Our previous post introduced you to the normal and abnormal capnogram and the end-tidal carbon dioxide (ETCO2). In this post, we will dive deeper into ETCO2 and explore how this relates to the clinically significant arterial partial pressure of carbon dioxide (PaCO2). Current guidance encourages the use of ETCO2 as a proxy for PaCO2. However, ETCO2 and PaCO2 do not always match, and the relationship between them may not be as predictable as previously thought.
How about a quick physiology primer?
Before exploring the differences between PaCO2 and ETCO2, it is important to recap some of the basics of ventilation and perfusion.
The lung can be divided into three zones based on the relationship between the airway pressure in the alveoli and the perfusion pressure of the arteries and veins. These zones are called West’s Zones (although apparently, West wasn’t the first person to describe the concept!).
In Zone 1, the alveolar pressure (air pressure) is highest, but arterial and venous pressures are lowest, partly because of gravity and the increasing distance the blood travels. This means that the alveoli in Zone 1 are less perfused than those in other lung zones, and gas exchange is less efficient. In contrast, in Zone 3, arterial and venous pressures are higher (leading to greater blood flow), whereas alveolar pressure is lower than in Zone 1.
The fraction of the tidal volume (the amount of air that moves in and out of the lungs with a normal breath) that does not participate in gas exchange is called dead space. Dead space can be split into apparatus, anatomical, and alveolar.
Changes in cardiac output alter blood flow to the lungs and alveolar perfusion. Increased cardiac output will result in improved lung perfusion, enhanced gas exchange, and a higher ETCO2. In contrast, decreases in cardiac output will reduce blood flow to the lungs, decrease alveolar perfusion, increase alveolar dead space, and result in a reduced ETCO2 (even though PaCO2 may be elevated).
What is the PaCO2-ETCO2 gradient?
The PaCO2-ETCO2 gradient is the difference between arterial and end-tidal carbon dioxide. This difference is due to the alveolar dead space, which is small in healthy children and young people. Alveoli that are ventilated but not perfused have a gas mixture that is nearly identical to that inspired. This dilutes expired carbon dioxide and reduces ETCO2, resulting in ETCO2 being slightly lower than PaCO2. Under normal physiological conditions, this gradient is approximately 0.5 kPa (3.8 mmHg).
What can increase the PaCO2-ETCO2 gradient?
The gradient can increase with changes in the perfusion of the lungs. Local perfusion changes may occur due to pulmonary embolism, infarct or contusion. More globally, reduced perfusion may occur due to hypovolaemia, hypotension, cardiac failure, pulmonary hypertension, or cardiac arrest. All these perfusion changes can increase the gradient. Since the publication of the ARDSnet trial in 2000, which demonstrated improved outcomes with lower tidal volume ventilation of 6 ml/kg, it is likely that we have observed a greater PaCO2-ETCO2 gradient owing to an increase in the ratio of physiological dead space and tidal volume. However, this has not been confirmed in a clinical study.
Can the ETCO2 be higher than the PaCO2?
Although less common, it is possible for ETCO2 to exceed PaCO2.
This tends to occur if the alveolar CO2 changes significantly over the course of a breath, when there is a high tidal volume and low respiratory rate, low functional residual capacity, or lung compliance, or if expired CO2 is rebreathed (which is particularly relevant for our anaesthetic colleagues). This has previously been described most commonly in pregnancy, obesity, and infants.
Can the PaCO2- ETCO2 gradient be predicted?
Current guidance recommends an ETCO2 of 4.0–4.5 kPa (30.0–33.8 mmHg) as a stand-in for a low-normal PaCO2 with an expected difference of 0.5 kPa (3.8 mmHg). These guidelines are based upon evidence extrapolated from healthy individuals, often in the controlled setting of an operating theatre. However, this difference may increase in the presence of ventilation-perfusion mismatch, acid-base disturbance, and haemodynamic instability.
More recent studies have demonstrated the gradient in unstable patients exceeds the expected difference of 0.5 kPa (3.8 mmHg). This has been observed to result in patients with high or low CO2 being misclassified as having normal CO2. Moreover, a correlation has only been moderate, meaning there is variation between the levels, making it difficult to predict the difference in the gradient.
Given that the gradient is more significant than previously thought in unstable patients and that there is a lack of a predictable relationship, the ETCO2 is not a suitable stand-in for the PaCO2 where pH or PaCO2 requires precise control. This is particularly relevant for traumatic brain injury, where high carbon dioxide can lead to raised intracranial pressure, whilst low carbon dioxide can lead to cerebral ischaemia. In such situations, arterial measurement may be more appropriate.
Take-Home Points:
In healthy patients the normal PaCO2-ETCO2 gradient is approximately 0.5 kPa (3.8 mmHg).
The gradient increases in unstable patients
Changes in the gradient may not be predictable
In critically ill patients where careful control of arterial carbon dioxide is required, arterial samples should be taken when practical
References
Acute Respiratory Distress Syndrome Network, Brower RG, Matthay MA, et al. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med. 2000;342(18):1301-1308. doi:10.1056/NEJM200005043421801
Hibberd O, Hazlerigg A, Cocker PJ, et al. The PaCO2-ETCO2 gradient in pre-hospital intubations of all aetiologies from a single UK helicopter emergency medicine service 2015–2018. Journal of the Intensive Care Society. 2022;23(1):11-19. doi:10.1177/1751143720970356
Kodali BS. Capnography. 2022. Accessed online at https://www.capnography.com
Price J, Sandbach DD, Ercole A, et al. End-tidal and arterial carbon dioxide gradient in serious traumatic brain injury after prehospital emergency anaesthesia: a retrospective observational study. Emerg Med J. 2020;37(11):674-679. doi:10.1136/emermed-2019-209077
West JB. Respiratory physiology: the essentials. 9th ed. Baltimore, MD, USA: Lippincott Williams & Wilkins, 2012.
Yang JT, Erickson SL, Killien EY, et al. Agreement Between Arterial Carbon Dioxide Levels With End-Tidal Carbon Dioxide Levels and Associated Factors in Children Hospitalized With Traumatic Brain Injury. JAMA Netw Open. 2019;2(8):e199448. doi:10.1001/jamanetworkopen.2019.9448
Yarstev A. The Respiratory System. 2022. Accessed online at https://derangedphysiology.com/main/cicm-primary-exam/required-reading/respiratory-system
The area referred to as “dead space” in the third figure is not necessarily true dead space…assuming the “ideal alveolar gas” approach was referenced (i.e., using PaCO2 as a surrogate for PACO2). This approach is more of a global index of V/Q abnormalities and essentially overestimates physiologic dead space when shunt is present (which affects the PaCO2, of course). This approach assumes that PaCO2 closely approximates PACO2, which is not the case in many critically ill and healthy subjects. The Dubois method, which utilizes “mean” PACO2, seems to be a more reliable/approachable method to estimate actual physiologic dead space. Each approach has its own space, but to refer to the PaCO2-PetCO2 as “dead space” is misleading.