I killed my COPD'r with an oxygen bottle!

[jrm818]

Have you killed a patient this year? Some here (maybe me!) might have...

I know that it's already common practice for some providers to try to titrate COPD'rs back to their normal (lowish) SpO2 values, but my impression is that the general rule is for EMS providers to aim pretty high with saturation values.

Similarly, there have been numerous discussions here about "high flow"/high concentration O2 and COPD, and these discussions generally end with a consensus that the "O2 kills because of hypoxic drive" theory is bunk, which seems to me to be accurate. However, nearly every discussion has one or more posters suggest (unchallenged) that "for the short time that EMS treats a COPD patient" high O2 concentrations are unlikely to do any harm.

I would have agreed with that a week ago. I don't like the "O2 for everyone mentality, and think its a poor excuse for "medicine," but these results still came as a bit of a shock to me:

Free full text access here, apparnetly the BMJ thought it was important enough to let all read:

http://www.bmj.com/content/341/bmj.c5462.abstract

I encourage everyone to read the full text. I don't think the abstract can tell you enough with this study

Austin et. al: Effect of high flow oxygen on mortality in chronic obstructive pulmonary disease patients in prehospital setting: randomised controlled trial

405 patients with "presumed acute exacerbations of chronic obstructive pulmonary disease" were treated with either bronchodialators nebulized by compressed O2 plus O2 via NRB at 8-10 L/min or treated with bronchodialators nebulized by compressed air plus oxygen delivered by nasal prongs and titrated to achieve oxygen saturations between 88% and 92%.

The shocking part (to me) degree of harm found (just from prehospital care!): "Overall mortality was 9% (21 deaths) in the high flow oxygen arm compared with 4% (7 deaths) in the titrated oxygen arm; mortality in the subgroup with confirmed chronic obstructive pulmonary disease was 9% (11 deaths) in the high flow arm compared with 2% (2 deaths) in the titrated oxygen arm. Titrated oxygen treatment reduced mortality compared with high flow oxygen by 58% for all patients (relative risk 0.42, 95% confidence interval 0.20 to 0.89; P=0.02) and by 78% for the patients with confirmed chronic obstructive pulmonary disease (0.22, 0.05 to 0.91; P=0.04)."

"The number needed to harm was 14; that is, for every 14 patients who are given high flow oxygen, one will die."​

The study does of course have limitations. It is a single center study, I'm not sure how I feel about the choice to randomize paramedics instead of patients, and there was a very high rate of non-compliance in study participating (although, as the authors note, that may minimize rather than inflate the risk of harm from oxygen administration), but I still think this is incredibly important.

Comments? Does this make you re-evaluate the way you practice? Are you surprised that EMS could have such a profound effect on patient course, even with short treatment time? What do you think about the quality of the study? Is this enough evidence to modify your behavior?

It was notable to me that every death occurred after arrival at the hospital, and only 2 occurred in the emergency department. I wonder if this says anything about the need for EMS to examine treatment effects that are not immediatly apparent, and may take days to present? My sense is that there is minimal appreciation in general as to the long term effects of the things we do in the field.

I know that I was shocked the first time I saw pictures of grossly edematous trauma patients a day or so after massive fluid infusions....EMS education doesn't seem to concerned with educating providers about such delayed effects. What do you do to attempt to keep in tune with the effects (immediate and delayed) of your treatments? Do you follow up with patients long-term course?

 

[reaper]

While COPD'ers fall in the same line as every other Pt. Assess and treat as needed. These Pts do not need high SPO2. They need to ventilate.

That study in no way proved a higher mortality rate from O2. Those Pts may have died from other complications.

Assess your Pt. Do not slap a NRB on every SOB and do not withhold from a COPD Pt from fear. Assess and treat. If they need O2, give it to them. Lack of ventilation will kill faster.

 

[EMSLaw]

Generally, the risk from Oxygen administration to a COPDer is characterized in terms of the hypoxic drive. Since only a small number of those with COPD have a hypoxic drive, and since the chances of causing respiratory arrest even in them is slight, EMS personnel are taught that there is no danger in giving oxygen to even COPD patients.

I'm curious as to how the study deals with comorbidities in COPD patients, who are, let us be honest, quite sick to begin with. I'm concerned that the corellation here does not imply causation.

If the patient is having respiratory issues, then the immediate danger of hypoxia is the concern. If, on the other hand, the patient is in no apparent distress (they have some non-respiratory issue, for example), then there's no reason to throw them on high-flow O2 just go get their oxygen saturation up. Like everything else in the ambulance, the PulseOx is a tool, and we shouldn't blindly be lead by the numbers. Patients that need oxygen should get oxygen, and those that don't probably shouldn't.

