Situation: A 68-year-old man with chronic obstructive pulmonary disease (COPD) and a 45 pack-year smoking history is admitted to the medical ward with an exacerbation: 3 days of increased breathlessness and more purulent sputum. His records show chronic carbon dioxide retention. On room air, his oxygen saturation by pulse oximetry (SpO2) is 84%. On day 3 he is on a nasal cannula at 1 L/min with an SpO2 of 90%. A worried relative turns the flowmeter up to 6 L/min. Two hours later the nurse finds him hard to rouse. He complains of headache, his skin is warm and flushed, his respiratory rate has fallen from 22 to 10/min, and his SpO2 is 98%. What is the MOST likely explanation?

In a client with chronic carbon dioxide retention, oxygen above the target worsens ventilation–perfusion matching and reduces carbon dioxide carriage by hemoglobin, so the arterial carbon dioxide level rises. Drowsiness, headache, warm flushed skin, and a falling respiratory rate with an SpO2 above target are signs of carbon dioxide narcosis. The nurse returns the flow to the prescribed rate without stopping oxygen, obtains an arterial blood gas, and notifies the physician.

Why the patient became drowsy after the oxygen was increased

The most likely explanation is that the high oxygen flow raised his arterial carbon dioxide level, producing carbon dioxide narcosis. This patient has chronic CO₂ retention from long-standing COPD. His respiratory drive no longer depends primarily on a falling PaO₂; instead, it is driven partly by a chronically elevated PaCO₂ and partly by hypoxemic stimulation of peripheral chemoreceptors. When a relative increased the nasal cannula from 1 L/min to 6 L/min, the inspired oxygen concentration rose sharply and his SpO₂ climbed to 98%. That level is well above the target range for a known CO₂ retainer.

In a patient with chronic hypercapnia, excessive oxygen worsens ventilation–perfusion matching and reduces carbon dioxide carriage by hemoglobin, so the PaCO₂ rises further. Two mechanisms are involved. First, high inspired oxygen abolishes the hypoxic pulmonary vasoconstriction that had been directing blood away from poorly ventilated lung units. Perfusion then increases to areas with poor ventilation, creating more dead-space ventilation and allowing CO₂ to accumulate. Second, when hemoglobin is already highly saturated, less reduced hemoglobin is available to carry CO₂ as carbaminohemoglobin, and the Haldane effect shifts CO₂ transport so that more CO₂ remains dissolved in plasma.

The clinical picture supports hypercapnia rather than improved rest or worsening hypoxia. The patient is difficult to rouse, complains of headache, has warm flushed skin, and his respiratory rate has fallen from 22 to 10/min. Drowsiness, headache, warm flushed skin, and a falling respiratory rate with an SpO₂ above target are signs of CO₂ narcosis. If the problem were simply better oxygenation allowing rest, the respiratory rate would not drop to 10/min with obtundation. Worsening brain hypoxia is also unlikely because the SpO₂ is 98%. Respiratory muscle fatigue from a new pneumonia would typically present with tachypnea and increased work of breathing, not bradypnea and depressed consciousness.

Watch out! A high SpO₂ in a known CO₂ retainer is not reassuring. The target SpO₂ for most COPD patients with chronic hypercapnia is 88–92%. A reading of 98% signals oxygen overshoot.

Key point! The nurse should return the flow to the prescribed rate without stopping oxygen completely. Abruptly removing oxygen can cause dangerous hypoxemia. An arterial blood gas should be obtained, and the physician must be notified.

The case reports reinforce this concern. One describes a patient with COPD who self-administered high-flow oxygen and developed hypercapnic respiratory failure with severe respiratory acidosis requiring ICU admission and non-invasive ventilation ^[1]. Another case series notes that high-dose oxygen therapy with FiO₂ greater than 50% can induce significant hypercapnia in elderly patients even without an identifiable additional cause ^[2]. These reports support the principle that unsupervised increases in oxygen delivery are hazardous in patients with chronic CO₂ retention.

Finding	Supports CO₂ narcosis	Does not support
SpO₂ 98% on 6 L/min	Oxygen overshoot above target	Rules out worsening brain hypoxia
Respiratory rate 10/min	Depressed respiratory drive from rising PaCO₂	Not typical of fatigue or pneumonia
Drowsiness, headache, flushed skin	Cerebral vasodilation and CO₂ narcosis	Not explained by restful sleep

The priority nursing action is to reduce the oxygen flow to the prescribed rate while maintaining some oxygen delivery, assess the patient’s level of consciousness and respiratory effort, obtain an arterial blood gas to confirm the PaCO₂ and pH, and notify the physician for further management such as consideration of non-invasive ventilation if respiratory acidosis is severe.