
Aman is a critical care medicine fellow at the Cleveland Clinic. Interests: Cardiopulmonary physiology, shock, POCUS, mechanical ventilation, and ARDS. Music genres: Blues, Rock, and Heavy metal

Emergency Medicine Resident and MedED Enthusiast. Learning and teaching medicine one doodle at a time!
Pre-brief
A 50-year-old female of height 152cm develops severe ARDS and is being passively ventilated with a tidal volume of 275cc (~6 cc/kg IBW) and RR of 30/minute. Her ventilator circuit includes a heat and moisture exchanger and a ventilator mount. Her PaCO2 is 80mmHg and pH is 7.14. What can be done here to significantly enhance decarboxylation without increasing the risk of lung injury?
Definitions
Simply put, dead space represents the volume of ventilated air that does not participate in gas exchange. This concept can be extended to include factors that cause a dead space effect. A certain amount of dead space is normally present in every person (this is known as anatomical dead space: see below). The fraction of the tidal volume that does not contribute to gas exchange is known as dead space fraction (VD/VT; where VT = tidal volume and VD = dead space volume). VD/VT is a common way to quantify dead space. At this point, it is helpful to define minute ventilation and alveolar ventilation:
Minute ventilation (V̇E; the subscript ‘E’ denotes ‘exhaled’) is the total ventilation in one minute (units: L/min). E.g. in a passive mechanically ventilated patient on volume control (VC) mode, V̇E = VT x RR.
Alveolar ventilation (V̇A; the subscript ‘A’ denotes ‘alveolar’) is the amount of ventilation occurring in the alveoli in one minute. Its calculation requires exclusion of ventilation occurring in the anatomical dead space. Mathematically, V̇A is total ventilation minus anatomical dead space ventilation. E.g. in VC mode, V̇A = (VT – VD) x RR.
The alveolar ventilation controls CO2 homeostasis according to the alveolar ventilation equation:
PaCO2 ∝ V̇CO2/V̇A
, where V̇CO2 = CO2 production by the body (units: cc/min) and V̇A = alveolar ventilation.
Classification
The nomenclature describing dead space can be quite confusing (see [1] for in-depth reading). A conceptual, simplistic categorization is as follows:
- True dead space
- Anatomical dead space
- Alveolar dead space
- Mechanisms that create ‘dead space effect’
- Increased heterogeneity of V/Q ratio
- “Shunt dead space”
- Anatomical dead space:
The entire airway circuitry all the way from mouth to the terminal bronchioles (~generation 14-16) is the conducting zone of the respiratory system. The remaining circuit: respiratory bronchioles to alveolar sacs (generation 23) participate in gas exchange and is called the respiratory zone. Anatomical dead space is thus defined as the volume of the conducting zone (Figure 1). The dead space in an average adult has been reported to be ~150 cc or 2cc/kg ideal body weight.

- Alveolar dead space:
Normally, ventilation and perfusion of respiratory zones (lung units) are nicely matched (normal V/Q ~0.9). Consider a scenario where a small embolus blocks all perfusion to a lung unit. Since the perfusion of the lung unit reduces to zero, the V/Q ratio for this unit is infinity (V÷0 = ∞). A V/Q of infinity is the definition of alveolar dead space.
- Increased heterogeneity of V/Q ratio:
This is the most difficult to grasp but also perhaps the most important factor in many disease states. Normally, there is only mild heterogeneity of V/Q ratios but this is increased in various pathological states. If a substantial proportion of lung units have a very high V/Q (e.g. >= 10), this can cause a ‘dead space effect’. Again, this is not truly alveolar dead space since the V/Q ratio is not infinity.
Let’s take the example of a patient with ARDS who is set on a high PEEP and high tidal volume. The dependent lung units are inflated normally but the non-dependent units are now overinflated. Lung overinflation compresses the capillaries and reduces blood flow to those units. Hence, the V/Q in non-dependent regions rises to 10. At the same time, most of the blood flow is now diverted to the dependent lung units, where the V/Q drops down to 0.5. You can think of low V/Q as ‘relative hypoventilation’ of that lung unit 🡪 alveolar pCO2 increases. Furthermore, since the majority of the perfusion is happening in the dependent region, the global gas exchange mirrors the gas exchange in this area – overall resulting in a dead space effect.
- Shunt dead space
This is a misnomer. A significant shunt fraction can cause a dead space effect but this has obviously nothing to do with true dead space. The basic idea behind this is that the CO2 in the shunted blood never makes it to the respiratory zone and is hence retained by the body. However, this is a mild effect and a very large shunt fraction is needed to cause a significant dead space effect.[2]
Physiological dead space: a cumulative measure of wasted ventilation
Physiological dead space (VDphys) is another term that is troublesome but is commonly used in the literature. The term “wasted ventilation” has been proposed as an alternative, which sounds less technical but is certainly more accurate.[3] It is a cumulative index that incorporates all aforementioned causes of true dead space and dead space effects. Its formula is as follows –
(PaCO2 – PECO2)/PaCO2
, where PECO2 is the mixed expired PCO2. The widely available index of end-tidal CO2 can be used as a surrogate for PECO2 to calculate physiological dead space but they are not interchangeable. Calculation of PECO2 requires a technique called volumetric capnography that can be performed by some ventilators or dedicated devices.
