Prone position is used during general anaesthesia for posterior spinal surgery. This position has influence on lung volumes, ventilation and lung perfusion. In obese patients and patients with acute lung injury, prone position increases functional residual capacity (FRC)  and improves oxygenation [2,3]. In healthy subjects, increased oxygenation has been related to an increase in FRC , more homogeneous lung perfusion  and reduction of ventilation/perfusion mismatching . However, no conclusive data have yet been reported regarding carbon dioxide (CO2) exchange in prone position. In one series of healthy patients with constant minute ventilation, the arterial carbon dioxide tension (PaCO2) did not change . In two other groups of patients, the gradient between PaCO2 and end-tidal PCO2 (Pa-etCO2) increased in the prone position, indicating an increase in the physiological dead space (VD) [6,7]. In each case, the measurements were made 15–20 min after turning the patients without later controls. There is a paucity of data on CO2 exchange in the prone position over an extended period of time.
In healthy patients anaesthetized in prone position, a reduction of compliance of the respiratory system has been noted [8,9], the magnitude of which seems to be directly linked to the surgical frame used to support the patient . In fact, when the support allows freedom of movements of the abdominal wall, the total compliance of lung and chest wall is unchanged . In obese patients, even though the total compliance does not change, an improvement in lung compliance is observed as a result of recruitment of atelectatic segments .
The aim of the present study was to investigate changes in alveolar dead space and oxygenation during general anaesthesia in prone position in spinal surgical procedures of long duration (more than 3 h).
The study was approved by the Institutional Ethics Committee and all the patients gave their written informed consent.
Fourteen ASA I-II patients (age 18–70 yr) with various spinal pathologies, scheduled for posterior spinal surgery in prone position with an estimated duration of more than 3 h, were studied prospectively. Exclusion criteria were pregnancy, morbid obesity (body mass index) BMI > 30 kg m−2 or history of cardiorespiratory disease that might affect the distribution of ventilation and perfusion.
All patients were premedicated with intravenous (i.v.) midazolam 0.02 mg kg−1 and fentanyl 0.15–0.20 mg. Anaesthesia was induced with propofol 2–3 mg kg−1 i.v. Muscle relaxation was achieved with succinylcholine 100 mg or cisatracurium 0.1 mg kg−1 i.v. The trachea was intubated with a reinforced tube (internal diameter 7.5–8.0 mm). Anaesthesia was maintained with continuous infusions of propofol (6–12 mg kg−1 h−1) and remifentanil (0.2–0.3 μg kg−1 min−1) as needed to maintain the bispectral index (BIS) between 40% and 60%, and heart rate (HR) and arterial blood pressure within 20% limits of the preinduction values. Muscle paralysis was maintained with an infusion of cisatracurium 0.1 mg kg−1 h−1.
Monitoring included electrocardiography, invasive arterial pressure (radial artery), pulse oximetry, and rectal or oesophageal temperature (Datascope Passport, Datascope Corp., Mahwa, USA). Depth of anaesthesia was monitored by means of BIS (BIS monitor, Aspect Medical Systems, Natick, USA), muscle relaxation by means of accelerometry (Tof-Watch, Organon Teknika BV, the Netherlands). Blood loss and urine production were measured hourly.
The patients were ventilated in a volume-controlled mode with a frequency of 12–14 breaths min−1 and a tidal volume (VT) of 8–10 mL kg–1. The aim was to maintain PaCO2 at 4–4.6 kPa. We used a non-rebreathing anaesthesia ventilator (Ergotronic, Temel SATM, Spain) that has a compressible volume (internal compliance) of less than 3 mL cm H2O−1 when measured with a standard external circuit. The I/E ratio was fixed at 1/2, resulting in an inspiratory time (TI) of 1.43–1.67 s. The inspiratory flow was adjusted to guarantee an inspiratory pause time of 30% of the TI (0.5 s) so that the plateau pressure (Ppt) would represent an estimate of the alveolar pressure at the end of inspiration. The expiratory time (TE) was long enough to permit complete emptying of the lung, avoiding intrinsic positive end-expiratory pressure (PEEP), as demonstrated in all the patients by a zero flow at end expiration. All patients were ventilated without PEEP, as previous studies have shown an increase in FRC and improvement in PaO2 with the change of supine to prone position [1–3]. FiO2 was initially set at 0.4 and maintained throughout the procedure in all patients. Small readjustments in VT were accepted in order to maintain the target PaCO2 of 4–4.6 kPa. Respiratory parameters were kept constant afterwards throughout the study and no lung-recruiting manoeuvres were performed.
