Situation: A 47-year-old woman is in the intensive care unit (ICU) with acute kidney injury after severe sepsis. Her blood pressure is now stable without vasopressors. She receives intermittent hemodialysis (HD) through a temporary dialysis catheter and parenteral nutrition (PN) by infusion pump through a dedicated lumen of a central venous catheter. The nurse is credentialed by the unit to perform arterial puncture. The client is receiving high-flow nasal oxygen with a fraction of inspired oxygen (FiO2) of 0.60, and the client's partial pressure of arterial oxygen this morning was 240 mmHg. The nurse draws an arterial blood gas (ABG) sample into a heparinized plastic syringe. If a small air bubble were left in this syringe before analysis, how would it MOST likely change the reported values?

Room air has an oxygen pressure of about 150 mmHg and almost no carbon dioxide, and an air bubble pulls the sample's gas pressures toward those room-air values. Carbon dioxide therefore always reads falsely low. The oxygen value moves toward 150 mmHg: it reads falsely high when the true value is below 150 mmHg, but falsely low in a client on high inspired oxygen whose true value is about 240 mmHg.

Preanalytical error in ABG sampling: air bubble effect

A small air bubble left in a heparinized arterial blood gas syringe acts as a contaminant that shifts measured gas tensions toward the composition of room air. Room air contains approximately 21% oxygen, corresponding to a partial pressure of about 150 mmHg at sea level, and essentially 0 mmHg of carbon dioxide. When an air bubble is present, gas molecules diffuse across the blood–air interface until a new equilibrium is approached: oxygen moves from the higher-pressure compartment into the lower-pressure one, and carbon dioxide does the same in the opposite direction.

For carbon dioxide, the direction of error is always the same. Because room air has virtually no CO₂, the bubble continuously draws CO₂ out of the blood sample. This makes the measured PaCO₂ falsely low regardless of the patient's actual CO₂ level. The magnitude of the decrease depends on bubble size, mixing, and time before analysis, but the direction does not change.

For oxygen, the direction of error depends on the patient's true PaO₂ relative to the room-air value of 150 mmHg. If the true PaO₂ is below 150 mmHg—as in a patient breathing room air or low-flow oxygen—the bubble adds oxygen to the sample and the reported PaO₂ is falsely high. However, when the patient is receiving high inspired oxygen concentrations, the true PaO₂ can exceed 150 mmHg. In this scenario, the client is on high-flow nasal oxygen with an FiO₂ of 0.60 and has a documented arterial PaO₂ of 240 mmHg. Because the sample's oxygen tension is higher than that of room air, oxygen diffuses out of the blood into the bubble, and the analyzer reports a value that is falsely lower than the true PaO₂.

The combined effect is therefore a falsely low oxygen value and a falsely low carbon dioxide value, which corresponds to option 4.

Analyte	True value in this client	Room air value	Direction of shift with air bubble	Reported result
PaO₂	240 mmHg	150 mmHg	Moves toward 150 mmHg (downward)	Falsely low
PaCO₂	Normal or elevated in sepsis/AKI	0 mmHg	Moves toward 0 mmHg (downward)	Falsely low

The studies on air contamination in blood gas samples support this bidirectional mechanism. In one investigation, blood tonometered to a PO₂ of 339 mmHg showed a decrease in measured oxygen tension when contaminated with room air, whereas samples with baseline PO₂ values below 150 mmHg showed an increase ^[2]. This confirms that the error is not a fixed overestimation or underestimation but depends on where the true value sits relative to the room-air oxygen pressure. Another study similarly reported that PO₂ increased in samples with lower baseline oxygen tensions when air bubbles were introduced, while the effect reversed at higher tensions ^[1]. The key teaching point is that Key point! an air bubble always lowers PaCO₂, but its effect on PaO₂ depends on whether the true PaO₂ is above or below 150 mmHg. In a patient on high FiO₂ with a PaO₂ of 240 mmHg, the bubble pulls oxygen out of the sample, producing a falsely low oxygen reading alongside the falsely low carbon dioxide reading.

Watch out! The common assumption that air bubbles always increase PaO₂ is only correct when the patient's true PaO₂ is below room-air levels. In critically ill patients receiving supplemental oxygen—especially high-flow systems or mechanical ventilation—the true PaO₂ often exceeds 150 mmHg, reversing the direction of the oxygen error. This is a frequent source of confusion in exam questions that pair an air bubble with a high-FiO₂ clinical scenario.

The presence of a temporary dialysis catheter and parenteral nutrition does not directly alter the gas diffusion physics of the air bubble; those details establish the ICU context and the reason for frequent ABG monitoring. The heparinized plastic syringe is appropriate for ABG collection, but heparin does not prevent the gas exchange caused by an air bubble. The nurse's credentialing for arterial puncture is relevant to the sampling procedure but does not change the preanalytical error mechanism.

In summary, the air bubble equilibrates the sample with room air. Carbon dioxide always falls because room air has almost no CO₂. Oxygen falls in this client because the true PaO₂ of 240 mmHg is higher than the room-air value of 150 mmHg, driving oxygen out of the sample. The reported ABG would show both oxygen and carbon dioxide falsely low.