Which is a long-winded way of saying I agree with Reaper that you treat the patient in front of you.

 

[EMTStudentwa]

First I gotta say its really neat to come to this forum and read about things I'm learing in my EMT class. I missed a question on a quiz because I thought that hypoxic drive was something to worry about. The quiz answer and response from my instructor, was that you never withhold o2 from a pt. He said sure, if they get sleepy or sluggish, turn down the o2 but its going to do more good than harm. And as it was said, these are already very sick patients, and you gotta do as you're trained.

 

[Veneficus]

Like many things in EMS, there is an exageration on the extreme patiens.

The pts who are have profound hypoxic drive are not usually encontered by EMS because they are already managed in a pulmonary ward or SNF.

This is an end stage that presents in the "blue bloater" subcategory.

It is not that these patients don't exist, it is that EMS has about as much chance of encountering one as encountering a pt that is helped by the effect of MAST occluding a ruptured abd aortic aneurism.

Even if they did encounter one, they might not even recognize it as any more specificthat an impending or actual respiratory arrest.

Not because EMS providers are at fault, but because the education revolves around "life saving" interventions for peri/arrest.

As a perfect example, how many surgical trauma patients do you see a year? From EMT/Medic class, you would think there was a need for multisystem surgical intervention on every patient who falls off the bed as a mechanism.

 

[EMSLaw]

As a perfect example, how many surgical trauma patients do you see a year? From EMT/Medic class, you would think there was a need for multisystem surgical intervention on every patient who falls off the bed as a mechanism.

Do you mean how many that live long enough to actually make it to a surgical suite?

 

[jrm818]

While COPD'ers fall in the same line as every other Pt. Assess and treat as needed. These Pts do not need high SPO2. They need to ventilate.

That study in no way proved a higher mortality rate from O2. Those Pts may have died from other complications.

Assess your Pt. Do not slap a NRB on every SOB and do not withhold from a COPD Pt from fear. Assess and treat. If they need O2, give it to them. Lack of ventilation will kill faster.

I don't think the issue here is so much that you should be afraid to give high concentrations of oxygen, but that as a provider you should be targeting oxygen therapy to reach a specific goal, and in the case of COPD patients that goal should be lower than in other patients - around 90% SpO2 (if the study is to be believed). If that takes high concentrations than OK, but the evidence here is that if high concentrations of O2 are more than what is needed, than maybe you SHOULD be afraid of giving too much.

This really agrees with the "Treat the patient" sentiment you and EMSlaw express....but the problem is that in EMS many providers (perhaps not any individual here, but I'd say most) think "treat the COPD patient" means "high concentration of O2," which may do harm in patients who do not require high concentrations (but might be OK if that's what's needed to get them to 90ish).



I would be interested in an elaboration on the bolded part of your statement - I don't really understand it. It is of course possible that patients died from "other complications," but in that case this evidence would suggest that there may be a higher rate of complications due to high O2 administration prehospital. The patients in each treatment arm appeared reasonably well matched, received generally the same care, etc.

It is possible that there was a random variation such that the high flow O2 group was sicker or had more co-morbidities, and thus died at a higher rate, but there is no evidence I see to support this hypothesis. However, that's the point of having large enough sample sizes to claim statistical power - to minimize (ie make statistically insignifigant) the possibility that a difference in treatment effects can be accounted for by random variation.

I'm no statistics expert, but barring a problem with the statistical analysis, the study appeared to use a sufficiently large and well matched population to be able to say that the difference in outcome between the two study arms is statistically unlikely to arise from any variation other than the variation in oxygen administration in the prehospital treatment phase.

Basically: The study seems to me adequately designed to detect a difference in morbidity due to two different techniques of oxygen administration, and I'd be interested in hearing any reason you think that is not so.



As for the hypoxic drive stuff:

Really, I don't think that is an issue here at all. I mentioned the theory only because most discussion about COPD'rs and oxygen involves the "myth," and I thought this study raised a different possible cause for harm due to oxygen administration. I agree that the hypoxic drive boogieman should go away and that it is not really a concern in the prehospital world especially.

What is a concern raised by this study is the possibility that relative hyperoxia may increase CO2 levels in these patient and induce respiratory acidosis. This is largely a different issue.

 

[CAOX3]

A doctor said it best to me, its about where they live, if they live at 88% then why are we striving for 100%. Titrate to comfort and stability. Your not going to reverse twenty years of lung damage on a ten minute trip.

