Why Rescue Breaths May Matter in Overdose CPR

This discussion was recorded on July 18, 2025. This transcript has been edited for clarity. 

Robert D. Glatter, MD: Hi, and welcome. I’m Dr Robert Glatter, medical advisor for Medscape Emergency Medicine. Joining me today to discuss an article published last month in JAMA Network Open, “Bystander CPR Technique and Outcomes for Cardiac Arrest With and Without Opioid Toxicity,” is Dr Sheldon Cheskes, medical director at Sunnybrook Centre for Prehospital Medicine in Toronto, Ontario. 

Also joining me is Dr Peter Antevy, a pediatric emergency physician medical director for Coral Springs-Parkland Fire Department in Palm Beach County, Florida. Welcome, Peter and Sheldon. It’s great to have you join me. 

Sheldon Cheskes, MD: Great to be here. 

Peter M. Antevy, MD: Good to be here, Rob. Thanks for having us. 

The Study: Better Outcomes With Rescue Breaths in Overdose Cases

Glatter: I want to discuss this study by Brian Grunau’s group in Vancouver, which looked at more than 10,000 out-of-hospital cardiac arrest cases with linkages to many provincial data sources to identify cases due to opioid toxicity. We’re well aware of the clinical scenario where a patient is found down, pulseless, and apneic. 

Assuming no apparent trauma, the guidelines have encouraged bystander compression-only CPR; this has been a beneficial approach and many persons now assist in such a scenario. The fear of bystanders attempting what we’ll call mouth-to-mouth, colloquially — risking a bodily fluid exposure — has really dissuaded the public in the past to get involved.

In the setting of an opioid overdose, would ventilations be helpful to improve survival in out-of-hospital cardiac arrest? Intuitively, we know the answer likely is yes, based on physiologic mechanisms. The reality is that the research here isn’t clear about the validity of having compressions with ventilations in this population of opioid toxicity and undifferentiated patients. 

Based on some sources, in about 10% of out-of-hospital cardiac arrests, there is some drug toxicity. Knowing that, it begs the question, would there be some value to having assisted ventilation in these scenarios?

Peter, can you describe the design and some of the outcome variables to our audience? 

Antevy: This is a great topic. Just to set the bar here so people know, every year there’s a large registry, called the Cardiac Arrest Registry to Enhance Survival (CARES) , in the United States. Last year, for example, about 135,000 cardiac arrests were actually worked on in the field.

Only one third of those patients received bystander CPR by the layperson. A majority of those happen because when the person calls 911, then the 911 call taker encourages them to do CPR. As many of us know, the reason that they went to compressions only is because they wanted to just increase the number of patients who actually got bystander CPR — and even that wasn’t able to do it.

We still are at a very big loss compared to, let’s say, the folks in the Netherlands, who have approximately a 65% or 66% bystander CPR rate. What you’re asking here is, for patients specifically with an opioid overdose, can we have the bystander provide ventilations with a mask or mouth-to-mouth? I think that’s going to be a difficult thing to do. However, it’s definitely an important study — to show that this is important. 

Just to talk about the entire design of the study, Brian Grunau is a great researcher out of Canada. It seems like all the Canadians are doing all the great things. As far as the actual design, I’m not an expert in this particular study, per se. It was a cohort study. They performed it from August 1, 2023, through December 31, 2024. 

They analyzed cases of adults with EMS-treated out-of-hospital cardiac arrests that occurred from December 1, 2014, all the way through March 31, 2020. These cases were basically classified as opioid associated based on a postmortem toxicologic investigation or death certificates, and then all the other cases were categorized as undifferentiated. They really had a good knowledge base here of which patients had opioids associated with their cardiac arrest and which patients did not.

With respect to the main outcomes here, they looked at favorable neurologic outcome and hospital discharge, basically looking at a CPC score of 1 or 2. Obviously, they did all the typical adjusted logistic-regression models and they looked at the outcomes then. They looked at 10,923 out-of-hospital cases. After removing 24 cases only treated with ventilatory support, there were 1343 opioid-associated cardiac arrests. Then the rest, about 9556, were in the undifferentiated group. 

When they looked at the favorable neurologic outcome on the adjusted odds ratio, it was 2.85. That is a significant increase when you are performing bystander CPR and you’re actually providing ventilations to the opioid group compared to the group that did not get that. It is significant and it would be nice if those patients could get that instruction when they call 911. 

Before I pass it over to Sheldon, I will say that in our system in the United States and in Canada, we use systems like APCO EMD or Medical Priority Dispatch System. When you call and you say, for example, that someone is in cardiac arrest or they’re not breathing, and you say the word “hanging” or “drowning” or “opioid,” it directs the call taker to recommend the bystander to give them breathing.

However, I could tell you, as an EMS medical director, that we rarely, if ever, see patients getting that from a bystander. I’ll stop there, but definitely I would agree that patients who have opioid-related cardiac arrest do need ventilations, from a physiologic perspective.

Should Narcan Be Routine in Cardiac Arrest Response?

Glatter: In that setting, would Narcan or naloxone have a role? In those scenarios, clearly, you’re going to obviously offer ventilatory assistance or intubation. Of the proportion of arrests where you come on scene and you’re just not sure, is there any downside to giving naloxone?

