IV versus IO for cardiac arrest (PARAMEDIC3 and IVIO)

Justin Morgenstern

1 year ago

Cite this article as:
Morgenstern, J. IV versus IO for cardiac arrest (PARAMEDIC3 and IVIO), First10EM, November 11, 2024. Available at:
https://doi.org/10.51684/FIRS.139138

In the most recent BroomeDocs podcast, we talk about an attempt to use the IM route to give cardiac arrest victims epinephrine faster. (The podcast and Research Roundups blog post might actually come out next week.) I was skeptical about the outcomes, partly because the evidence that epinephrine helps at all in cardiac arrest is rather shaky. However, if you really want to give epinephrine quickly, it seems like the intraosseous route is the better way to do that during an arrest. This month we saw two RCTs looking at IV versus IO medication administration for patients in out of hospital cardiac arrest, so let’s have a look at both.

Paper #1: PARAMEDIC-3

Couper K, Ji C, Deakin CD, Fothergill RT, et al. A Randomized Trial of Drug Route in Out-of-Hospital Cardiac Arrest. N Engl J Med. 2024 Oct 31:10.1056/NEJMoa2407780. doi: 10.1056/NEJMoa2407780. PMID: 39480216 ISRCTN14223494

The Methods

The PARAMEDIC 3 trial is a pragmatic, open-label RCT from 11 EMS services in the United Kingdom.

Patients

Adult patients with out of hospital cardiac arrest who required vascular access for medication administration during ongoing CPR. The only exclusion was known or apparent pregnancy.

Intervention

Intraosseous-first vascular access strategy.

Comparison

Intravenous-first vascular access strategy.

Shared procedures

The assigned vascular access was used until return of spontaneous circulation (ROSC), termination of resuscitation, hospital arrival, or the cannula was dislodged. If the paramedic could not obtain vascular access within 2 attempts, they could proceed with whatever option they wanted.

Outcome

The primary outcome was 30 day survival.

The Results

The trial was powered to demonstrate a 1% absolute difference in mortality between the groups, and required 15,000 patients to do so. The trial was stopped early because of slow recruitment and loss of funding, and so only 6,082 patients were enrolled. (It is therefore dramatically underpowered.)

The mean age was 68 years, with about 65% of patients being male. Most of the arrests occurred at home, with bystander CPR in about half of patients, and 20% of arrests with a shockable rhythm.

Median time from EMS arrival to vascular access was the same in both groups (12 minutes), and time to drug administration also looks to be the same (14 vs 15 minutes). I am not sure how that is possible, as the IO route had a 95% first attempt success rate, as compared to only 65% first attempt success with IV, so it really seems like IO should have been faster.

For the primary outcome of 30 day survival, there was no statistical difference, with 4.5% of the IO group and 5.1% of the IV group alive (aOR 0.94, 95% CI 0.68-1.32).

Favourable neurologic outcomes were seen in 2.7% versus 2.8%. ROSC and sustained ROSC to hospital handover were both slightly higher with IV (2% absolute).

Thoughts on PARAMEDIC 3

The big problem with this trial is obviously that it was stopped early and ended up significantly underpowered. That being said, they were hoping to find a 1% difference between the groups, and the absolute difference here was only 0.6%, so there is a good chance this trial would have been negative even with the full 15,000 patients.

I imagine there will be some debate about the small difference in ROSC seen here. Some people argue that, even though we have never seen good outcomes with medications in cardiac arrest, we should take all of the ROSC events we can get, because patients can’t possibly survive if we don’t get ROSC. I disagree with that argument. If no one is surviving long term, and especially no one is surviving with good neurologic outcomes, this seems like pure harm, with additional resource use and potentially harm to patients who have no hope of a positive outcome. (Patients don’t want to spend their final days in an ICU with no hope of survival).

The flip side to that argument is that more ROSC means more possibility of organ donation. I am not sure that is a great argument for epinephrine use in this setting, but certainly opens a fun ethical argument.

I would love to hear some practical details from the study authors, if any are out there, because it makes absolutely no sense to me that time to medication administration was the same in both groups when the IO was a far more successful approach. I assume something in their protocol was delaying IO use after it was placed? If so, these outcomes could potentially be improved by simply improving the IO protocol.

As an interesting EBM side note, there is a graph in the appendix of this trial that beautifully demonstrates the problem with stopping trials early. By random chance, survival in the IV group was dramatically higher in the first couple hundred patients, and survival in the IO group was dramatically lower. If this trial had been stopped after 1,000 patients, they might have erroneously concluded that IV is better. I have discussed this phenomenon many times below, but this might be one of the best graphical examples I have ever seen:

I am assuming this graph is explained by random chance. This paper doesn’t really describe previous experience with IOs, so another possibility is that we are seeing improved skill with the IO over time, which makes you wonder what this graph would look like with the full 15,000 patients. However, I think random chance is the much more likely explanation.

Paper #2: IVIO

Vallentin MF, Granfeldt A, Klitgaard TL, et al. Intraosseous or Intravenous Vascular Access for Out-of-Hospital Cardiac Arrest. N Engl J Med. 2024 Oct 31. doi: 10.1056/NEJMoa2407616. PMID: 39480221 NCT05205031