If their comfortable at 88% then I'm happy, if it takes 10 liters or two.

 

[MrBrown]

http://www.sciencedaily.com/releases/2010/10/101018211824.htm

 

[mycrofft]

Good points above.

Also:
1. Were pt's getting untitrated oxygen because they were crashing (hence more prone to die) and the titrated ones were more stable? (Co-morbidities).
2. Length of oxygen application?
3. Small retrospective study. How about comparative pts who did not receive oxygen at all?
4. Official causes of death? How far post-induction of oxygen treatment were pts followed? (I.E., were they tracking out to one week after the point the pt was picked up, or two weeks, or what? At one week maybe a bigger difference of outcome than at two weeks). What other factors came into play in that period between pickup and the cutoff point?
5. In a study this small, percentage differences might not really be significant except to suggest further larger, longer and better-controlled studies. Only four pts would constitute nearly 1% of this total sample.

Yes, treat the best for each pt, don't slap on O2 to feel useful, but buy enough time to do the fine-tuned stuff.

 

[jrm818]

These objections/questions are really all answered in the full text of the study

Also:
1. Were pt's getting untitrated oxygen because they were crashing (hence more prone to die) and the titrated ones were more stable? (Co-morbidities).

No. This is a prospective randomized study. The paramedics in the system were randomized to one of the two treatment arms and treated every presumptive COPD patient they saw according to their random assignment. Paramedics and patients were all reasonably matched.

2. Length of oxygen application?

treated in the prehospital phase for an average of 45 minutes. The authors tried to get the hospital to continue the same treatment, but with varying compliance

3. Small retrospective study. How about comparative pts who did not receive oxygen at all?

A: No. 400 patient randomized prospective study. B: Placebo control here would be rather unethical...active control is really the only option, and gave much more interesting results really. I'm sure no oxygen is not a good thing....but the questions "is titrated oxygen better than blind high concentration oxygen" is clinically useful.

4. Official causes of death?

"Respiratory failure was the primary cause of death in all cases [...]"

How far post-induction of oxygen treatment were pts followed? (I.E., were they tracking out to one week after the point the pt was picked up, or two weeks, or what? At one week maybe a bigger difference of outcome than at two weeks).

Followed to death or discharge. 70% of all deaths occured within 7 days. The remarkable thing about the study is that it demonstrated that O2 administration technique can have an effect on long-term morbidity and mortality.

What other factors came into play in that period between pickup and the cutoff point?

Probably quite a few. The issue is that these "other factors" would occur with equal frequency in both treatment groups.

5. In a study this small, percentage differences might not really be significant except to suggest further larger, longer and better-controlled studies. Only four pts would constitute nearly 1% of this total sample.

Again, I'm no statistics expert, but the sample size was calculated to detect this difference with statistical power. they did achieve statistical significance. Also...4 is a percentage...21 actual patients died in the control group (9%) and 7 in the active group (4%)

Yes, treat the best for each pt, don't slap on O2 to feel useful, but buy enough time to do the fine-tuned stuff.

Yes, do the best for each patient...but the only way you know what is best is to have a clear idea of the effects (acute and long term) of your treatment.

 

[jrm818]

A doctor said it best to me, its about where they live, if they live at 88% then why are we striving for 100%. Titrate to comfort and stability. Your not going to reverse twenty years of lung damage on a ten minute trip.

If their comfortable at 88% then I'm happy, if it takes 10 liters or two.

Agreed, except that even if they are uncomfortable at 88, it may be time to consider the possibility that more oxygen A) won't help, their problem is probalby ventilation now and B) may kill them at a significantly higher rate than holding back on oxygen would.

Maybe this is a time that you need to use the tools you have and actually do treat the monitor to an extent? Yes, treat the patient in front of you, but if the monitor says they have enough oxygen, than "treat the patient" might not have to include more O2, even if dyspnea persists.

 

[CAOX3]

Agreed, except that even if they are uncomfortable at 88, it may be time to consider the possibility that more oxygen A) won't help, their problem is probalby ventilation now and B) may kill them at a significantly higher rate than holding back on oxygen would.

Maybe this is a time that you need to use the tools you have and actually do treat the monitor to an extent? Yes, treat the patient in front of you, but if the monitor says they have enough oxygen, than "treat the patient" might not have to include more O2, even if dyspnea persists.

Are we comparing death rates from hypercarbia and hypoxia?

Hypoxia is going to kill more pts then hypercarbia will.

