I have two teaching points in mind.
(1) Coffee-ground emesis isn’t really a gi bleed.
(2) Big anion gap with normal pH = etoh-related illness, nearly all of the time.
Point one. When I think of a gi bleed, i think of blood everywhere, and the patient might die. This means hematemesis or bright red blood from below (aka brbpr). Coffee-ground emesis is generally a sign of gastritis and the patient is systemically ill from something else – sepsis, dka, uremia, etc. Coffee-ground emesis usually means I don’t have to call GI, surgery, or radiology to stop the bleeding in the next hour.
The important point here is that when you see coffee-ground emesis, find out what is really wrong with the patient. GI bleed is not the primary and final consideration.
Point two. How often do you see a combination of a huge anion gap and (a) a person that doesn’t look like death, and (b) a normal-ish vbg / pH? Almost never. The most common dx I see is alcohol ketoacidosis with metabolic alkalosis, by far. The ddx here is rather narrow, though there are other sexy dx that can cause this combination of acidosis-alkalosis. You can go crazy trying to figure out compensatory changes, delta-delta, base deficit, etc with these labs.
My general approach (when I see labs like these) is check an aspirin level and give iv fluids, sugars, anti-emetics; and repeat labs in 2-3 hours or so. I check ketones too (and sometimes a lactate), but that’s not too important. I would almost-never check for a toxic alcohol; they just don’t present with labs like this (usually an isolated anion-gap metabolic acidosis). After sugar, fluids, and supportive treatment, the labs generally head toward normalization, so that you know you’re on the right track.
The patient got fluids, sugar, zofran. Her labs got better over 12 hours or so. She did fine with a little benzo for her etoh withdrawal. See the figure for improvement in her labs with nothing more than supportive therapy. If someone’s gap goes away with supportive therapy, it wasn’t a serious toxin.