In order to maintain body temperature during surgery, all patients were covered with forced–air warming systems (Warmtouch, Mallinckrodt MedicalTM, Ireland, EU), and all intraoperative fluids were delivered via a Hotline fluid warmer (Level 1 Technologies Inc., Rockland, MA, USA). A crystalloid solution was infused at a rate of about 8 mL kg−1 h−1, titrated to maintain CO above 3.5 L min−1, after replacing fluid loss caused by fasting (1.5–2 mL kg−1 h−1 of fasting). When necessary, blood loss was compensated by infusion of colloid solutions in equal amounts.
After the induction of anaesthesia, the patients were placed in prone position on the corresponding thoracic–pelvic support . In all patients, a standard orthopaedic frame (Wilson Frame, OSITM, Union City, CA, USA) was used, placed on the surgical table and separating the lateral supports to guarantee freedom of abdominal movement.
Simultaneous measurements of respiratory volumes and CO2 were made using an automated volumetric capnograph and pulmonary mechanics monitor (NICO, RespironicsTM, Wallingford, CT, USA). This non-invasive monitor consists of a mainstream capnometer, a variable orifice pneumotachometer, a signal processor, and computer software with capability for both on- and off-line data analysis. The CO2 signal is provided by a mainstream, non-dispersive, infrared capnometer complete with an analogue output module. The pneumotachometer is a disposable, variable orifice, differential pressure device. The volumetric capnography or single breath CO2 curve displayed by the monitor (Fig. 1) is a dynamic curve in which expired CO2 (y-axis) is plotted against expired volume (x-axis). It is divided into three distinct phases, as first described by Fowler . Phase 1 represents expired gas from the conducting airways, which contains no measurable CO2. Phase 2 represents the mixing of the terminal gas from conducting airways and alveolar gas from acini with the shortest transit times. Phase 3 represents gas from the alveoli and includes the alveolar plateau. The physiological, airway and alveolar dead spaces are calculated cycle by cycle from this curve by the NICO monitor, provided that an arterial PCO2 from blood gas analysis is introduced. The details of the waveform analysis and the potential sources of error of the method are described in the work of Arnold and colleagues . The NICO monitor also reliably estimates cardiac output (CO) by application of the Fick principle through a CO2 rebreathing period [13–15]. The flow pneumotachometer was calibrated with a metred 100-mL syringe (Hans RudolphTM, Kansas City, MO, USA) and the CO2 analyser was calibrated to a standard calibration gas (5% CO2).
For the measurement of arterial gases a previously calibrated blood gas analyser (OSM3, RadiometerTM, Denmark) was used. These calibrations were repeated at the beginning of the experimental protocol for each patient.
Measurements were carried out at four time points: in supine position, 20 min after beginning mechanical ventilation (supine), and in prone position at 30, 120 and 180 min after positioning. HR, systemic arterial blood pressure and CO were recorded, and arterial blood gases were taken for PaCO2 and PaO2 measurements. The values of PaCO2 were introduced into the NICO as indicated above. The recorded values were airway dead space, VDaw (mL); physiological dead space, VD (mL); tidal volume, VT (mL); alveolar tidal volume, VTalv (mL); alveolar dead space, VDalv (mL); VD/VT ratio (%) and VDalv/VTalv ratio (%). In order to avoid any interference of rebreathing on dead space measurements, CO was always measured after dead space recordings and not less than 30 min before the next dead space measurements.