When a COPD pt quits breathing, in my experience its usually because the provider failed to recognise or address (with medications or PPV) how sick they truly were.

 

[Veneficus]

Do you mean how many that live long enough to actually make it to a surgical suite?

I mean how many actually require emergency surgical interventionby a critical care or trauma surgeon.

The numbers are rather low and decreasing nation wide in the Level I centers, so I am guessing that based on the fact they get patients from mulitple sources and EMS only gets 1 source, that the amount of providers working prehospitally see an abysmally small number of surgical trauma patients.

So much so, that the EMS curriculums are improperly geared towards such extremes.

I would concede ortho injuries are technically "surgical" in nature, but few still go straight from ED to surg and almost none are ICU admits.

 

[usalsfyre]

So we're back to defending outdated EMS practice in the face of evidence that it might be harmful.

This is a small study, and needs to be quantified by further research. However I have a strong inkling the results will be repeated in a larger sample population. EMS needs to start thinking like clinicians who understand the effects of treatments they initiate will have long-ranging consequences.

We know, without a doubt, hypoxia will kill. What's the point of a comparative other than to make ourselves feel better about high-flow, untitrated O2?

 

[usalsfyre]

Are we comparing death rates from hypercarbia and hypoxia?

Hypoxia is going to kill more pts then hypercarbia will.

When a COPD pt quits breathing, in my experience its usually because the provider failed to recognise or address (with medications or PPV) how sick they truly were.

Does it matter? What we're comparing is death rates in people who were initially in respiratory distress. One group received high-concentration oxygen, the other received titrated oxygen PRN. The group who got "balls to the wall" O2 died at a higher rate. Other complications should have occurred at the same rate between the two sample groups.

This is not the old "hypoxic drive" argument. This is a recognition that oxygen is NOT a benign treatment as has been taught for years. We're just now starting to figure out how harmful O2 therapy can be. O2 is indeed a medication, and should be treated as such. "More is better" is not the philosophy used with pressors, why is it with O2?

 

[Veneficus]

We know, without a doubt, hypoxia will kill. What's the point of a comparative other than to make ourselves feel better about high-flow, untitrated O2?

This is a great way to make this point.

But I think there is some confusion in the EMS world between hypoxia and anoxia.

The terms are not the same.

 

[Veneficus]

"treated in the prehospital phase for an average of 45 minutes. The authors tried to get the hospital to continue the same treatment, but with varying compliance"

In my very first year of medical school, what seems like an eternity ago, we were specifically taught about oxygen and its uses. A point was made that research had shown for years it was not benign and "more was not better."

The topic was frther expanded upon with the handful of specific conditions where high flow o2 actually helps.

Furthermore, anyone with a basic understanding of biochemistry and characteristic reactions of organic chemistry, could not possibly come to the conclusion that high flow o2 should be or ever have been the standard.

But as is readily apparent, EMS is the only part of medicine that puts great effort into holding on to practices rather than furthering them.

In the words of a very wise person here,

"five is four"

 

[jrm818]

Furthermore, anyone with a basic understanding of biochemistry and characteristic reactions of organic chemistry, could not possibly come to the conclusion that high flow o2 should be or ever have been the standard.

I wonder if this trial would even have been completed if the high flow mantra never existed (a bit of a hypothetical situation here....) and the study had been run the other way around..ie with titrated O2 as the "standard" and high flow as the experimental arm.

In this land of make-believe I bet the authors would have terminated the trial early on ethical grounds once such a significant level of harm started appearing up in the "experimental high flow" group.

 

[Veneficus]

I wonder if this trial would even have been completed if the high flow mantra never existed (a bit of a hypothetical situation here....) and the study had been run the other way around..ie with titrated O2 as the "standard" and high flow as the experimental arm..

Recall that when the initial theory and protocols for high flow o2 in EMS were written, molecular biology was still an obscure form of "pseudoscience."

More has been learned in what is now the last 15 years in medicine than in the history of medicine previously. Prior to the last 20 years, even the physicians didn't know any better.

With the exception of some tweaks in CPR and cardiac resuscitation, much of what EMS does was written in the early 80s or even 70's. You will not find a hospital in the developed world practicing that kind of medicine nor any reputable medical school (one where you cannot by your diploma first day) teaching medicine that old.

In this land of make-believe I bet the authors would have terminated the trial early on ethical grounds once such a significant level of harm started appearing up in the "experimental high flow" group.

Perhaps, if the trial even got the "go ahead" to begin with. It would be like starting a trial comparing the outcome of blood letting wounds with techniques in wound closure.