It is easier to understand the concept of VDphys concept if we consider a lung model with only two units – one with normal perfusion and other with no perfusion (V/Q = ∞). Physiological dead space is then equal to the proportion of tidal volume that would have to be delivered to the unperfused unit to account for the measured difference between PaCO2 and PECO2. Although it is an excellent global measure of wasted ventilation, it doesn’t differentiate between the relative contribution of the four factors listed above.
Practical implications
The practical implications of dead space in disease states are immense and impossible to cover in a limited space. I will highlight just two important examples that relate to anatomical dead space, as it is the most intuitive.
Impact of breathing pattern on dead space fraction
Clinical example: A patient is mechanically ventilated and paralyzed on VC mode. Initial settings: TV = 700cc, RR = 10/min, V̇E = 7L/min. To enforce lung protective ventilation, the tidal volume is halved (350cc) and respiratory rate doubled (20/min), thereby achieving the same V̇E of 7L/min. Will PaCO2 be unaffected after this change?
To learn the answer, we need to calculate the alveolar minute ventilation in each case. Let’s assume the anatomical dead space (VD) is 150 cc/min, and there is no alveolar dead space or V/Q heterogeneity.
- V̇A with initial settings = (VT – VD)xRR = (700–150)x10 = 5,500cc/min
- V̇A after changing settings = (VT – VD)xRR = (300–150)x20 = 3,000cc/min
Hence, even though the minute ventilation remained the same, the alveolar ventilation reduced by 45%! Another way to look at this is to calculate the dead space fraction (VD/VT) for each breath. In the first case, it is 150/700 = 0.21. After changing the settings, it would be 150/300 = 0.5. Hence, for the same minute ventilation, a rapid shallow breathing pattern increases anatomical dead space fraction. It is important to be mindful of this factor while managing mechanical ventilation.
Impact of apparatus (instrumental) dead space
Another important factor often overlooked in clinical practice is that of apparatus dead space (VDapp). In a non-intubated patient, the source of gas is the ambient air and the anatomical dead space starts at the mouth opening. In an intubated patient, the source of gas is the ventilator and its tubing (inspiratory limb). The basic anatomy of this circuit is presented in Figure 2. The key point to appreciate is that the inspiratory limb is constantly filled with fresh gas and can be thought of as an extension of the ventilator itself. So practically speaking, the source of gas in an intubated patient is the wye (Y-piece) of the ventilator circuit. Any equipment distal to the wye constitutes apparatus dead space. At minimum, this includes the endotracheal tube (see appendix). Other potential sources include heat and moisture exchanger (HME), sensors (flow, end-tidal etc.), catheter mount (ETT extension) etc.

Apparatus dead space can be thought of as an extension of the anatomical dead space. The biggest contributor to VDapp are HMEs. An average HME adds ~50cc of dead space but the exact amount depends on the type. An alternative to using HME would be to use active circuit humidification, which is installed proximal to the inspiratory limb and hence does not contribute to dead space. A catheter mount is an extra tube that reduces the drag on the ETT. It can easily add ~25cc of dead space. I strongly recommend a smartphone app called VentilO[4] that provides detailed information on various sources of VDapp (Figure 3).

The importance of VDapp becomes magnified if the set tidal volume is low: e.g. lung protective ventilation pediatric patients and short adults. Clinical example:
Let’s revisit the case in the pre-brief: To recap, lung protective ventilation is being used for a female with severe ARDS: tidal volume of 275cc (~6 cc/kg IBW) and RR of 30/minute. Her ventilator circuit includes an HME and a ventilator mount, with a total VDapp of 75cc. On these settings, her PaCO2 is 80mmHg and pH is 7.14. What are some of the options to enhance decarboxylation?
- RR cannot be increased any further due to development of auto-PEEP
- An increase in tidal volume was attempted but that drove up plateau and driving pressures to unsafe levels (assume PEEP already optimized).
- Instead, the medical team decides to minimize VDapp by removing the catheter mount and switching to active circuit humidification instead of an HME. Can this simple maneuver provide any major benefit without touching any dials on the ventilator? Let’s do the math (assuming anatomical dead space = 100cc or ~2cc/kg IBW).