For the comparative analysis of the results within the groups, an analysis of variance (ANOVA) test for repeated measures followed by Wilcoxon's signed rank sum test was used, being considered statistically significant when P < 0.05. Bonferroni's test was applied for multiple comparisons. In the text and the tables, the values are expressed as mean with standard deviation (SD). The data were processed with the statistical package SPSS 13.0 (SPSS Inc.TM, Chicago MC, 2004).
There were 8 females and 6 males with mean ± SD age 45 ± 6 yr, height 165 ± 6 cm and weight 64 ± 8 kg. The diagnoses were scoliosis 6 patients, lumbar spine degeneration 3 patients, vertebral metastasis 2 patients and cervical instability 3 patients. The results of the respiratory changes are presented in Table 1. Tidal volume was reduced in 4 patients after turning to prone position in order to maintain the target PaCO2 of 4–4.6 kPa. The VD/VT ratio and the alveolar dead space/VT ratio did not change. A statistically non-significant increase in the PaO2/FiO2 ratio was observed in the prone position. Total dynamic compliance was significantly reduced after turning the patient, without further variations during maintenance of prone position. This was due to a decrease in tidal volume and an increase in plateau pressure.
Haemodynamics and rectal temperature remained stable in all patients during the study period without statistically significant changes (Table 1).
We have used the volumetric capnography technique to demonstrate that the physiological and alveolar dead spaces had not changed in anaesthetized and paralysed patients after 3 h in the prone position. On the other hand, oxygenation tended to improve without use of PEEP or recruitment manoeuvres.
Volumetric capnography for measurement of airway and VD was introduced by Fowler  and has since been validated in anaesthesia under various conditions of mechanical ventilation [16,17]. This technique, with a monitor similar to ours, was further investigated by Arnold and colleagues  in an experimental lung model, in which tube segments were added to simulate gradual increases in airway dead space. They found that the measured airway dead space correlated significantly with the actual circuit dead space (r2 = 0.99). In the same study, the VD/VT obtained with volumetric capnography in sheep was compared with values calculated using the Bohr–Enghoff equation. The correlation was good (r2 = 0.84) with a mean percent difference of 2.4% between the two methods. Kallet and colleagues  showed that VD/VT measured by volumetric capnography had a precision of 0.05 and a bias of 0.02, compared with VD/VT measured with a metabolic monitor technique and corrected for estimated compression volume. This grade of accuracy was very reasonable for the use of the monitor in clinical studies . In fact, volumetric capnography has been validated as a reliable method of measuring VD/VT during mechanical ventilation [19,20].
Potential sources of error in online dead space measurements were described by Fletcher and colleagues . The two most important ones are phase delay between capnometry and pneumotachography, which is related to the flow rate, and release of compressed gas during expiration. The response delay of CO2 measurement was insignificant in our study because a mainstream device with a sampling rate of 87 Hz was used. The flow signal in this monitor precedes the CO2 signal by a maximum of 10 ms (at peak flow at the onset of expiration) to 30 ms (at end-expiration). In this way the error in dead space calculations over a wide range of VT was estimated to be 2–5 mL (0.2–0.9%), which we regard as negligible . In connection with the potential influence of compressed gas, NICO measures expired CO2 at the Y-adapter of the ventilator circuit, thus eliminating this source of error.
In our patients, the relative magnitude of airway and VD in the supine position were similar to values previously obtained with the volumetric capnography method in patients mechanically ventilated with similar tidal volumes and anaesthetized with a comparable total i.v. anaesthetic technique [17,21]. After turning to the prone position, no significant changes were observed in PaCO2, Pa-etCO2, or physiological and alveolar dead space values during the study period (180 min).
