The NEM conference this week sparked some debate on twitter with @EMManchester questioning whether you can call yourself an up to date Emergency Physician if you can’t and don’t use POCUS in your resus room and @POCUSFOAMED proclaiming it borders on the negligent citing that it’s the standard of care now for EPs in the UK (given that Level 1 US is a requirement for CCT).
So what about some evidence?
We turn to the REASON trial this month for our JC.
Title of the Paper: Emergency Department Point of Care Ultrasound in out of hospital and in ED cardiac arrest.
Published: Resuscitation 2016
You’re working in resus, the bat phone rings……
Out of Hospital Cardiac arrest……
Immediate bystander CPR. PEA.
ETA 5 mins. You prep the team.
The patient arrives, you all work quickly and quietly, you know the drill, you’ve rehearsed this cognitively and practically multiple times. The presenting rhythm is PEA. You get more adrenaline ready, its due in the next cycle (a separate debate i know!!!).
You do a bedside ECHO during your next rhythm check.
Not a flicker.
You continue on for a further 16 minutes and decide to stop. He’s now in asystole and has been down for 30minutes. As you debfrief the team you begin discussing whether that ECHO added anything, or if it could have told you to stop at arrival.
Clinical Question: In adults presenting in a non shockable rhythm does cardiac activity on bedside ECHO predict survival to hospital admission (or discharge?).
Population: Adults presenting in atraumatic Cardiac Arrest (Non Shockable Rhythm)
Intervention: POC-ECHO at arrival and end of resuscitation attempt
Outcome: Hospital Admission
The Scan: Done by the treating team, who were unblinded to the outcome of the scan. A sub-Xiphoid or Parasternal long axis view were taken at the rhythm check during the cardiac arrest management. Have a look here for some ECHO pearls and here for the basics.
Generalisability: This was a Prospective observational study done across 20 sites in the United States and Canada between 2011-14. Does this mean its not generalisable to our patients in God’s own County. Well maybe not. The US and Canada are geographically huge, the prehospital set up is very different with intubating prehospital services, mechanical CPR devices and long transfer times. In our neck of the woods you’re rarely more than 10-15 minutes away from a hospital on blue lights. So maybe something happens en route when you’re en route for longer? Our paramedics are fantastic and have protocols in place to allow them to ‘call it’ out in the community given certain variables – such as a prolonged asystolic arrest, so we may not see the same cohort of patients as studied here.
Exclsuions: These are discussed in the paper and caused some discussion at the face to face journal club. Patients were excluded if they were found to have a DNACPR order (entirely appropiate) but were also excluded if resuscitation attempts were short (< 5 mins) or if the team decided to stop after the initial US. We're not quite sure why, and potentially this group of patients could have had a significant effect on results should they have been included - there were 148 of them. The reasoning for this isn't explained in the paper as far as I can tell. Patients: This was a 3 year and 20 centre study, yet only 953 patients were enrolled. We’re not told how many were screened, and what happened to those that were. I would assume that the number of OOHCA across all those centres and time frame would be higher than 953. I wish they’d told us….(i’m also happy to be corrected!)
Results: So 793 patients were eventually included. Most were men (62%) and mean age was 64.2. 33.8% had bystander CPR (despite 42.1% being a bystander witnessed arrest) make of that what you will……PEA was the most common presenting rhythm (414 vs 379).
For predicting non survivors to admission the absence of cardiac activity on bedside US in asytole had a PPV of 94% (93-95%) and in PEA (0.86-93%). So pretty narrow 95% Confidence Limits when you have asystole and no cardiac activity on initial US. The figures for survival to discharge are even more bleak asystole 0.99 (0.99-1) and PEA 1 (0.98-1). Wow. That’s pretty bleak, and those confidence limits are lovely and narrow.
An area of real interest to me was the patients who were found to have some evidence of the commonly recited resuscitation council Hs and Ts of reversibility – namely pericardial effusion or signs consistent with a PE (or Thromboembolism to give it’s “T”). POCUS potentially stops you ‘guessing’ at the presence or absence of these factors. The patients in these small subgroups in the study (34 for pericardial effusion) had a higher survivor rate that the rest of the cohort (15.4%).
They used multivariate analysis to determine that the presence of cardiac activity on ECHO was most predictive factor of survival in arrested patients (OR 5.7 95% CI 1.5-21.9) but with pretty wide confidence limits! There is no comment made as to the quality of survival of those individuals due to resource constraints within the trial – this is a shame as neurologically intact survival must be the best POO (patient orientated outcome) going!
I like this study. It adds weight to the merits of POCUS especially in the resus room. The primary outcome measure is patient orientated which is great, and I think in a perfect world we’d see and RCT of POCUS vs no POCUS with the primary outcome measure being survival to neurologically intact discharge, but I reckon the numbers needed and the potential confounders would be vast and thus such a trial wont happen.
For the time being, for me, this study is the best evidence we have that tells us POCUS is a useful adjunct in resuscitation decisions and predicting survival from non shockable arrests, but it shouldn’t be the only tool we use to make those decisions.
Make sure you check out the St Emlyn’s review of the same paper here (They’re far more learned than I am!)
You can download the one minute wonder here.
Keep fighting the good fight.
Emergency Department Point of Care Ultrasound in out of hospital and in ED cardiac arrest http://dx.doi.org/10.1016/j.resuscitation.2016.09.018