Secondary Logo

Journal Logo

Technology, Computing, and Simulation: Research Reports

Measurement of Dead Space in Subjects Under General Anesthesia Using Standard Anesthesia Equipment

Badal, John J. MD; Chen, Kyung J. MD; Loeb, Robert G. MD

Author Information
doi: 10.1213/ANE.0b013e3182025a9d

Pulmonary dead space is the volume of gas that is delivered to the lungs but does not participate in gas exchange. It is made up of anatomic dead space and alveolar dead space. Knowing pulmonary dead space in patients under general anesthesia is clinically useful because it can aid in detecting disease processes such as pulmonary emboli or low cardiac output states. Because almost all of cardiac output is delivered to the lungs, low cardiac output states cause a decrease in pulmonary blood flow and thus an increase in alveolar dead space. Dead space can be calculated using the Bohr equation; however, accurate measurement of mixed exhaled carbon dioxide (PECO2) is needed. PECO2 is difficult to obtain from a standard anesthesia machine. It is usually measured using a Douglas bag. One technique to measure PECO2 is using volumetric capnography, but this requires additional expensive equipment. Soro et al.1 used volumetric capnography, along with arterial samples of partial pressure CO2 in arterial blood (PaCO2), to measure dead space in prone subjects under general anesthesia. Alveolar dead space can be calculated by subtracting anatomic dead space, estimated by the Fowler equal area method, from physiologic dead space.2 This is, however, a complicated mathematical model. Hardman and Aitkenhead,3 using a physiologic simulator, found that the Bohr-Fowler technique for the measurement of alveolar dead space/alveolar tidal volume (VT) to be marginally superior to volumetric capnography, but it requires the collection and analysis of expired gas over time. Other investigators constructed a device that attaches to the inspired outlet and the expired inlet of the circuit that can measure PECO2.4 This device, known as the Bymixer, requires additional equipment that adds weight to the circuit and it must be cleaned between uses. Previously, a study at our institution measured PECO2 by sampling the gasses in the bellows of a standard anesthesia machine.5 In this study, we used PECO2 measurements from the bellows to calculate dead space using the Bohr equation. We theorized that we would be able to measure a change in dead space after adding known amounts of apparatus dead space volume to the breathing circuit.


The study was approved by the Human Subjects Protection Program at the University of Arizona. All subjects gave written consent. A standard CO2 sampling line (#625N; Criticare Systems, Inc., Waukesha, WI) was placed inside the bellows of a Narkomed GS anesthesia machine (Dräger Medical, Lübeck, Germany). This was done by punching a hole through a male-female circuit adaptor (Westmed, Inc., Tucson, AZ) just large enough to thread the CO2 sampling line through it. This adaptor was then placed between the bellows canister and the circuit so that the sampling line was positioned inside the bellows (Fig. 1). The sampling line was connected to a capnometer (Poet IQ; Criticare Systems, Inc.) so gasses from the bellows could be measured. All subjects underwent general endotracheal anesthesia, volume controlled ventilation, and had radial arterial catheters. After induction, fresh gas flow was set at 2 L oxygen and VT values between 8 and 10 mL/kg. Inspiratory-to-expiratory ratio was constant at 1:2. Measurement of PECO2 and PaCO2 (via the arterial catheter) was taken 10 minutes after induction, which we refer to as baseline. VT was recorded from the ventilator's display. Dead space was calculated using the Bohr equation (alveolar dead space [VDS]/VT = [PaCO2 − PECO2]/PaCO2). Next, an additional 100 mL of apparatus dead space was added to the circuit by connecting a portion of a breathing circuit (96-in. expandable adult circuit, #7-8103-10-68; Westmed, Inc.) with adaptors (Westmed, Inc.) between the endotracheal tube and the “Y” piece of the circuit. The volume of the added circuit was determined by measuring the amount of water it could contain. After 10 minutes of equilibration, PECO2 and PaCO2 were measured, VT was recorded, and the patient's dead space was calculated. Finally 200 mL of apparatus dead space was added to the circuit and after 10 minutes of equilibration time, PECO2 and PaCO2 were measured. VT was recorded and dead space was calculated.

Figure 1
Figure 1:
Carbon dioxide sampling line placed into the ventilator's bellows via a male-female adaptor.


