"Take the oxygen off, you'll knock out their hypoxic drive" is the most confidently wrong sentence in respiratory teaching. Here's the real mechanism.
The textbook version: chronic CO2 retainers rely on hypoxemia (not hypercapnia) to drive respiration, so giving oxygen removes that drive and causes hypoventilation. This turns out to explain only a small slice of what's actually happening.
Aubier et al.'s landmark study directly tested this by measuring minute ventilation in COPD patients given supplemental oxygen: ventilation dropped transiently, then recovered to near baseline — while PaCO2 kept climbing regardless, with no correlation between minute ventilation and the rising PaCO2. A follow-up study confirmed respiratory drive and minute ventilation both stayed within normal limits on supplemental O2. The conclusion the field eventually converged on: hypoventilation from lost hypoxic drive is not the primary cause of oxygen-induced hypercapnia in COPD.
This is the piece the traditional teaching skips, and it's the biggest contributor by a wide margin.
Poorly ventilated alveoli (mucus plugging, air trapping, destroyed lung units) normally trigger local hypoxic pulmonary vasoconstriction — the lung's own protective mechanism that redirects blood flow away from those units toward better-ventilated alveoli, optimizing overall V/Q matching. High-concentration supplemental oxygen raises alveolar oxygen tension even in those poorly-ventilated units, reversing that protective vasoconstriction. Blood flow returns to alveoli that still can't ventilate effectively, worsening dead space and V/Q mismatch, and CO2 continues to accumulate. The British Thoracic Society explicitly ranks this as the primary mechanism of oxygen-induced hypercapnia in COPD, above both hypoxic drive suppression and the Haldane effect.
A real, quantified contribution — smaller than V/Q mismatch, but larger than hypoxic drive suppression.
Deoxygenated hemoglobin has a higher affinity for carbon dioxide than oxygenated hemoglobin (it's a better proton acceptor). When supplemental oxygen saturates hemoglobin, CO2 gets displaced from hemoglobin binding sites into the plasma, raising PaCO2 independent of any change in ventilation or perfusion. In the original Aubier study, the Haldane effect accounted for approximately 25% of the observed rise in PaCO2 — a real, measurable contributor, just not the dominant one.
No — it's real physiology, just clinically overstated.
None of this changes the actual bedside recommendation: titrated oxygen to an SpO2 target of 88–92% remains the correct approach in COPD exacerbation, exactly as taught. What changes is why: the goal isn't primarily to protect a fragile hypoxic drive from being "knocked out" — it's to avoid reversing hypoxic pulmonary vasoconstriction and worsening V/Q mismatch, with a smaller assist from the Haldane effect. Ripping oxygen off a hypoxic, hypercapnic COPD patient out of fear of "knocking out their drive" is the wrong reflex either way — the treatment for oxygen-induced hypercapnia is titration and ventilatory support if needed, never withholding oxygen from a hypoxemic patient.