ABG Interpretation Doesn't Have to Be a Rite of Passage
I spent years watching residents and even attending physicians mess up basic acid-base cases because they tried to memorize algorithms instead of understanding what was actually happening. The traditional approach of step-by-step flowcharts tends to break down when you encounter mixed disorders, which is essentially every real patient. What actually works is building a mental model of how the lungs and kidneys interact with bicarbonate and CO2, then practicing until you stop second-guessing yourself on straightforward cases. First, pull the sample. You need arterial blood, not venous. Radial artery stick is standard, though brachial and femoral work when radial access fails. Once you have the sample, make sure it goes to the analyzer immediately or sits on ice if there will be a delay. Red blood cells keep metabolizing after the draw, so a sample that sits at room temperature for 30 minutes can show a pCO2 that is artificially high and a pH that drops by about 0.03 units. That matters when you are trying to distinguish a mild respiratory acidosis from a normal variant. Here is what you look at in order. Check the pH first. Normal range is 7.35 to 7.45. If it is below 7.35, the patient is acidemic. Above 7.45 means alkalemic. Then look at the pCO2. This is the respiratory component, and it normally runs between 35 and 45 mmHg. A low pCO2 means the patient is blowing off CO2, which pushes pH up. A high pCO2 does the opposite. Next, check the bicarbonate, either the calculated HCO3 or the standard bicarbonate value from the analyzer. Normal is 22 to 26 mEq/L. Low bicarbonate points toward metabolic acidosis, high bicarbonate toward metabolic alkalosis.
The hard part is figuring out whether compensation is appropriate or whether a second disorder is hiding in plain sight. For acute respiratory acidosis, bicarbonate goes up about 1 mEq/L for every 10 mmHg rise in pCO2 above 40. For chronic respiratory acidosis, that number climbs to roughly 3.5 to 4 mEq/L per 10 mmHg. Acute respiratory alkalosis drops bicarbonate by about 2 mEq/L per 10 mmHg fall in pCO2. Chronic respiratory alkalosis drops it by 4 to 5 mEq/L per 10 mmHg. These numbers are guides, not laws, but they catch most mixed disorders quickly. I remember a case a few years ago where a patient came in with COPD exacerbation and the initial ABG read pH 7.31, pCO2 68, bicarb 30. Looks like acute on chronic respiratory acidosis at first glance. But when I ran through the compensation math, the expected bicarbonate for an acute rise from 40 to 68 in pCO2 would only be around 33, and this patient was already at 30. That meant there was a concurrent metabolic acidosis dragging the bicarb down. I checked the lactate and it was 5.2. Sepsis from a pulmonary source. The initial reading had almost tricked me into missing the metabolic component because the elevated bicarbonate from chronic compensation masked it.
Common Pitfalls and What They Actually Mean
One thing people consistently get wrong is assuming that a normal pH rules out a problem. A patient can have a profoundly deranged acid-base status with a near-normal pH if two disorders are canceling each other out. A patient with vomiting and diarrhea, for example, can present with metabolic alkalosis from the vomiting and metabolic acidosis from the diarrhea, landing somewhere around pH 7.40 while both processes are actively damaging them. Always look at the individual values, not just the pH. Another trap is relying on the anion gap without adjusting for albumin. The standard anion gap formula is sodium minus chloride minus bicarbonate. Normal is about 12, but that assumes a normal albumin level of 4.0 g/dL. For every 1 g/dL drop in albumin below normal, the anion gap drops by roughly 2.5 points. A patient with an albumin of 1.5 and an anion gap of 10 actually has a corrected anion gap closer to 16, which means there is a hidden high anion gap metabolic acidosis that would be missed entirely. I once had a situation where the portable blood gas analyzer gave a reading that did not match the laboratory reference. The pO2 on the ABG was 85 mmHg while the patient was on 2 liters nasal cannula, which should have been higher. It turned out the sample had a large air bubble trapped near the needle hub. Air has a pO2 of about 150 mmHg and a pCO2 near zero, so even a small bubble skews the results. The fix was simple: redraw the sample, expel any visible bubbles through the needle before connecting to the syringe, and double-check that the plunger is seated tightly. I started doing a visual bubble check on every sample after that and it eliminated about half the repeat draws I used to do.
Get the Full Details
What This Approach Cannot Do
ABG interpretation is useful for assessing acid-base status, oxygenation, and ventilation in the moment, but it has real limitations. It is a snapshot. A single ABG tells you what is happening right now and maybe the last few minutes, not what the patient will look like in an hour. Serial measurements are necessary if the clinical picture is changing. The test also only samples arterial blood from one point in the body. In patients with significant shunting or uneven ventilation-perfusion matching, a radial artery sample might not reflect gas exchange in other parts of the body accurately. Capillary blood gas analysis is an alternative in pediatric populations and some adult situations, but it correlates poorly with arterial values when perfusion is compromised. Cold extremities, shock states, and peripheral vasoconstriction make capillary sticks unreliable. Venous blood gas can give you pH and pCO2 information, which is often sufficient for managing chronic conditions like COPD, but it cannot assess oxygenation and the bicarbonate values run slightly different from arterial samples. For clinicians who want a quick reference tool alongside actual ABG analysis, there are several freely available calculators and interpretive aids online. The WebNephrology.org ABG analysis tool and the MDCalc acid-base calculator are reliable resources that walk through the steps systematically. These are helpful for educational purposes and for checking your own work, but they should not replace understanding the underlying physiology. A calculator will tell you the compensation is appropriate or inappropriate, but it will not explain why a particular clinical scenario produces a mixed disorder.
