During sign out rounds, the resident reports on the next patient as you pass her room.
“75 y/o COPD patient, admitted, waiting for a bed, CXR negative for infiltrate. When her husband brought her in, she looked pretty bad with sats only 82% on her 2L oxygen from home. She was really working to breathe, but now looks so much better!” reports the resident.
You glance in the room. Her husband is reading the paper, and she is sleeping. The TV is on, and all appears calm in the room.
A quick glance and you note her current vital signs on the monitor:
Blood pressure: 138/88, P=108, Oxygen Saturation=99%
“When was her last breathing treatment?” You ask the resident.
“About 2 hours ago, but she doesn’t look like she needs one now.”
1. Why should you worry?
Patients on chronic home oxygen should not have a resting oxygen saturation of 99%. Clearly they are on more oxygen than usual (or they would never get their sats up to 99%), which can improve oxygenation, but sometimes at the expense of ventilation.
When these patients first arrive, and have clear air hunger, improving oxygenation is important. Many of these patients may arrive with oxygen saturation levels below 87%. However, it’s also important to realize their normal saturation is not the same as a person without COPD. If they are on home oxygen, their normal saturation may be 88-93% on oxygen. Many patients on chronic home oxygen will be able to tell you what their normal oxygen saturation is (on oxygen therapy). I usually place them on high flow oxygen (if I don’t know their saturation) as I’m doing my initial evaluation and starting treatment (Beta agonists, steroids, Bi-pap). By the time they get hooked up to all the monitors, I can see where the oxygen saturation is. I shoot for a target of their baseline saturation. If they don’t know what their baseline oxygen saturation is, I simply shoot for 93%.
If they come in on high flow oxygen with an elevated saturation (98%), I’ll turn down the oxygen to titrate to a saturation target of 93%.
There are several theories regarding why individuals with COPD will become acidotic and hypercarbic when given too much oxygen. Two common theories include the risk of respiratory depression from depression of hypoxic drive and ventilation/perfusion mismatch. (1) The first theory relates to the concern that patients with COPD don’t have the normal respiratory drive to breath related to increasing carbon dioxide (CO2) (as people with normal lungs do), but rather rely on decreasing oxygen levels to stimulate breathing. (1). In COPD patients increasing the oxygen level results in decreasing the respiratory drive because of hypoventilation, with subsequent hypercarbia and acidosis. In the second theory, the hypoxemia that is normally present in COPD patients helps to maximize matching ventilation and perfusion match. Increasing the oxygen level eliminates this and the resulting V/Q mismatch results in hypercarbia and acidosis. (1).
So, when physicians wanting to correct the hypoxemia provide too much oxygen, the COPD patient can begin to hyperventilate (breath less) because the hypoxic drive to breath is not there anymore. Unfortunately, the natural nurse or physician response to low oxygen saturation is to turn up the oxygen. Then the COPD patient continues to have a decreased respiratory rate, and eventually the Carbon Dioxide level can rise to a critical level, frequently referred to “CO2 narcosis.”
Patients with CO2 narcosis appear very comfortable and frequently look like they are sleeping. However, when you try and wake them up, you will realize something is wrong. Once you see it, you will always remember it. The best way I can describe it is that they appear to be drugged and trying to fight through the ‘fog’ to answer your questions. They will struggle to open their eyes, and when not stimulated, fall back into a ‘sleeping’ appearance. Unfortunately, although they may look OK, they are not. As the ventilation decreases, the patient becomes more acidotic.
A recent systematic review of patients with acute exacerbations of COPD found that those who received high flow oxygen from EMS during transport found that patients transported with high flow oxygen had a higher mortality. (2) So if they arrive with high flow oxygen, consider turning it down until you know where you are with regards to oxygen saturation. The NICE guidelines suggest hospital oxygen therapy to keep saturation above 90% (3). The newer NICE review (Nov 2008) recommends EMS not exceed oxygen saturation of 93% during transport. (4)
Exam of the lungs usually reveals continued bronchospasm, which should be a red flag, if they still have significant bronchospasm, and then they should be breathing hard with retractions and tachypnea, and not look comfortable with a normal respiratory rate.
So, there is something very wrong with this picture. You must make sure the patient is really sleeping because she is improved, and not because she is severely hypercarbic and acidotic.
2. What can you do now?
If you do suspect CO2 Narcosis and the patient has an abnormal mental status, you will need to do several things:
1. Check a blood gas. You can quickly determine where you are in the acid base status of the patient. I usually get a venous gas because the nurse can quickly draw and send it, but many pulmonologists at our institution prefer an arterial gas. A recent Best Bets review found that a capillary blood sample accurately reflects arterial blood gas measures of PO2, PCO2 and pH. (5) In asthmatics, venous blood gases also correlates very well with arterial levels of pH and pCO2. (6)
However, do not wait for the blood gas to initiate treatment.
2. Check an accucheck, likely overkill, but easy to do to rule this out as the reason for the altered mental status.
3. Turn down the oxygen to increase the respiratory drive. I shoot for 91-93% sat.
4. Treat the bronchospasm – I usually give another breathing treatment.
5. Noninvasive positive-pressure ventilation (NIPPV) works great to help ‘blow’ off the CO2, increase the respiratory rate, and decrease the work of breathing. (1) If the patient’s mental status is significantly decreased, this may not be an option. In that case, you may need to intubate.
3. The transporter is here to take her to the floor, what do you tell him?
The patient cannot be transported until you confirm the status is stable. If you find she was really sleeping and feels fine and sounds improved, then she can go (although remember to turn down the oxygen so it doesn’t become a problem on the floor).
If she is hypercarbic, then treat her. Depending on her status, if her condition is severe, she may need to go to the ICU.
Blood gas revealed pH of 7.18, pCO2=95. She was arousable, but very tired. She was awake enough to try bi-pap, and we turned her oxygen down to maintain sat at 90%. She did improve, and repeat gas 1 hour later. It showed the pH up to 7.25 and pCO2 improved at 80. She continued to wake up, and the ICU team did not feel she needed to go to the ICU.
1. Kamangar N, Nikhanj NS, Sharma S. Chronic Obstructive Pulmonary Disease. Emedicine. Feb 19, 2010. Accessed at: eMedicine Medscape
2. Huq S., Sloan J. High flow oxygen in acute exacerbation of COPD. Best Bets. September 2005. Access at: Best Bets
3. Chronic Obstructive Pulmonary Disease: Quick Reference Guide. NICE. Feb. 2004. Access at:
4. NICE. COPD Management NICE guidelines (draft for consultation). November 2009. Access at:
5. Murphy R, Harrison M. Capillary blood gases are comparable to arterial gases in COPD. Best BETs Oct 2000. Accessed at: Best BETs
6. Hassan Z., Subramonyam D. Venous blood gas in adult patients with diabetic ketoacidosis. Best BETs. May 2002. Access at: http://www.bestbets.org/bets/bet.php?id=472