In Intravenous drug administration during out-of-hospital cardiac arrest: a randomized trial,^[1], ^[2] I missed some important information. I think I expected more detail to be in the text, than in the charts and figures, so I did not read Figure 2 carefully enough. Regardless of the reason, I did miss some important information.

In response, Anonymous left this comment -

RM: I was doing some reading on this article and came across your commentary. I am unsure if you will even see this since your blog was posted a few months ago, but I just thought I would clear something up for you and anyone else who reads your commentary. You seem to have done an inadequate job of reading and understanding this article.

Yes, I did.

Specifically, you state that a major shortcoming of the study is that 10% of the no-IV group received IV medications. You actually go on make a joke about "falling on a IV" as being an unacceptable excuse.

Sometimes I think I am funnier than I actually am.

If you had read the article closely you would have noticed a perfectly adequate explanation for this, as well as the 18% of the IV med group who did not receive meds.

Yes. It is not hidden. It should not have required a close reading, but I did miss both of these.

The reason the no-IV did receive drugs was because of ROSC (Return Of Spontaneous Circulation) then subsequent new cardiac arrest (27), hospital admission (13), breach of protocol (5)...45/433=10%. I would argue that 5/433 (1%), is pretty good for a study of this magnitude.

I agree. However, if the breach of protocol had been 10%, that would have been ten times higher and reason for comment. Clearly the breach of protocol was not that high and my criticism was inappropriate.

Paramedics often make mistakes. Arrests are stressful. It seems unreasonable for you to criticize this study for 5 episodes. It would be difficult if not impossible for the researchers to have controlled that any better short of being in every ambulance. I hope this sheds a little light on the topic for you.

I agree. I was wrong. Thank you for correcting me.

On the other hand, I think that cardiac arrests should not be stressful. We spend so much time on teaching the use of medications, advanced airways, and IVs. We have no evidence that any of these lead to better survival. These unproven interventions are just more to distract EMS from what has been shown to work by good evidence. We create unnecessary stress.

What you pointed out only lends more support to conclusion of the study - that IV medications do not improve outcomes from cardiac arrest. There still is no evidence that routine use of IV medication does anything to improve survival from cardiac arrest. Without evidence to show improved survival, these treatments should only be considered experimental.

In the bizarre world of medical research in the US, these experimental treatments have become the standard of care. The lack of research cannot be overturned with anything less than the highest quality research. That research is more likely to be done outside of the US, because it is considered unethical to deprive US study participants of a standard of care, even though there is absolutely no evidence of improved survival with the standard IV medications in cardiac arrest.


Anyway, back to my errors. I made the same mistake with the protocol deviations in the IV arm of the study.

442 Randomized to intravenous administration group

418 (95% of 442) Included in primary analysis

24 (5% of 442) Excluded due to predefined exclusion criteria

17 (4% or 442) Cardiac arrest witnessed by ambulance crew

6 (1% or 442) Resuscitation not attempted

1 (<1% or 442) Traumatic etiology

344 (82% of the 418 Included in primary analysis) Intravenous drug administration established and administered as randomized

74 (18% of 418) Intravenous drug administration not established prior to end of resuscitation

42 (10% of 418) Restoration of spontaneous circulation before intravenous administration

12 (3% of 418) Inability to establish intravenous access

12 (3% of 418) Intravenous administration considered futile

8 (2% of 418) No explanation given

For the information already mentioned by Anonymous -

474 Randomized to no intravenous administration group

433 (91% of 474) Included in primary analysis

41 (9% of 474) Excluded due to predefined exclusion criteria

17 (4% of 474) Bystander physician ordered treatment

14 (3% of 474) Cardiac arrest witnessed by ambulance crew

5 (1% of 474) Resuscitation not attempted

4 (1% of 474) Traumatic etiology

1 Asthma-induced cardiac arrest

388 (90% of 433 Included in primary analysis) No intravenous drug administration established or administered as randomized

45 (10% of 433) Intravenous drug administration occurred

27 (6% of 433) Restoration of spontaneous circulation and new cardiac arrest

13 (3% of 433) Hospital admission

5 (1% of 433) Breach of protocol

I am a bit confused by the meaning of Hospital admission. Were these EMS responses to treat patients in hospitals? Elsewhere Hospital admission is used as a measurement of short term outcome - whether the patient survived to the hospital. Survival to the hospital is just a short term outcome that has led to the adoption of treatments which have later been shown to increased long term harm. therefore, these short term outcomes probably should be ignored, rather than highlighted.

For the IV group, there were 2% listed as no explanation given, but nothing specifically listed as protocol violations.

Rather than an IV/No IV approach, blinded randomization to use of a placebo syringe/active drug syringe should not be that much more difficult.

And we need to stop the ethicists from forcing experimental treatments on unsuspecting uninformed patients in the name of ethics.


Footnotes:

^ ¹ Intravenous drug administration during out-of-hospital cardiac arrest: a randomized trial.
Rogue Medic
Article


^ ² Intravenous drug administration during out-of-hospital cardiac arrest: a randomized trial.
Olasveengen TM, Sunde K, Brunborg C, Thowsen J, Steen PA, Wik L.
JAMA. 2009 Nov 25;302(20):2222-9.
PMID: 19934423 [PubMed - indexed for MEDLINE]

If you want to read the entire study, this link opens it in PDF.


.