Before removing Vdapp:
Total dead space = anatomical dead space (100cc) + apparatus dead space (75cc) = 175cc
VD/VT = 175/275 = 0.63
Alveolar ventilation (V̇A) = (275-175).30 = 3,000cc/min
After removing Vdapp:
Total dead space = anatomical dead space = 100cc
VD/VT = 100/275 = 0.36
Alveolar ventilation (V̇A) = (275-100).30 = 5,250cc/min
Hence, alveolar ventilation increased by 75% simply by removing the HME and catheter mount! This effect would not be as dramatic if the baseline tidal volume is much larger than the VDapp.
The Debrief
- Dead space represents the volume of ventilated air that does not participate in gas exchange.
- The causes of true dead space are (a) anatomical dead space and (b) alveolar dead space. Conditions that create a ‘dead space effect’ are (a) high V/Q units in a heterogenous lung, and (b) shunt.
- Physiological dead space = (PaCO2 – PECO2)/PaCO2. It is the global measure of wasted ventilation that incorporates true dead space as well as conditions that create a dead space effect.
- Rapid shallow breathing pattern and respiratory equipment distal to the ventilator wye increase the anatomical dead space fraction.
References
- Robertson HT. Dead space: the physiology of wasted ventilation. Eur Respir J. 2015 Jun;45(6):1704-16.
- Wagner PD. Causes of a high physiological dead space in critically ill patients. Crit Care. 2008;12(3):148.
- Petersson J, Glenny RW. Gas exchange and ventilation-perfusion relationships in the lung. Eur Respir J. 2014 Oct;44(4):1023-41.
- VentilO. Site: https://play.google.com/store/apps/details?id=ca.qc.iucpq.ventilo&hl=en_GB Accessed on: 11-3-2020.
Brilliant post, thank you so much. One thing I’ve wondered though is how VDapp affects the lung mechanics and whether we couldn’t achieve the same thing by increasing delivered volumes?
As a thought experiment if I had a balloon attached to a 10cm piece of hose pipe, and an identical balloon (unchanged compliance) to a 20cm hosepipe and inflated both to the same volume, wouldn’t each balloon experience the same pressure (Pplat), with the increased Pinsp being secondary to increased resistance to flow?
If so couldn’t I just use the longer hose, deliver a larger volume (accounting for increased VDapp) and achieve the same ventilation with no deleterious affects?
(Sorry hope my semi-incoherent rambling/metaphor makes sense!)
Thanks Tim. This is in fact far from rambling and thought experiments like this are important to get a better grasp at things.
Firstly, we can easily leave resistance’ out of the discussion. Resistive load is a dynamic phenomenon that is only present *during* air flow. When you’re talking about compliance and Pplat, you’re discussing elastic properties of the respiratory system at zero flow (e.g. at end-inspiration).
Allow me to refine your thought experiment a bit:
You are ventilating a balloon with 10 cm hosepipe. Hosepipe = anatomical dead space; volume of hosepipe = dead space volume. For simplicity sake, let’s say tidal volume is twice the volume of the hosepipe (dead space volume).
Scenario 1:
You change the hosepipe to 20 cm and keep tidal volume same, your Pplat will be the same (same TV and balloon compliance). However, your dead space volume has now doubled (from 10cm to 20cm hosepipe) and it now equals tidal volume. Since tidal volume = dead space volume, “alveolar ventilation” is zero. This is equivalent to not ventilating at all.
Scenario 2:
You change the hosepipe to 20 cm and double the tidal volume to keep the alveolar ventilation same. In this case, CO2 will remain the same. However, since tidal volume has doubled, Pplat will double for the same balloon compliance.
Another insight that helps with this – Dead space volume is not vacuum. When a breath is delivered, gas in the dead space enters the lungs first (see Figure 1). So in scenario 2, gas in the longer hosepipe (20 cm) enters the balloon first before the ‘extra’ tidal volume can reach the balloon.
Bottomline: effect of a given tidal volume on Pplat only depends on balloon compliance. It doesn’t care about the dead space. So if you increase tidal volume, Pplat will go up and overdistension will occur. Hope that helps.
Thank you for this clear, practical explanation of dead space, will be using this as a student resource.
Curious of your thoughts on this, … re the final calculation, specifically the anatomical dead space, rather than approx 2cc/kg IBW, might the anatomical dead space be calculated as approx 1 cc/kg IBW with the ETT in place, considering that the conventional understanding re an appropriate sized ETT in situ reduces anatomical dead space by approx 50%? In that case the “average” size adult’s anatomical dead space = 150 cc, reducing this by 50% = 75 cc then adding 75 cc of VDapp brings the airway dead space back to 150 cc.
Respectfully submitted
Hi Michael.
In an average adult, a typical endotracheal tube has been shown to reduce anatomical dead space by about 40% so your analysis in indeed spot on.
Obviously, this doesn’t change the fact that you’d wanna keep it as low as possible in a sick ARDS patient with an otherwise large dead space fraction.
Thank you for your response Aman.
VDapp “as low as possible in a sick ARDS, …” absolutely!