Sodium bicarb in a code -Still no evidence

Not sure why, but an EMS Facebook page decided to highlight a JEMS article written over a year ago. It got me to thinking, however.

The JEMS article is titled "Sodium Bicarbonate: Should it be considered as a treatment?" The author is Jim Davis, a FF/RN/EMT-P, and he covers some of the physiologic rationale for the use of bicarb in cardiac arrest, as well as some of the animal and clinical evidence. In the end, he comes out fairly strongly in favor of using it. He concludes:

The routine use of NaHCO3 may warrant a greater role in the delivery of prehospital care for cardiac arrest patients in the prehospital setting.

Until other options exist for the prehospital assessment and treatment of acidosis, EMS providers need to pay more attention to the length of the patient’s down time and recognize the importance of considering acidosis early on, as well as recognizing that NaHCO3 may be a viable treatment option.

I earned my first ACLS card in 1996, long after bicarb had fallen out of fashion, so I had never reviewed the data personally that supported its use in cardiac arrest. Given the number of comments that followed the FB post, however, it appears that this old drug still has a number of "fans" (125 likes!), so I thought that I should acquaint myself with the studies.

Didn't take long - there are only 2 human clinical trials that the AHA cites in the 2010 ACLS guidelines. Let's look at the first.

The first was from 2005, "Improved resuscitation outcome in emergency medical systems with increased usage of sodium bicarbonate" It reviews data from a trial that took place from 1990 to 1992, and it has an interesting design.

First of all, the study (named BRCT III) was primarily designed to look at high-dose versus escalating-dose epinephrine. (High-dose turned out to be a bad idea.) So, basically, the whole paper is an exercise in "data-dredging," because the trial was never designed to look specifically at the use of sodium bicarbonate in cardiac arrest.

They had 16 EMS agencies involved in their study, and some used bicarb almost all the time, while others hardly ever busted out the big yellow box.

 So, they compared the EMS agencies that were "high users" against those that were "low users"

SB = sodium bicarb; ACLS = first defibrillation

 When they identified these agencies who gave bicarb more often and earlier in cardiac arrest (purple box), they found that their patients were more often alive, and neurologically intact, 6 months later (red box). On the other hand, these same "high SB users" were also much better at getting in the first defibrillation faster (blue box), as well as getting in a line, etc. These agencies, simply put, seemed to better at running cardiac arrests in general!

Given the study design, there's no way to piece apart how much, if anything, the bicarb helped.


Anyway, there were a lot of other interesting points in the paper (e.g. giving bicarb earlier doesn't have a physiologic rationale...), but let's move on to paper #2!

The design of this study is even shakier than the preceding study. Weaver et al., conducted a study that ran from 1983-1985 with Seattle EMS, looking at whether it was better to give lido or epi in refractory VF. To summarize the whole paper, I'll just show you this graph:

No difference! Hey, what about the bicarb!? Wasn't there supposed to be bicarb in the study?

Here's how they got bicarb in the study. They decided to look at the cardiac arrests that Seattle EMS had treated in the 2 years proceeding the study, i.e. in the years 1981-1982. Ancient history, indeed.

One of the medics in the study, preparing the bicarb.

During that time period, the common practice had been to start a slow sodium bicarb drip, and not give epi or lido. So they compared the patients from their "true" study with this historical cohort - that's a pretty messy study! And the results were:

Hey, wasn't this supposed to be one of the papers supporting bicarb in cardiac arrest? Instead, the authors conclude that:

In this study, there was no clinical evidence to support any form of drug therapy for initial treatment of persistent ventricular fibrillation.

The Bottom Line
The Bridgeport protocols (PDF download) suggest the use of sodium bicarb in only a few situations:

• Hyperkalemia
• Tricyclic OD
• Metabolic Acidosis 

 Based on this review of the studies cited by the AHA, there does not appear to be any reason to add cardiac arrest to the list. If you hear of any better evidence, send it this way!