Wahba and colleagues have previously reported  that PCO2 did not vary but that end-tidal CO2 decreased, producing a significant increase in Pa-etCO2 (from 0.49 ± 0.13 to 0.78 ± 0.01 kPa), after 15 min of prone position in a group of 20 anaesthetized patients. Similar results were found by Casati and colleagues  in a group of 24 anaesthetized patients after 20 min in the prone position. They observed an approximately 10% increase in VD/VT (calculated using a modification of the Enghoff equation) and an increase in Pa-etCO2 compared with a control group that was kept supine (0.52 ± 0.27 vs. 0.82 ± 0.27 kPa). These studies are in contrast to our findings. Both Wahba and Casati attribute the observed increases in dead space ventilation in the prone position to a presumed influence on the intrapulmonary distribution of blood flow and alveolar gas. However, it cannot be disregarded that these relatively small increases in VD/VT could be due to reduced lung perfusion secondary to a fall in CO after turning to prone position. Moreover, in both studies anaesthesia was maintained with isoflurane, which has effects on pulmonary vascular regulation and bronchial muscle tone that have recently been claimed to be an important mechanism for increases in VDalv and VDaw observed in supine general surgical patients . Finally, we cannot exclude that the different technique of dead space measurements in these previous studies may explain the contrasting results compared with ours.
Our results appear to be more consistent with the increased FRC in the prone position that has been observed in awake healthy volunteers  and in the anaesthetized patient [2,3]. Prone position with a thoracic–pelvic support frame allows free expansion of the abdomen. This shifts the chest wall elastic recoil curve to the left, resulting in an equilibrium position of the respiratory system at a greater lung volume than in supine position.
With surgery of long duration as in our patients, an increase in FRC after turning to the prone position may also be attributed to a recruitment of dependent lung areas, where atelectasis have been shown to appear shortly after anaesthetic induction and muscular relaxation . With constant VT, ventilation of the previously atelectatic lung areas would result in a redistribution of the alveolar VT, increasing ventilation of poorly ventilated-dependent areas (low V/Q) at the expense of relatively underperfused areas (high V/Q). Moreover, when turning to prone position, the gravitational gradient of blood flow is reduced  and lung perfusion is more uniform . Both factors would generate increased homogeneity in the ventilation/perfusion relationship, leading to better oxygenation and maintained dead space in spite of a decrease in VT. Even in the absence of changes in FRC, the better homogeneity of V/Q in prone position could explain the significant increase in PaO2 observed by several investigators in various patient groups [25–27].
In our study, CO was stable throughout the study period and would not have influenced VD. The high mean PaO2/FiO2 in our patients indicates a very low shunt, thus minimizing its effect on VD/VT measurements with the volumetric capnography method .
After turning to prone position, we observed a reduction in dynamic compliance of the respiratory system, which was probably attributable to compression of the chest wall by the frame used. We also found a rise in airway pressure and the subsequent increased gas compression in the ventilator circuit resulted in a slight but significant reduction in VT.
In conclusion, our study demonstrates that patients undergoing surgery in prone position for a duration of 3 h under general anaesthesia, muscle relaxation and mechanical ventilation without PEEP have stable haemodynamics and no significant changes in alveolar dead space ventilation. There was a non-significant trend towards improved oxygenation.
