The drugs don’t work…

At least in cardiac arrest that is.  Maybe…

Published today in the NEJM is the latest offering from the Resuscitation Outcomes Consortium (ROC), entitled Amiodarone, Lidocaine, or Placebo in Out-of-Hospital Cardiac Arrest.

Switched on readers will already suspect that this study looks at the effect that anti-arrhythmic have in OOHCA, which to date has been largely unknown – it just seemed like a good idea.

I have no doubt that this study will be dissected in depth by minds far greater than mine, but I thought I would put the paper out there with some rambling, semi-coherent thoughts to go with it.

First of all, anti-arrhythmics seem like a good idea on the surface, but with a bit of reflection, they actually seem a bit weird.  We know one thing will help in VF/VT – defibrillation.  CPR can help keep things ticking until we hit the big orange button, but it’s really all about the joules.  When we have recurrent VF it seems intuitive that giving a drug to stop the VF happening again is a good thing.  But… the only thing that works in VF is electricity.  Amiodarone raises the defib threshold – i.e. it makes it harder to effectively defib.  So if we can’t stop the VF, we make it harder to stop the VF by trying to stop the VF (and of course we are giving heaps of adrenaline, which makes VF more likely, at the same time as trying to stop the VF and making the only effective treatment less likely to work)

My head starts spinning at this point.

Back to the study:  This looks like a well done piece of work to me (not being a mathemagician).  It is a genuinely blinded, randomised, multi-centre trial in what looks like high performing services throughout North America.  The interventions are in line with what we would do Down-Under, so it is likely generalisable to our setting.  I can’t comment on the types of statistical analysis performed, as I cannot count past 10 with my shoes on, but the populations seem well matched.
Importantly, the outcomes measured are good – survival to discharge and neurological function at discharge.  Gone are the days of accepting survival to hospital as a reasonable outcome: it’s not.

So, on to the results.  Anyone with any interest in this sort of thing will not be surprised to discover that the outcomes are not statistically different between the groups.  That is, the drugs don’t work.  Or at least these drugs don’t (and lets face it, no-one is fooled by adrenalines charade any more either).

However, there are some interesting points raised in the discussion.  First of all is the fact that we provide the same treatment to patients who will not, and cannot survive, as to the patients who may actually survive.  This has the potential to dilute the treatment effect, as many of the patients will not survive whatever we do for them, and any positive effect of treatment may be washed out.

Also, if there is an actual treatment effect of 3% for amiodarone (and the study may be underpowered to detect this), then this may actually be a significant result.  Although not “statistically” significant, given the massive burden of cardiac arrest internationally, this may nonetheless be numerically significant.  As the authors noted, this could represent approximately 1800 lives saved in North America alone, which is pretty impressive for an “insignificant” result.  Nevertheless, this study cannot actually answer that question.

This looks like a good study, and it certainly seems that the ROC people turn out some good work.  I’m not convinced that I will see any change to my practice coming down the line any time soon (given I still have to give sodium bicarbonate to arrests), but one never knows…
I do, however, look forward to someone who knows what they are talking about dissecting this paper for me.

As always, comments, questions, marriage proposals, or general abuse below.