Ten subjects took part in this study, 9 women and 1 man. The mean age was 53 ± 13 years (range, 40–79 years). The mean height was 167.6 ± 13.9 cm, weight 83.2 ± 16.3 kg, and body mass index 30.4 ± 6.7 kg/m2. Procedures included open hysterectomies, 1 laparoscopic hysterectomy, 1 laparoscopic prostatectomy, and 1 lumbar posterior fusion. Table 1 summarizes the results. Figure 2 shows calculated dead space for all 10 subjects starting at baseline, then increasing as we increased the dead space. The mean baseline dead space was 265 ± 47 mL (mean ± SD). After the addition of 100 mL of known dead space, the measured increase in dead space was 110 ± 46 mL. After the addition of 200 mL of known dead space, the measured increase was 158 ± 39 mL. Paired Student t tests showed that there was a statistically significant difference between the groups (P = 0.000).

Table 1
Table 1:
Means and Standard Deviations for All 10 Subjects
Figure 2
Figure 2:
Dead space volume on the y-axis, plotted against baseline measurement of dead space (VDc), dead space measurement after adding 100 mL of dead space (VD 100), and after adding 200 mL of dead space (VD 200).


Baseline dead space measurements were 265 ± 47 mL, which is expected in adults under general anesthesia. When a known amount of dead space was added, the calculated dead space increased in every subject. Although this demonstrates that a change of at least 100 mL in dead space can be detected, the sensitivity to smaller changes in dead space was not examined. However, this ability to detect a change in dead space may alert the provider of a possible pulmonary embolism or decrease in cardiac output.

It is interesting that added apparatus dead space was more accurately measured when 100 mL was added than when 200 mL was added. In the latter case, the amount of added dead space tended to be severely underestimated (Fig. 2). We can think of 2 explanations. Perhaps we did not allow enough time for equalization of CO2 in the bellows. Equalization time may not be as simple as circuit size versus flow rates. Ten minutes of equilibration time may not be adequate when dead space changes dramatically. Another explanation for the underestimation is that dead space gas may preferentially exit the breathing system via the ventilator relief valve with the introduction of a higher apparatus dead space. In a semiclosed circle breathing system, the exhaled gas traverses the expiratory tubing and enters the ventilator bellows; when the bellows reaches its capacity, excess gas exits through the ventilator relief valve. The ventilator relief valve opens at the end of exhalation, but the composition of the gas exiting the valve depends, in part, on the length of the expiratory tubing. In our method, dead space is underestimated if dead space gas preferentially exits the valve (which would explain why the response to adding dead space is nonlinear), and it is overestimated if alveolar gas preferentially exits the valve. It would be interesting to determine whether this was a source of error by investigating the effects of tubing length, VT, and fresh gas flow rate on the CO2 concentration of the gas exiting the ventilator relief valve. This is an area for future research. In addition, if the subject's dead space changed between measurements, this would also account for error; however, we believe that it is reasonable to assume that the subject's dead space did not significantly change as we added apparatus dead space. However, we did not assess the amount that PECO2 or dead space changes in the stable subject or during changes in cardiac output.

We present a simple way to detect trends in dead space in ventilated patients. Although much research is needed in this area, this technique could be used on the Narkomed GS anesthesia machine in operating rooms today.


Name: John J. Badal, MD

Contribution: Study design, data collection, data analysis, manuscript preparation.

Attestation: I am the archival author, and I have reviewed the original data and analysis.

Name: Kyung J. Chen, MD

Contribution: Data collection, manuscript preparation.

Name: Robert G. Loeb, MD

Contribution: Study design, data analysis.

Attestation: I have reviewed the original data and analysis.


1. Soro M, Garcia-Perez ML, Belda FJ, Ferrandis R, Aguilar G, Tusman G, Gramuntell F. Effects of prone position on alveolar dead space and gas exchange during general anaesthesia in surgery of long duration. Eur J Anaesthesiol 2007;24:431–7
2. Kline JA, Kubin AK, Patel MM, Easton EJ, Seupal PA. Alveolar dead space as a predictor of pulmonary embolism. Acad Emerg Med 2000;7:611–7
3. Hardman JG, Aitkenhead AR. Estimating alveolar dead space from the arterial to end-tidal CO2 gradient: a modeling analysis. Anesth Analg 2003;97:1846–51
4. Rosenbaum A, Breen PH. Novel, adjustable, clinical bymixer measures mixed expired gas concentrations in anesthesia circle circuit. Anesth Analg 2003;97:1414–20
5. Badal JJ, Loeb RG, Trujillo DK. A simple method to determine mixed exhaled CO2 using a standard circle breathing circuit. Anesth Analg 2007;105:1048–52
© 2011 International Anesthesia Research Society