By Beau Stokes, MD, PGY-2, with commentary by Dr. Sage Whitmore, MD
The topic of acid base physiology came up in a recent M&M. I know that one of my biggest weaknesses was acid/base physiology until I sunk a bunch of time into reading about it. It was just conceptually hard for me to wrap my mind around it for a while. I’m hoping that this post can help save other people the amount of time it took me to sort through all of this complicated physiology.
For the longest time, I couldn’t understand why giving bicarb wasn’t the treatment for acidosis. There have been multiple studies that have shown in various situations where giving bicarb in acidosis not only does not improve mortality, but can even increase mortality. It seemed backwards that giving a base to someone with acidosis could actually increase mortality.
It turns out that it all comes down to physiology. When you give someone bicarb, it buffers pH through the equation we all learned back in med school: HCO2 + H < > H3CO2 < > H2O + CO2. When you give someone bicarb, the only way it can make the pH more “basic” is by turning the H+ ions into water and generating carbon dioxide. However, if that patient can’t breathe out the carbon dioxide, they will just become hypercapnic, and the excess CO2 will diffuse intracellularly which traps acid inside body compartments, like the CSF and myocardium. Most acidemic patients already have increased respiratory rate and are already maximally compensated from a respiratory standpoint. If you give them bicarb, all you are doing is causing increased CO2 that they can’t ventilate out. This will then cause worsening hypercapnia and potentially worsened acidosis (which is a possible explanation for the paradoxical increased mortality in some patients given bicarb).
This also applies to the fluids we give. There has been a movement towards LR as a more “balanced” fluid with less potential to cause acidosis. This is not because of the lactate conversion to bicarb in the liver, but rather because of the composition of LR. We’ve all heard by now that normal saline has the propensity to cause a hyperchloremic metabolic acidosis, which you don’t get with LR. If you also look at the pH of LR versus normal saline, the pH of normal saline is around 5.5, and the pH of LR is around 6.5. The strong ion difference of LR is also closer to physiologic range compared to normal saline. There’s just less acidosis caused with LR and it is indeed a more balanced fluid, although as we can imagine, there would be certain clinicaly situations depending on the physiology of the patient where NS would be preferred. Although there is not much data about improved mortality in these acidotic patients with LR vs. NS (that I could dig out, seems limited), physiologically it just makes sense to use LR more that we probably do.
This is a great podcast for acid/base of fluids and how they effect physiology: http://emcrit.org/podcasts/acid-base-4-use-of-fluids/
This is a great podcast for why giving bicarb physiologically doesn’t help in these patients: http://emcrit.org/podcasts/enough-with-the-bicarb-already/
I’d also recommend listening to Jan. 2015 EMRAP they had a nice segment about giving bicarb to these patients and why it doesn’t work. https://www.emrap.org/episode/2015/january/criticalcare (starts at 19 minutes in for DKA management, 24:30 for why not to use bicarb).
Commentary by Dr. Sage Whitmore, MD (critical care trained EM faculty):
I completely agree. Pushing bicarb is indicated for NO OTHER condition than for frank hemodynamic collapse. A sudden jump in HCO3 and pH can cause decreased oxygen delivery to tissues (leftward shift of oxy-hgb curve and tissue arteriolar constriction), increased lactate production, decreased ketone clearance, hypocalcemia, and hypokalemia. “Slow push” might be needed for QRS widening with hyperkalemia or sodium channel blocking agents, or as a poor-(wo)man’s hypertonic fluid for impending brain herniation. In NON-anion gap metabolic acidoses, as well as renal failure (bicarbonate wasting processes), you can give bicarb but in the form of a drip: 150mEq (3 amps) in a liter of D5W. If your patient is on a ventilator, ALWAYS be thinking about how you’re going to ventilate the iatrogenic CO2 load off. Otherwise, just say no; never for a lactic acidosis or respiratory acidosis, unless there is cardiovascular collapse. Not to mention, most patients can tolerate a much worse pH than you as their physician can tolerate to look at on your blood gas
Regarding NS versus LR, there was just a recent meta-analysis associating mortality from severe sepsis/septic shock with normal saline use, and plenty of other literature on worse systemic inflammation, worse bleeding, worse acid base status, worse renal function, increased need for dialysis, etc. all associated with normal saline.
NOW, for the exceptions: there are FIVE groups of patients who cannot tolerate acidemia or permissive hypercapnia, and we must correct their pH whether by improving their minute ventilation or by giving bicarb: Cardiogenic shock, pulmonary hypertension, elevated ICP, certain overdoses—specifically salicylate and sodium channel poisons, and pregnancy. Furthermore, there are groups of patients for whom LR is NOT appropriate (due to its lower sodium concentration): intracranial pathology at risk for cerebral edema, chloride losses such as severe vomiting, hypovolemia accompanied by alkalemia, adrenal failure (hyponatremia/hyperkalemia), hypercalcemia, and in small children/infants. Furthermore, for logistics, know that LR cannot be run in with blood products (but plasmalyte can), and additives such as KCl are not easily added (which is why I begrudgingly still use half-normal saline for DKA).