Antevy: This is a great topic. One of my favorite papers of last year was by Dr David Dillon in Northern California, and they looked at thousands of patients who received naloxone for their cardiac arrest. They found that if you had opioids on board, then you got significantly better than if you didn’t. However, they found, interestingly, that patients who were in cardiac arrest without any opioids in their system did better with Narcan. That begged the question of, should we be adding Narcan into the algorithm? Is there a downside?

I will say that the lay public has Narcan. It feels like it’s everywhere now. The police have Narcan. My only stipulation here is that if you’re an EMS professional and you’re getting to the scene, and I’m your medical director, I want you to do the basics first. If someone’s not breathing and they’re in cardiac arrest, the treatment is not naloxone. The treatment is high-quality basic life support (BLS) CPR with high-quality chest compressions, the appropriate ventilations, and then you should give Narcan.

In the future, I think Narcan will be added into the algorithm for every single cardiac arrest, irrespective of the cause. The reason is because physiologically the body produces endorphins. Think of an endogenous morphine. When you’re in a trauma, for example, your body releases that so you don’t feel pain, but it also relaxes the vessel that causes your diastolic pressure to drop. 

Narcan, it turns out, reverses even the endogenous morphine or opioid substances and therefore gives you a bump in your blood pressure, specifically your diastolic blood pressure, which actually gets you back to life.

I am very heavily leaning toward that direction, and I predict that my own protocols — I’ll finally get them in 2026 — will have Narcan in the algorithm. Just later, not earlier. 

The Public Health Challenge: More Complexity, Less Action?

Glatter: Sheldon, I’d love to get your take on what you’re doing in Canada. 

Cheskes: There’s lots of stuff to unpack there. Certainly, Brian Grunau’s group is outstanding. Whereas I’m a defibrillation guy, he has done a great job on the work. The linkage to autopsy or coroner data is a very strong methodologic mechanism to do that.

The use of Narcan or ventilations by the lay public is the big question here, right? As Peter has already mentioned, compression-only CPR, from the lay public point of view, is here to stay. There are two groups that are likely suffering the most from compression-only CPR: this group of opioid-associated cardiac arrests and the pediatric group. The pediatric group really requires ventilations, and that’s kind of been lost in this move to compression-only CPR. Again, the balance, as Peter has brought up, is to get enough people to actually do CPR.

Here in Canada, when I first started as a medical director, we were down at 20%-30% bystander CPR. We’re now, at CanROC [Canadian Resuscitation Outcomes Consortium], probably over 55%-60% across Canada for bystander CPR. We really made some great strides.

The big piece is to separate out the lay responder from the bystander responder or the trained responder. For the trained responder, all these things are things that should be done. For the bystander, it’s going to be very difficult to convince them in an opioid-associated cardiac arrest to actually do mouth-to-mouth or do ventilation. I think there’s going to be a challenge there. 

Looking Ahead: Scoring Tools and Future Research

Cheskes: I’m going to give a plug to Hania Siddiqui. She’s my PhD student, and her work is right on this. Can we develop a score, based on information available when we assess the patient, that says this is a high risk of being an opiate-associated cardiac arrest vs undifferentiated? It’s pretty similar to the work Brian did. 

I’m really excited for her and hopefully, in that work, we’ll be able to develop a score based on some of the work we did here in Toronto, some of the work Brian’s done and the San Francisco group has done, all in this particular area. I think we’re going to be able to identify these people better once we have the score in place.

The piece about Narcan as standard in cardiac risk management is a really interesting one. That’s a randomized trial question. Do we give Narcan in all undifferentiated cases or to people with unwitnessed cardiac arrest? That’s another group that also had this. I think that’s a great question for a randomized trial.

One of the weaknesses of these studies on the use of Narcan is that they’re very weak on the timing of Narcan. What I mean by that is you don’t know in those studies if they got it pre-arrest or if they got it during the arrest. I think that’s a big differentiating factor.

Certainly, from the point of view on the use of Narcan, once the patient has already arrested, they didn’t find a great deal of evidence there. I think this is because of so much weakness in knowing exactly when Narcan was given.

I know Peter is swinging the pendulum one way. I’m probably a little bit more to the middle, as a scientist, as we all are, right? 

Glatter: I want to thank both of you for such an enlightening conversation. Certainly, it’s pushing the boundaries and really making us think about new horizons and alternatives in the realm of patient care research, and how we can move the needle.

What to Know as a Bystander:Compression-only CPR is still the default recommendationIf you suspect overdose and feel comfortable, rescue breaths may improve survivalAlways call 911 and follow dispatcher instructions

Robert D. Glatter, MD, is an assistant professor of emergency medicine at Zucker School of Medicine at Hofstra/Northwell in Hempstead, New York. He is a medical advisor for Medscape and hosts theHot Topics in EM  series. 

Peter M. Antevy, MD , is a pediatric emergency medicine physician and medical director for Davie Fire Rescue and Coral Springs–Parkland Fire Department in Florida. 

Sheldon Cheskes, MD , is a professor in the Division of Emergency Medicine, Department of Family and Community Medicine, at the University of Toronto and a scientist at the Li Ka Shing Knowledge Institute at St. Michael’s Hospital, Unity Health Toronto. He is the medical director for the Regions of Halton and Peel with the Sunnybrook Centre for Prehospital Medicine and affiliate scientist with the Sunnybrook Research Institute in Toronto.