that your COPD patients rely on their low oxygen levels for their respiratory drive.  As you give them higher FiO2, and they have higher oxygen content, they will breathe less.  Less breathing leads to higher CO2 levels.  Higher CO2 levels leads to hypercarbic respiratory arrest.  Be comfortable with oxygen saturations of 88-92% in your COPD patients.  More and more studies are showing too much oxygen can be a bad thing, especially in COPD and ROSC arrest patients.  More on this down the road.  

Have a question, a different opinion, know more than I do, or know of a study that proves me wrong?  Great!  Leave a response to this post, send me a message, or post a reply of your own!  

1. Emerg Med Australas. 2010 Dec;22(6):493-8. doi: 10.1111/j.1742-6723.2010.01344.x.  Review article:  Can venous blood gas analysis replace arterial in emergency medical care.  Kelly AM.

2. Emerg Med Australas. 2006 Feb;18(1):64-7.  The case for venous rather than arterial blood gases in diabetic ketoacidosis.  Kelly AM.

3.  Am J Emerg Med. 2011 Sep 9.  Venous vs arterial blood gases in the assessment of patients presenting with an exacerbation of chronic obstructive pulmonary disease.  McCanny PBennett KStaunton PMcMahon G.

4.  J Res Med Sci. 2011 Feb;16(2):188-94.  The effects of hypotension on differences between the results of simultaneous venous and arterial blood gas analysis.  Shirani FSalehi RNaini AEAzizkhani RGholamrezaei A.

5.  JAMA. 2010 Jun 2;303(21):2165-71.  Association between arterial hyperoxia following resuscitation from cardiac arrest and in-hospital mortality.  Kilgannon JHJones AEShapiro NIAngelos MGMilcarek BHunter KParrillo JETrzeciak SEmergency Medicine Shock Research Network (EMShockNet) Investigators.