To recap
50 something man admitted to our service after being found unresponsive. He is well known to our hospital, is homeless, an alcoholic and has a seizure disorder.
In the ER he is found to have bilateral lower lobe pneumonia, and an alcohol level of 426. He becomes hypotensive, requires intubation and pressors.
The resident presents these laboratory tests:
| Na | 138 | Cl | 109 | BUN | 15 | glu | 151 |
| K | 3.9 | CO2 | 15 | creat | 1.0 |
On 50 % oxygen:
| pH | 7.29 |
| pCO2 | 33 |
| pO2 | 86 |
| HCO3 | 16 |
Over 24 hours we were able to extubate the patient and discontinue the pressors. He became alert, but without any memory of his admission.
I had speculated about his acid base status on presentation. I needed more information. His albumin was 4. Previous electrolyte panels were normal (no previous RTA). The patient had had no diarrhea, but some vomiting.
Day 2 he had these numbers:
| Na | 138 | Cl | 106 | BUN | 12 | glu | 99 |
| K | 4.0 | CO2 | 24 | creat | 0.7 |
On 6 liters oxygen:
| pH | 7.43 |
| pCO2 | 40 |
| pO2 | 112 |
| HCO3 | 27 |
CXR was consistent with aspiration pneumonia – no ARDS. Given this information:
1. What was his likely acid-base disorder and etiology?
Clearly he had a metabolic acidosis. With a gap of 14 and a HCO3 of 15 my resident first suspected a normal gap acidosis. The patient had no diarrhea or previous GI surgery, so the resident suspected an RTA. However, old records confirmed that he had normal acid base balance within the last couple of months, making a new RTA less likely.
The patient had some ketones in his urine, supporting alcoholic ketoacidosis or starvation ketosis. I suspect (but could not prove because the patient had such poor venous access) that the patient had a mild lactic acidosis. He was hypotensive and on pressors and thus had major risks for lactic acidosis. I remembered this article – The anion gap does not accurately screen for lactic acidosis in emergency department patients
Note that on day 2 the anion gap is 8.
Now I am not certain if I am correct. I present this patient as a challenge, because trying to understand the labs challenged me. I have presented my hypothesis, and welcome your critiques.
Several teaching points are relevant to acid base 301 or 401. First, we can only estimate the normal anion gap – normal is not an absolute concept here. Second, we the clinical situation suggests lactic acidosis do not be dissuaded by a "normal gap". Third, in this patient my obsession over the acid base disorder did not help the patient. We treated the patient without regard to the acid base problem and he and the acid base disorder resolved.
{ 3 comments… read them below or add one }
Blood gas people (you know who you are):
ABG interpretation is becoming a lost art (as is blood gas drawing) which is regrettable because in sick patients it is the single most valuable test we have, Give me one test in a sick patient and I’d make it an ABG. You can tell more form it than a CBC or a chem profile. A couple of basic principles that go along with this case:
1. Blood gases are like any other test, EKG, CT x-ray – it’s best to interpret them in a clinical context. I can guess as good as the next guy but that’s what it is without clinical info- guessing. Interpretation is looking at the numbers with clinical information and saying what it means.
2. Don’t be a slave to formulas. some of your readers may argue with me – but formulas have limited value. IF you see a lot of ABGs – you are going to see a lot of divergence form the classical formula. They are useful but not dispositive (as this case indicates). My opinion is ABG interpretation often gets bogged down in formulas at the expense of good common sense. knowing the formulas is good but knowing their limitations is better. Blood gas formulas, indeed all clinical formulas, are theories. And the first rule is don’t fit the facts to the theory- fit the theory to the facts. (It’s like the fluid formulas they give you for hypernatremia – some of the formulas introduce correction factors of 25% – once your formula is off by 25% it’s not much good clinically)
3. Simple is better- old Occam was right most of the time. There are extremely complicated blood gases with double and triple disturbances – but they are the exception, not the rule. It’s best to remember that especially when teaching new students. Learning basic acidosis and alkalosis and acute and chronic compensation takes care of 95% of ABG interpretation. That other 5% is left to experts.
4. Trends are more important than single values. You learn more from the movie than you do from the snapshot. In this case the first ABG could have radically different implications for the patient. No one can say which of those implications is likely to be right until you have at least one more (and sometimes several more). It turned out to be a more benign interpretation than the worst-case situation but only a follow-up ABG could tell you that.
Good case, quite instructive, like to hear what others think.
I notice that you often take the CO2 value (when available) to compute the anion gap, instead of using HCO3. Since there’s always a difference of 1-2 between those 2 values, this can put a gap over the top and make the difference between non-anion and anion gap acidosis…
What’s your take on this. Use CO2 when available ? Always use HCO3 ? Don’t bother since a difference of 1-2 mmol/L is not clinically relevant ?
I vote for #3. I can’t imagine too many times where it would make a difference.