Capnography in the ED

Continuous quantitative waveform capnography, also known as end-tidal carbon dioxide, PetCO₂, or ETCO₂, is a measurement of the partial pressure of CO₂ in the exhaled breath. This technology has been around since the mid-19^th century and only relatively recently has its potential in emergency medicine begun to be explored.

Cardiopulmonary Resuscitation

The value of capnography during CPR was recognized as far back as 1987, specifically in a paper by Garnett et al¹ showing that end-tidal CO₂ (ETCO₂) was often the first indicator that return of spontaneous circulation (ROSC) occurred. However it wasn’t until 2010 that the AHA guidelines for CPR included the use of capnography as a Level I recommendation for ET tube verification, Level IIa for detecting ROSC, and Level IIb for monitoring CPR quality².  This was again stressed in the 2013 AHA consensus statement³, regarding the use of capnography:

• Most reliable method of confirming and monitoring placement of the ETT (see also Silvestri et al⁴)
• Should be primary physiologic metric when neither an arterial nor central line is in place
• Potential to guide optimization of compression depth and rate and to detect fatigue in the provider performing compressions: consider improving technique or switching provider if ETCO₂ <10 mmHg with a goal of >20 mmHg
• Consider an abrupt sustained increase to a normal value (35-40 mmHg) as an indicator of ROSC
• Failure to maintain ETCO₂>10 mmHg during adult CPR reflects poor cardiac output and strongly predicts unsuccessful resuscitation
• Avoid unnecessary pulse checks if the level is not compatible with organ perfusion

Heradstveit et al⁵ published a retrospective study of 575 cardiac arrest patients in 2012. They found that the ETCO₂was significantly different in patients with and without ROSC, but depended on the cause of the arrest. Respiratory arrests had higher levels than cardiac, and pulmonary embolism had the lowest levels. For these reasons they concluded it will be difficult to establish a single initial cut-off prognostic value. Also, given the low ETCO₂ found in patients later determined to have PE, they postulated that persistently low levels combined with clinical suspicion of PE during ACLS may be an indication for thrombolytics.

A systematic review of 23 observational studies of the prognostic value of ETCO₂ in CPR by Touma and Davies⁶ in 2013 again concluded that no single cut-off value as a predictor of death could be established. However, they said ETCO₂ values could be used in conjunction with other prognostic factors and clinical findings to support the decision-making process to terminate CPR.

A prospective study of 80 patients published in 2014 by Akinci et al⁷ looked at ETCO₂  levels after 5,10, 15, and 20 minutes of CPR and showed that levels after 20 minutes to be the most reliable for differentiating patients who will achieve ROSC from those who will not. The best intersection point was 28 mmHg (sensitivity 86%, specificity 80%). No one with level <14 mmHg survived. There was no statistical difference among the presenting rhythms.

The most recent study supporting ETCO₂ as a marker of ROSC concluded that a sudden increase of ETCO₂ >10 mmHg is likely to indicate ROSC and a pulse check is warranted at this time⁸. They noted similar findings were shown previously by Grmec et al⁹. They also noted levels <10 mmHg indicates a very low chance of successful resuscitation, which was also shown previously by Wayne et al¹⁰.

Edelson et al¹¹ looked at capnography as a tool for measuring ventilation rate during CPR. They concluded that it was more reliable than standard chest wall impedance algorithms, although both methods tended to underestimate the true rate. This is particularly important as the literature supports the fact that over-ventilation is a common problem despite the AHA recommendation of a respiratory rate of 8-10bpm during ACLS.

To summarize, capnography should be used during CPR to:

Finally, keep in mind that ETCO₂ levels can be transiently affected by administration of both bicarb and pressors.

Septic/Critically Ill Patients

New literature has emerged in the last couple years suggesting a use for capnography in evaluation of septic/critically ill patients. Hunter et al^12-13 found that pre-hospital capnography was more predictive of in-hospital mortality than any other vital sign. It was associated with serum bicarbonate, anion gap, and lactate. ETCO₂ was significantly lower in those who died. Levels from 31-41 mmHg were considered normal and values outside this range had 93% sensitivity for mortality (specificity 44%, NPV 99%). They also found that in patients with suspected sepsis, there were three independent predictors of mortality:

1. Use of pressors
2. Mechanical ventilation
3. Abnormal ETCO₂

They concluded that ETCO₂ may perform similarly to lactate as a predictor of mortality in patients with suspected sepsis (excluding patients with COPD and asthma). It is also faster, less invasive, and may reduce time to recognition of the severity of illness.

Another study looked at ETCO₂ and volumetric carbon dioxide (VCO₂= volume of expired CO₂) as predictors of fluid responsiveness in hemodynamically unstable patients¹⁴. They determined either of these parameters could be used as an indicator of fluid responsiveness following a preload challenge. However, absolute changes in ETCO₂ were small and thus VCO₂ may be more useful.

Other proposed uses for capnography include monitoring of patients with DKA, recognition of respiratory failure in seizing and post-ictal patients, and monitoring response to treatment of CHF, COPD, and asthma¹⁵.

Pulmonary Embolism

The potential role of dead space ventilation in the evaluation of PE was proposed back in 1959. PE creates dead space, which is then translated as a drop in alveolar and thus end-tidal carbon dioxide.  The dead space can then be calculated using arterial CO2 and ETCO₂. A meta-analysis¹⁶ of 14 trials found that when all the data was pooled, the overall sensitivity and specificity was 0.8 and 0.49, respectively. Due to heterogeneity and variation in measurement techniques (some used dead space as an end-point while others used ETCO₂) they were unable to calculate an optimal diagnostic threshold. They concluded that capnography might permit the exclusion of PE in cases in which the pre-test probability is <10%.

One trial¹⁷ in the meta-analysis found that an ETCO₂ ≥ 36 mmHg, when combined with a Well’s score ≤ 4, had a NPV of 97.6%.

Another trial¹⁸ found a statistically significant difference in ETCO₂ in patients with and without PE. However, they also concluded that the sensitivity is not sufficient to rule out PE, but may help reduce unnecessary testing when combined with D-dimer and other scoring systems.

Procedural Sedation and Analgesia

The current literature regarding use of capnography for procedural sedation and analgesia (PSA) in the ED can be summed up by the 2014 ACEP policies subcommittee on PSA statement 19:

“Although the routine use of capnography appears to decrease the incidence of hypoxia and respiratory events as defined in these studies (Level B recommendation), currently there is a lack of evidence that capnography reduces the incidence of serious adverse events during procedural sedation and analgesia such as neurologic injury caused by hypoxia, aspiration, or death. Future studies should focus on these areas to provide a better understanding of these outcomes.”

For further reading refer to studies by van Loon²⁰, Deitch²¹, and Waugh²². Also see the pro and con editorial in Annals of Emergency Medicine June 2013²³.

Sources // Further Reading