Capnography as a predictor of ROSC

This post can also be found at 510 Medic.

It’s no secret that I’m a fan of capnography.  One of the reasons I started a blog was to pass on what I felt were the best practices in EMS. Really that’s a huge part of the EMS 2.0 movement.  So let me just say again, for the record, that capnography absolutely qualifies as a “best practice” and may be one of the greatest tools added to our arsenal in my time in the field.  Though not universal, many EMS systems are now utilizing waveform capnography for confirmation of advanced airway placement in cardiac arrest patients.  But what else can capnography do?  Really, the question should be, what can’t it do?

Most providers are aware that higher capnography readings during a resuscitation are associated with a pending return of spontaneous circulation[1].  But is this true for all patients?  In a study published in Critical Care, researchers looked at initial and serial CO2 readings in full arrest patients with both cardiac and respiratory origins [2].  The findings were interesting:

• Patients in asphyxial cardiac arrest had higher initial EtCO2 readings
• This high initial reading was not a ROSC predictor
• Asphyxial arrest patients experienced a gradual decline in EtCO2 readings, then an increase in patient who eventually experienced ROSC
• In cardiac arrest of cardiac origin, patients who eventually regained pulses had a higher initial EtCO2 than those who did not
• In both groups, the EtCO2 reached a prognostic value after 5 minutes of CPR

So what does this mean for us?  Picture the following scenario:

You arrive on scene of a cardiac arrest first.  After assessing the patient and finding no pulse, your partner begins chest compressions and you do one of the following next:

• A.  Place the patient on the monitor
• B.  Place an OPA and start to ventilate the patient

The answer to that question ultimately depends on the cause of the arrest right?  A patient in cardiac arrest from primarily cardiac causes is likely to be in VF/pulseless VT (depending, of course, on downtime, bystander CPR and the like) and will benefit from early defibrillation.  If a patient is in cardiac arrest from primarily respiratory causes (for instance a choking), they are more likely to need oxygenation.  Ultimately, the history of the patient and their current medical condition affects the order of your steps on a call.

Now skip ahead a bit in the call.  The next due unit has arrived and you have all the manpower you could possibly need.  The patient is intubated and on the ECG monitor (which is showing PEA).  Your initial post-intubation EtCO2 reading was 30 mmHg, but it has dwindled over a period of minutes to 15 mmHg.  One of the other responders mentions that this might indicate that the resuscitation will be terminated.  Is this true?  It all depends on the patient’s history.  If this was a cardiac arrest of respiratory origin, this may be the expected dip in the EtCO2 before an anticipated rise and, ultimately, a return of spontaneous circulation.  If this is a cardiac arrest of primarily cardiac origin, it may well mean that the patient will not regain pulses.  The take home lesson in this study is that capnography is really just a tool and findings must be interpreted in conjunction with a thorough history.

STUDY LIMITATIONS
The authors cite several limitations to their study including the need for a larger cohort study and the fact the EtCO2 is only an approximation of cardiac output.  I’d like to add another one for application of this study to EMS:  time frame.

The study found that EtCO2 values in cardiac arrest from both primary cardiac and asphyxial causes reached a prognostic value for ROSC in five minutes.  Even in an urban system (with lots of paramedics), I would guess that it is probably unlikely that intubation has been performed with regularity within five minutes of the start of a code, even less likely within the three  minutes required to see the drop in EtCO2 for asphyxial arrest.  There is a way around this, however.  Our crews have had good luck placing the capnography fitting between the BVM and facemask when using a BLS airway.  This could allow for collection of capnography readings immediately upon start to run the code.  Does anyone else use capnography inline with the BVM and facemask?  Any feedback or stories?

CITED ARTICLES
[1] - Weil MH: “Partial pressure of end-tidal carbon dioxide predicts successful cardiopulmonary resuscitation in the field”. Crit Care 2008, 2:90.

[2] – Lah, K; et. al: “The dynamic pattern of end-tidal carbon dioxide during cardiopulmonary resuscitation – difference between asphyxial cardiac arrest and ventricular fibrillation/pulseless ventricular tachycardia cardiac arrest”.Crit Care 2011, 15:R13.