As in Incident command...

IC established!
We'll have several different sections reporting in - recent research, local topics, or highlighting areas of the Sponsor Hospital Council of Greater Bridgeport protocols.

*** Keep in mind - this website does not replace your protocols, and these posts do not reflect SHCGB or Bridgeport Hospital policies. This is a place to discuss research, controversies, or discuss possible future protocols. When in doubt, check your current protocols through the official source.

Tuesday, September 11, 2012

Should we increase the duration of CPR before "calling it?"

If you have worked in EMS for longer than one shift, you know how most codes will end up.

True story: During my tech/medic days, I got to tie a few toe tags myself.

This is old news - if EMS doesn't get a pulse back by the time the patient is moved onto the longboard, the prognosis is grim.

That's why the authors of the ACLS guidelines support appropriate field termination-of-resuscitation, writing in the 2010 Ethics portion of the ACLS guidelines:
Field termination reduces unnecessary transport to the hospital ... , reducing associated road hazards that put the provider, patient, and public at risk. In addition field termination reduces inadvertent paramedic exposure to potential biohazards and the higher cost of ED pronouncement. More importantly the quality of CPR is compromised during transport, and survival is linked to optimizing scene care rather than rushing to hospital.
They provide a suggested algorithim for field termination by paramedics, based on the most recent evidence:

The protocols for the Sponsor Hospital Council of Greater Bridgeport reflect this evidence. 

The relevant section, 3.16,  provides criteria for termination, which are, in part:

Most EMS systems around the country have similar protocols, and comfort with "calling it" onscene has been increasing in the EMS community.

And then this happens...

You can always trust what you read on Twitter!
Oh boy - are we going to go back to transporting them all to the ED?

So what's this new study?

This study used data from a retrospective registry, collecting data from 435 hospitals over a 9-year period (2000-2008). They ended up with about 65,000 patients who had had a cardiac arrest while in the hospital.

First, they looked at how long patients received resuscitative efforts, and calculated the average duration of CPR at each hospital for patients who did not have ROSC. 

That last part is key - when the news reports talk about "how long patients got CPR," what they really mean is "the average duration of CPR at each hospital for nonsurvivors." Subtle, but important difference.

Next, they figured out a way (regression analysis) to make all the patients and hospitals "equal." For example, patients who came in to the hospital with CHF were clearly at higher risk of cardiac arrest than those who had cellulitis, so they figured out the difference, and calculated it in mix.

Last step - they looked at the "average" cardiac arrest patient at the hospitals with the longest average duration of resuscitation, and compared the survival rate with patients at the hospitals with the shortest average durations.

What did they find?
First, they found a some variation in how long hospitals would attempt to resuscitate patients. The hospitals that "coded" non-survivors for the shortest time did so for about 16 minutes, while those hospitals in the more persistent group ran resuscitations for an average of 25 minutes. These averages hide, of course, a lot of individual variation.

So how did the patients at the 16-minute hospitals (quartile 1 in the figure below) fare compared to those at the 25-minute (quartile 4) hospitals?

There was a 12% difference. 

Put another way, for every 100 patients that survived a code at the 16-minute hospital, the 25-minute hospitals (on average) would have 112 survivors.

So, does this apply to my patients in the field?
First, keep in mind that this study only looked at in-hospital cardiac arrest. Importantly, they excluded arrests that occurred in the ED, and EMS codes were not included either. We already know that in-hospital cardiac arrest patients are different from those in in the pre-hospital realm, so the results are not immediately applicable.

Second, the results are modest, and may be as small as only a 2% survival advantage if you look at the confidence intervals (95% CI) in the figure.

But probably the biggest reason to not let this study change your practice is that it wasn't comparing patients - it was comparing hospitals. And those hospitals were likely doing more than just doing CPR for longer periods of time. They may have been employing better CPR, getting people and resources to the patient faster, or using better post-resuscitation care. We don't know anything else about the hospitals except that they tended to do CPR for longer on the non-survivors!

The Bottom Line
Data from retrospective, registry trials like this is provocative, not least because it can generate statistical results from the large numbers of patients included. It's impossible, however, to explain those results, or show cause and effect. For that we need prospective trials. 

I don't see anything in the current study that could justify changing our current approach, so keep following the current protocols for termination, provide quality CPR, and if you do have a reason to transport while continuing resuscitative efforts, do so safely!


  1. Agreed! Another important issue I had with this study is that they only looked at survival to hospital discharge. I don't have the full-text, but since, by and large, these are not the sudden cardiac arrest cases we see in the field with hearts "too good to die," but usually sick patients with multiple co-morbidities, I wonder what something more useful like one-year survival would show.

  2. They gave there results for not just ROSC, but for neurologically-intact hospital discharge as well, so I think that's about as much as we can hope for in a cardiac-arrest study!

    Generally, though, you are right - the in=patient arrest population is usually sicker, and has a worse short & long-term prognosis than the prehospital arrest.

  3. I wonder if they took all comers or VF/VT arrests only? In the field that certainly makes a difference and the Utstein Template only looks at VF/VT arrests.

  4. The funny thing is that the effect of longer CPR duration was MORE pronounced for the PEA/asystole patients. (Or rather, at hospitals where CPR was performed for a longer duration on non-survivors, there was a larger benefit for PEA/asystole patients than for VF/VT patients.)

    Given the study design, I wouldn't read too much into that, though.

  5. I never read anything into once study. I know that some people like to do that, but there are so many variables (not to mention fraud) that I like to see some consistent results before I adopt something new.

    1. Wise policy. Of course, half the fun in medicine is in running with the latest trend!

      BTW, if you want a pdf of the study, you can email me at brooks.walsh, a Gmail account. (Or with any other EMS related study.)

  6. Thanks for your information, it was really very helpful: Antipodes