1. Lumb AB, Nunn JF. Respiratory function and rib cage contribution to ventilation in body positions commonly used during anesthesia. Anesth Analg
2. Pelosi P, Massimo C, Calappi E et al
. The prone positioning during general anesthesia minimally affects respiratory mechanics while improving functional residual capacity and increasing oxygen tension. Anesth Analg
3. Pelosi P, Croci M, Calappi E et al
. Prone positioning improves function in obese patients during general anesthesia. Anesth Analg
4. Nyren S, Mure M, Jacobsson H, Larsson SA, Lindahl SG. Pulmonary perfusion is more uniform in the prone than in the supine position: scintigraphy in healthy humans. J Appl Physiol
5. Mure M, Domino KB, Lindahl SG, Hlastala MP, Altemeier WA, Glenny RW. Regional ventilation-perfusion distribution is more uniform in the prone position
. J Appl Physiol
6. Casati A, Salvo I, Torri G, Calderini E. Arterial to end-tidal carbon dioxide gradient and physiological dead space monitoring during general anaesthesia: effects of patients' position. Minerva Anestesiol
7. Wahba R, Tessler M, Kardash K. Carbon dioxide tensions during anesthesia in the prone position
. Anesth Analg
8. Lynch S, Brand L, Levy A. Changes in lung-thorax compliance during orthopedic surgery
9. Safar P, Aguto-Escarraga L. Compliance in apneic anesthetized adults. Anesthesiology
10. Palmon S, Kirsch J, Depper J, Toung T. The effect of the prone position
on pulmonary mechanics is frame-dependent. Anesth Analg
11. Smith RH. One solution to the problem of the prone position
for surgical procedures. Anesth Analg
12. Fowler WS. Lung function studies. V. Respiratory dead space
in old age and in pulmonary emphysema. J Clin Invest
13. Arnold JH, Thompson JE, Arnold LW. Single breath CO2
analysis: description and validation of a method. Crit Care Med
14. Rocco M, Spadetta G, Morelli A et al
. A comparative evaluation of thermodilution and partial CO2
rebreathing techniques for cardiac output assessment in critically ill patients during assisted ventilation. Intensive Care Med
15. Jover JL, Soro M, Belda FJ, Aguilar G, Caro P, Ferrandis R. Measurement of cardiac output after cardiac surgery
: validation of a partial carbon dioxide rebreathing (NICO) system in comparison with continuous thermodilution with a pulmonary artery catheter. Rev Esp Anestesiol Reanim
: 256–262 (Spanish).
16 Fletcher R, Jonson B. Deadspace and the singlebreath test for carbon dioxide during anaesthesia and artificial ventilation. Effects of tidal volume and frequency of respiration. Br J Anaesth
17. Taskar V, John J, Larsson A, Wetterberg T, Jonsson B. Dynamics of carbon dioxide elimination following ventilator resetting. Chest
18. LaMantia KR, O'Connor T, Barash PG. Comparing methods of measurement: An alternative approach. Anesthesiology
19. Blanch L, Lucangelo U, Lopez-Aguilar J, Fernandez R, Romero PV. Volumetric capnography in patients with acute lung injury: effects of positive end-expiratory pressure. Eur Respir J
20. Kallet RH, Daniel BM, Garcia O, Matthay MA. Accuracy of physiologic dead space measurements in patients with acute respiratory distress syndrome using volumetric capnography: comparison with the metabolic monitor method. Respir Care
21. Praetel C, Banner MJ, Monk T, Gabrielli A. Isoflurane inhalation enhances increased physiologic deadspace volume associated with positive pressure ventilation and compromises arterial oxygenation. Anesth Analg
22. Mahajan RP, Hennessy N, Aitkenhead AR, Jellinek D. Effect of three different surgical prone positions on lung volumes in healthy volunteers. Anaesthesia
23. Hedenstierna G, Rothen HU. Atelectasis formation during anesthesia: causes and measures to prevent it. J Clin Monit Comput
24. Walther SM, Domino KB, Glenny RW, Hlastala MP. Positive end-expiratory pressure redistributes perfusion to dependent lung regions in supine but not in prone lambs. Crit Care Med
25. Pelosi P, Croci M, Ravagnan I et al
. The effects of body mass on lung volumes, respiratory mechanics and gas exchange during general anesthesia. Anesth Analg
26. Mure M, Lindahl SG. Prone position
improves gas exchange – but how? Acta Anaesthesiol Scand
27. Radstrom M, Loswick AC, Bengtsson JP. Respiratory effects of the kneeling prone position
for low back surgery
. Eur J Anaesthesiol
28. Tang Y, Turner MJ, Baker AB. Effects of alveolar dead-space, shunt and V/Q distribution on respiratory dead-space measurements. Br J Anaesth