1 Department of Surgical Sciences, Section of Anaesthesiology and Intensive Care Medicine, Uppsala University, Sweden. 2 Department of Anaesthesiology and Surgical Intensive Care Medicine, Klinikum Augsburg, Germany. 3 Department of Medical Sciences, Section of Clinical Physiology, Uppsala University, Sweden. 4 Department of General and Visceral Surgery, Klinikum Augsburg, Germany. 5 Department of Anaesthesiology and Critical Care Medicine, Freiburg University, Germany
Corresponding author. E-mail: m.lichtwarck-aschoff@t-online.de
Accepted for publication: April 9, 2003
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Methods. Calculated Ptr was compared with measured Ptr during controlled ventilation and assisted spontaneous breathing in 18 healthy and surfactant-depleted piglets. Their lungs were ventilated using different flow patterns, tidal volumes (VT) and levels of positive end-expiratory pressure.
Results. In terms of the root mean square error (RMS), indicating the average deviation of calculated from measured Ptr, the difference between calculated and measured Ptr was 0.6 cm H2O (95%CI 0.580.65) for volume-controlled ventilation; 0.73 cm H2O (0.720.75) for pressure support ventilation; and 0.78 cm H2O (0.750.80) for bi-level positive airway pressure ventilation.
Conclusion. The good agreement between calculated and measured Ptr during varying conditions, suggests that calculating Ptr could help setting the ventilator and choosing the appropriate level of support.
Br J Anaesth 2003; 91: 23948
Keywords: lung; lung, tracheal pressure; ventilation, controlled mechanical; ventilation, spontaneous
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Study plan
One group of piglets (n=12) was allocated to volume-control ventilation (VC) before and after broncho-alveolar lavage, to study whether lavage-induced changes of the impedance of the respiratory system would affect the reliability of the calculated Ptr. We later added a second study with assisted spontaneous breathing. We felt that in clinical practice, tracheal pressure is more relevant with assisted ventilation than with VC, and that the complex flow patterns of assisted modes would be a more vigorous test of the value of the Ptr algorithm than the constant inspiratory flow of VC. Three healthy piglets therefore received pressure support ventilation (PS), and another three received bi-level positive airway pressure ventilation (BiPAP). Ventilation was given for 20 min before measurements were performed. The relationship between calculated and measured Ptr was assessed.
Tracheal pressure was calculated using two different methods: according to Guttmann and colleagues,4 henceforth called calculated PRohrer and according to Jarreau and colleagues,2 henceforth called calculated PBlasiusIto (see Appendix). When referring to calculated Ptr in general, irrespective of the method to calculate it, and emphasizing the contrast to measuring Ptr, the simpler term calculated Ptr is used.
Ventilator settings
In the VC group, tidal volumes (VT) of 150, 250, 300, 400, and 500 ml and either zero PEEP (healthy animals; group labelled ZEEP), or 12 cm H2O PEEP (surfactant-deficient animals; group labelled PEEP) were consecutively applied. Ventilator frequency was 20 min1. In the PS group (healthy animals; PS), PEEP was set to 5 cm H2O and the pressure support to 5 cm H2O above PEEP level.
In the BiPAP group (healthy animals; BiPAP) the settings were: PEEP at 5 cm H2O for 3 s; upper pressure level 10 cm H2O for 3 s; set frequency 6 min1. FIO2 was 0.5 throughout. In the PS and the BiPAP group, trigger was set to the maximal flow-trigger sensitivity.
Respiratory measurements
The animals lungs were ventilated through a tracheal tube (#7, Mallinckrodt, Athlone, Ireland), connected to a Servo 300 ventilator (Siemens-Elema, Solna, Sweden) by a 60 cm rigid tubing system. In the PS and BiPAP group, airway pressure and flow were continuously measured with a Fleisch pneumotachograph (Pneumotachometer Series 1110; Hans Rudolph, Kansas City, MO) placed between the tracheal tube and the Y-piece of the ventilator circuit. In the VC group, flow and airway pressure were measured with the Servo 300 built-in transducers. A comparison between flow/pressure data simultaneously recorded with the Fleisch head and the Servo 300 transducers showed excellent agreement during VC. Tracheal pressure was directly measured 23 cm above the carina with a catheter tip transducer (OLM intracranial pressure monitoring kit, model 110-4B, Integra Neurosciences Camino, San Diego, CA), positioned outside the tracheal tube under direct inspection after the tracheostomy, before placing the tracheal tube. The Camino transducer measures pressure at the tip of the catheter, in an axial direction. End-inspiratory and end-expiratory hold manoeuvres (5 s) were also performed. Assuming that pressure would equilibrate within 5 s, the pressure values at the end of a 5-s hold were used to correct the measured Ptr data if necessary. If after a 5-s hold a pressure difference was detected, this difference was treated like an offset, and it was subtracted from the pressure readings of the Camino sensor.
All signals were sampled at 200 Hz, passed to a data acquisition system (AcqKnowledge, version 3.7.2, BioPac System, Inc., Santa Barbara, CA), and then exported to an MS-Excel worksheet for off-line analysis. At each measurement period, data were continuously sampled for 2 min. The measured Ptr was compared off-line to the calculated Ptr using the pressureflow data acquired at the proximal outlet of the tracheal tube.
Anaesthesia and fluid management
After pre-medication (tiletamine 3 mg kg1, zolazepam 3 mg kg1, xylazine 2.2 mg kg1, atropine 0.04 mg kg1 intramuscularly), anaesthesia was induced with 500 mg ketamine, 0.5 mg atropine, and 20 mg morphine i.v.. Anaesthesia was maintained with ketamine infusion (20 mg kg1 h1) and morphine 0.5 mg kg1 h1, and neuromuscular block obtained by continuous infusion of pancuronium bromide (0.25 mg kg1 h1). The animals were given a solution of 4.5 g litre1 NaCl with 25 g litre1 glucose (Rehydrex, Pharmacia Infusion AB, Uppsala, Sweden) at 10 mg kg1 h1 and a 5 mg kg1 bolus of dextran-70 (Macrodex 70, Pharmacia Infusion AB, Uppsala, Sweden) to ensure normovolaemia.
Lavage
Lavage was done with 11 broncho-alveolar lavages (1.21.5 litre normal saline, corresponding to 5060 ml kg1) which resulted in a PaO2/FIO2 of less than 20 kPa. After lavage, the animals were allowed to stabilize for 20 min.
Re-expansion
Immediately after lavage the lungs were re-expanded by a 5-min period of pressure-controlled ventilation (PEEP 25 cm H2O and a peak inspiratory airway pressure of 50 cm H2O).
Data presentation
Results are presented as mean and 95% CI (95% lower to 95% upper) if not otherwise indicated. The agreement between measured and calculated Ptr was assessed in terms of the root mean square error (RMS), giving the mean deviation for all data points under consideration. Bias and limits of agreement for distinct single pressure points were assessed according to the Bland and Altman method.3 The data were defined as follows. During VC: point 1=0.1 s past start of inspiration; point 2=peak pressure; point 3=0.25 s past peak pressure; point 4=0.5 s past peak pressure; point 5=1 s past peak pressure; point 6=last end-expiratory point. During BiPAP: point 1=0.2 s past start of inspiration; point 2=peak pressure; point 3=0.2 s past peak pressure; point 4=0.15 s past start of expiration; point 5=0.25 s past start of expiration: point 6=last end-expiratory point. During PS: point 1=0.10 s past start of inspiration; point 2=peak pressure; point 3=0.10 s before end of inspiration; point 4=0.15 s past start of expiration; point 5=0.25 s past start of expiration; point 6=last end-expiratory point. Statistical significance was assumed with P0.05.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Comparison of two methods to calculate tracheal pressure
The average deviation of PRohrer4 from PBlasiusIto2 was 0.3 cm H2O (95% CI 0.150.4) over the whole respiratory cycle. Agreement was equally good for inspiration and expiration and for VC, PS, and BiPAP, respectively (see Figs 123).
|
|
|
BlandAltman for PRohrer (see Fig. 4 and Tables 25)
The comparison of single pressure points indicated good overall agreement, with the most bias during VC with ZEEP for the data point 0.25 s past peak in expiration (0.63 cm H2O (0.75 to 0.49)) and the smallest bias for the pressure point 0.5 s past peak pressure (0.07 cm H2O (0.150.01)). The situation was similar for VC with PEEP 12. The limits of agreement for VC (both without and with PEEP) in the worst case ranged from 1.23 to 1.53 cm H2O (peak inspiratory pressure, VC ZEEP) and in the best case from 0.81 to 0.79 cm H2O (1 s past peak pressure, VC 12 PEEP). Figure 4 shows BlandAltman plots for six selected pressure points. For the spontaneous modes, bias and limits of agreement were greater for all pressure points under consideration with a maximal bias of 1.31 cm H2O (1.271.34) (0.1 s after start of inspiration during pressure support).
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Methodological considerations
Some methodological problems could not be solved to our satisfaction.
First, the transducer for measuring Ptr is not specifically designed for that purpose. The device is designed for measuring intracranial pressure and has a temperature-related drift,5 6 which in these animals, whose temperature was 37.5°C at start and sometimes became 41°C, could be important (the manufacturers manual indicates a temperature coefficient of 3 mm Hg over temperature range 2238°C).
As our concern was not the reliability of direct Ptr measurement, we did not systematically screen all the measured Ptr data for malposition and potential artefacts to determine the absolute number of erroneous measurements to which our current equipment is obviously prone. Despite the basic methodological concerns with direct Ptr measurement discussed below, we know, that a dedicated tool for measuring Ptr7 would probably have given better-measured Ptr data.
The second concern was with the bias flow of the Servoventilator 300. Flow-triggering requires this bias flow of 40 ml s1. The algorithm does not discriminate between the main flow and the bias flow, and the sum of both is included in the calculation of PRohrer. During early inspiration, flow is high (>1500 ml s1) with PS or BiPAP, and the error introduced by the bias flow is relatively small. However, at end-expiration the flow rate is getting small and the relation between the bias flow and the entire flow increases. This can explain the weak agreement of measured and calculated Ptr at some of the pressure points. As far as device-specific methods for regulating flow on the calculation of Ptr can be taken into account, the present results can be extended to other types of ventilators, too.
Cardiogenic oscillations were another disturbing factor. Measurement of Ptr was performed at the carina close to the beating heart. Calculation of Ptr, in contrast, uses pressure and flow data sampled at the proximal tracheal tube outlet, distant from the beating heart with the dampening effect of airways and tracheal tube interposed. Cardiogenic oscillation affects the measured Ptr more than the calculated Ptr, and an association between measured and calculated Ptr data may be considered at the very moment when a heartbeat distorts the measured Ptr. At low lung volumes the heartbeats influence the measured Ptr even more strongly, which can explain the weak association of some of the pressure points at end-expiration.
The combination of two studies, one with VC and an additional one with assisted spontaneous breathing, PS and BiPAP, comprising different numbers of animals is certainly open to statistical criticism. We justify this with our intention to test the agreement between calculated and measured Ptr under an extended spectrum of flow patterns.
Finally, this is a short-term study only. A clinical validation would have to take long-term effects like progressive tracheal tube narrowing by secretions into account. An estimation of the potential error caused by tracheal tube narrowing is therefore added (see below).
Can direct measurement of tracheal pressure be regarded as a gold standard at present?
Apart from the above, direct measurement of Ptr has drawbacks that make it difficult to regard it as a true gold standard. Thin measuring catheters are easily obstructed, and thick catheters affect flow pattern and increase the tracheal tube resistance. (To avoid the latter, the catheter in this study was placed outside the tracheal tube, whereas in all previous studies that we know of the Ptr catheter was placed inside the ETT.) Catheters that measure Ptr at the tip of the catheter in an axial direction (like the one used in the present study) overestimate the expiratory pressure because of the Bernoulli effect:8 With expiratory gas flowing towards the measuring site of the catheter, there is an additional dynamic pressure component that depends on the square of flow velocity. Moreover, the cross-sectional area changes abruptly at the transition of the tracheal tube to the trachea. Here kinetic energy dissipates by flow separation, leading to additional pressure loss.8 9 As a consequence, Ptr should be measured where the gas flow is fully developed. This requires the catheter to be advanced at least 3 cm beyond the tip of the tracheal tube.4 Positioning the measuring site radial to the flow direction (using catheters with a lateral opening) also has drawbacks because the velocity profile is inhomogeneousit is parabolic when flow is laminar, and it is flat when transitional flow is present. Apart from those basic methodological concerns it is not easy to position the catheter correctly, and one might easily find the catheter stuck against the tracheal wall or obstructed by secretions. We also observed that the position of the catheter could change with lung volume (that is, with tidal volume and/or with PEEP).
Different methods to calculate tracheal pressure
Different methods have been proposed for calculating Ptr.2 4 813 From a theoretical standpoint it is certainly more satisfying to take all relevant variables (gas physical properties, the inner geometry of the ETT and its curvature) into account2 rather than limiting the analysis to the mathematical description of empirical pressureflow curves using the Rohrer coefficients (depending on length and inner diameter (ID) of the native tracheal tube measured ex-vivo).4 14 When it comes to the bedside, however, it is cumbersome, if not completely impossible, to determine the curvature of the tracheal tube or the changes in gas physical properties oscillating within the respiratory cycle. Jarreau and colleagues2 have suggested dimensional forms of the more complex equations given in their paper for practical estimation of the pressure decrease over the tracheal tube. In the dimensional form of the equations only variables available at the bedside (flow, length, and ID of the native tracheal tube) are used and whether flow is turbulent or not is estimated from a rough approximation. