The Hypoxic Drive Theory…DEBUNKED!

breathing hypoxic drive

We’ve all heard it at some point or another: “Don’t give that COPD patient too much oxygen”. It has led us to believe that oxygen is a REALLY bad thing in patients with chronic obstructive pulmonary disease (COPD). This has caused nurses to be afraid to put oxygen on a COPD patient with low oxygen saturations (SpO2); yes, even in patients with SEVERE hypoxia! Why? Because of a little something known as the hypoxic drive…

Let’s Start With Respiratory Physiology

Central chemoreceptors are found in the cerebrospinal fluid (CSF) and are sensitive to changes in pH. Changes in pH can occur as a result of changes in the level of carbon dioxide (CO2) within the body. When there is an increased level of CO2 in the body (hypercarbia), CO2 will bind with water (H2O) within the CSF to convert into carbonic acid (H2CO3 ) which in turn releases hydrogen ions (H+) that lower the pH. When there is a decreased level of CO2 in the body (hypocarbia), hydrogen ions will bind with bicarbonate ions (HCO3) and convert into carbonic acid and then convert into CO2. This equilibrium is maintained by the respiratory centre establishing a respiratory rate that maintains CO2 within the normal range of 35 – 45 mmHg and pH within the normal range of 7.35 – 7.45.

As the respiratory centre becomes aware of a decreasing pH from the central chemoreceptors, it will innervate the diaphragm causing the person to take a breath. In people with chronically high levels of CO2, such as COPD patients that are CO2 retainers, the normal breathing mechanics involving central chemoreceptors would result in an unwanted high respiratory rate. The body gradually starts to develop a progressive ventilatory desensitisation to CO2, as bicarbonate ions increase to bind with hydrogen ions in order to achieve a pH level that is within the normal range. This is known as metabolic compensation. As a result, the body switches over to what is known as the “hypoxic drive” and utilises the peripheral chemoreceptors found in the aortic arch and bifurcation of the carotid arteries instead. Peripheral chemoreceptors are sensitive to low oxygenation levels. As the respiratory centre becomes aware of a PaO2 falling below 60 mmHg which approximately correlates with an oxygen saturation of 90%, it will innervate the the diaphragm in order to make the person take a breath.

A person breathing using the oxygen generated by a tree inside breathing apparatus

The Hypoxic Drive Theory

So the hypoxic drive theory is as follows: if you give a CO2  retainer too much oxygen, their PaO2 will increase and you will knock out their hypoxic drive to breathe therefore causing apnoea. However, their central chemoreceptors are still working enough to signal the brain to breathe to bring their CO2 levels back into a range that is normal for them. Therefore, the hypoxic drive theory is debunked! And even if you are unconvinced that this is the case, the hypoxic drive only starts to theoretically be affected when the patient’s oxygen saturations are more than 90%. So if your COPD patient has low oxygen saturations, put oxygen on them until they have an oxygen saturation at least 90%. There have been many a MET call that I have arrived to with a COPD patient that has oxygen saturations of 75% with 2 L/min of oxygen via nasal prongs because “they are a CO2 retainer and I didn’t want to give them too much oxygen”. Want to know the first thing I do when I arrive? I put on a hudson mask on this COPD patient with oxygen saturations of 75% and I turn the oxygen flow up, up and up. If there is only one thing that you remember after reading this article, let it be this:

Your patient will die of HYPOXIA before they die of APNOEA!

The current evidence for managing an acute COPD exacerbation is to titrate oxygen therapy to an oxygen saturation between 88 – 92%. This is most likely due to our long time belief in the hypoxic drive and a small study of 405 patients in 2010 that demonstrated a lower mortality rate in patients titrating oxygen to oxygen saturations of 88 – 92% compared to the patients receiving high flow non-titrated oxygen. This has further cemented the view that giving oxygen to a COPD patient is bad. However, giving high flow non-titrated oxygen to ANY patient is bad! The increased mortality rate with the high flow non-titrated oxygen COPD group could therefore be attributed to the deleterious effects observed with oxygen toxicity. Oxygen toxicity has been associated with various clinical consequences including diminished lung volumes, hypoxemia due to absorptive atelectasis, accentuation of hypercapnia, and damage to airways and pulmonary parenchyma. Until we have further evidence that may demonstrate that higher saturation levels in the COPD patient are non-detrimental, aim for 88 – 92%.

Bottom line: if your patient is hypoxic, turn up the oxygen! Get them to this acceptable range of 88 – 92% and then wean back your oxygen to the lowest possible flow rate to maintain this range.  The only time you are going to see a CO2 retainer lose their drive to breath due to an increase in oxygen is when you have someone who is struggling to breathe, is worn out, and has fatigued their respiratory muscles to the point where they have no further capacity to blow off the excess CO2. But guess what? This is called type 2 respiratory failure and a “normal” patient will also have a similar outcome in the same situation!


Consider the scenario of a COPD patient that arrived to the emergency department with a respiratory rate of 45 bpm and oxygen saturations of 70% on room air. An arterial blood gas was done and showed a PaO2 of 44 mmHg, PaCO2 of 66 mmHg and pH of 7.35. The nurse on the shift put 10 L/min of oxygen via a face mask on the patient and after a few minutes, the patient’s respiratory rate was 26 bpm with a much less laboured respiratory effort. Another arterial blood gas revealed a PaO2 of 90 mmHg, PaCO2 of 80 mmHg and pH of 7.25. Did we knock out the patient’s drive to breathe? Or did we relieve the hypoxia sufficiently that it allowed the patient to return to a normal breathing pattern for them? But what about the rise in CO2? This is actually attributed to a ventilation/perfusion (V/Q) mismatch and something known as the Haldane effect.

V/Q Mismatch in the COPD Patient

The body is very clever at trying to keep us alive! So in the chronic COPD patient with some alveoli that may not ventilate as well, the body overcomes this through hypoxic vasoconstriction of the pulmonary vasculature to direct blood to the alveoli that are functioning well. Therefore, the V/Q match is enhanced and gas exchange is optimised. Giving these patients oxygen overcomes the hypoxic vasoconstriction observed, and there is an increased blood flow to the poorly ventilated alveoli that results in a V/Q mismatch for the patient. As a result, gas exchange is compromised affecting both PaO2 and carbon dioxide.VQ mismatch in asthma due to hyperoxia

The Haldane Effect in the COPD Patient

Deoxygenated haemoglobin tends to hold on to carbon dioxide with a greater affinity than oxygenated haemoglobin. So when our patient arrived in the emergency department with oxygen saturations of 70% and a PaCO2 of 66 mmHg, the deoxygenated haemoglobin could have been holding on to a fair amount of carbon dioxide that wouldn’t show up on the arterial blood gas. The Haldane effect occurs when the oxygenated haemoglobin release the carbon dioxide that they were holding in their deoxygenated state, causing a rise in the PaCO2. If the patient has not fatigued their respiratory muscles at this point, hypoxic drive or not, their central chemoreceptors will sense the carbon dioxide level rise outside the normal parameters for the patient, and will increase their respiratory rate to clear the additional carbon dioxide. If the Haldane effect causes the carbon dioxide to be persistently high with a low pH, we need to be on the look out for impending type 2 respiratory failure and have our non-invasive or intubation equipment ready. 


What Does That Mean For Your Practice?

It is possible to survive a severe respiratory acidosis with no adverse outcomes provided that oxygenation and circulation are maintained. The issue with a patient that is not breathing, is that they are not oxygenating. Remember: your patient will die of HYPOXIA before they die of APNOEA! So when in doubt, oxygen is the good guy! But as with everything in life, all in moderation! Let’s not leave our poor COPD patient’s in a hypoxic state any longer; the hypoxic theory has been debunked!


If you liked this post, why not subscribe to Blogging for your Noggin? Be one of the first to know when new content becomes available! I promise there will be no annoying spam! 

Enter your email address to follow this blog and receive notifications of new posts by email.


16 thoughts on “The Hypoxic Drive Theory…DEBUNKED!

  1. I am a COPD patient who almost died when hospitalized for aspiration of a foreign object. Nurses kept turning down my 02, quoting the Hypoxic Drive Theory, which I insisted has been debunked. My husband physically pushed them out of the way when they tried to turn down to 2L and I could not breathe. Too much 02, you know. Nurses insisted that they had been taught the Hypoxic Drive Theory and therefore it was correct. The RRT and MD were just as uninformed. I increased the 02 to where my sats were at least 94%, and that was 6L at that time. After I finally coughed up the object after 10 days, I could decrease the 02 back to 2L. Ignorance kills. My husband saved my life..he would not leave my side.


  2. Thanks for the blog.

    I like your style of explaining the oxygen/carbondioxide issue without making it too complicated. Keep up the good work. I’ll be looking at other articles for tips on how to communicate other topics in a better way. Of the three possible mechanisms for hypercapnia/hypercarbia, the reason why V/Q mismatch and Haldane effect are rarely mentioned is that they are harder to explain than hypoxic drive. I am not sure how well I understand the first two.

    I work in Respiratory Medicine and I have changed my message to focus on “Target Oxygen Saturations” for the patient in front of you in their current state of disease and less focus on oxygen and carbon dioxide. I use the haemoglobin threshold for blood transfusion as an analogy. It could be 70 g/L or 90 g/L depending on the circumstances and the patient.

    Lawrence, Belfast

    Liked by 1 person

    • Hi Lawrence,

      I absolutely agree with VQ mismatch and Haldane being concepts that are harder to explain, and I feel we do that to ourselves because we are more comfortable with explaining the hypoxic drive.

      I think it’s a really important that clinicians have an understanding of the hypoxic drive in a normal person, then understand the physiology being swapping to a hypoxic drive in a person that is a CO2 retainer, then understanding that our normal mechanism of breathing to maintain a pH within a normal range doesn’t actually change between two! If this is clear in the mind of the clinician, there would be a lot more empowered clinicians out there!

      Oxygen, as with everything in life…is all about moderation!


  3. The study out of Tasmania alleging that untitrated O2 is worse than “titrated” had an endpoint of mortality. The ambulance rides are long and no ABG data reported. My best guess is that (like the Plante study using NIV utilization) that a PaCO2 threshold was the “guiding light” to intubation—with the resultant bad outcome after “the discussion”. Colonel Mustard in the Kitchen with a laryngoscope.

    Some other points–O2 toxicity was discussed in an article in Clinical Pulmonary Medicine many years ago. It appears that mammalian evolution confers on us an advantage over rats as far as O2 tolerance goes (also the pressures used may have been the culprit). The Sasquatch of the ICU. The article looked at studies of people on FiO2’s of over 0.60 and no untoward results. As far as MI’s, yes IF there is reperfusion after ischemia what basically happens is that the cells don’t really die after 6 minutes of no O2, but rather revert to “it’s every cell for themselves” and go anaerobic. So then a full onslaught of O2 is toxic (Hemoglobin originally was to scavenge O2 back in the evolutionary day before O2 used for energy). So instead you want to flush. Perhaps in the future we’ll be giving lower than 21% O2 to some!

    But what to do for a patient with an ischemic bowel and carbon monoxide poisoning then?

    Kudo’s to this author! But you’re really arguing religion. Or saying that the Emperor has no clothes and is butt ugly to boot. So many say “but I’ve seen it happen” as they retrofit a theory to explain ABG’s and/or clinical events. Much like a primitive culture with an unscientific accounting for thunder and lightening which always allows them to “see it happen”.

    Liked by 1 person

    • Hi Jerry,

      Thank you for the link to this recently published article; it was an enjoyable read. Similarly, the author of that article has attributed the clinical picture of hypercapnia in a COAD patient receiving oxygen to a V/Q mismatch secondary to hypoxic vasoconstriction and the Haldane effect.

      As the author of that article noted, the current literature advocates an oxygen saturation of 88 – 92% in COAD patients. This is based on a study comparing titrated oxygen saturations of 88 – 92% to non-titrated high flow oxygen. Important to note that the non-titrated high flow oxygen group did not have saturations disclosed in the article. The study goes in with a preconceived notion that high flow oxygen in COAD patients is bad. However, high flow oxygen therapy in ANY patient is bad. Take for example the recent AVOID study demonstrated a greater myocardial infarct size giving supplemental oxygen to a STEMI patient with normal oxygen saturations, compared to those titrated with oxygen to achieve normal oxygen saturations. The rationale for this was due to the free radicals that circulate in hyperoxic states; could this same rationale not be applied in the COAD patient? There are numerous articles out there associating oxygen toxicity with various clinical consequences including diminished lung volumes, hypoxemia due to absorptive atelectasis, accentuation of hypercapnia, and damage to airways and pulmonary parenchyma.

      Nonetheless, the current evidence is indicating we maintain oxygen saturations between 88 – 92%. Until we have more research into this matter, we will continue to do this. How quickly we get up to that range is each person’s prerogative. Personally, I want to make my hypoxic patient no longer hypoxic in the shortest space of time and would achieve this by turning the oxygen flow rate up. Once in the desired range, I would wean the oxygen to the lowest flow rate to maintain that range…as I would with any other patient cohort.



  4. Not so fast O2 and hypoxic drive

    On Wed, Jul 6, 2016 at 4:48 AM, Blogging For Your Noggin wrote:

    > Joanne Reading posted: “We’ve all heard it at some point or another: > “Don’t give that COAD patient too much oxygen”. It has led us to believe > that oxygen is a REALLY bad thing in patients with chronic obstructive > airways disease (COAD). This has caused nurses to be afraid to put” >


  5. Excellent! Very helpful. Thank you for responding to the request so quickly. Julie Mathews is absolutely right, you have a knack for explaining the essence of the issues in a succinct way that stimulates understanding and interest in the topic.
    Leslie Eddowes

    Liked by 1 person

Do you have a question? Or a suggestion for a new topic?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s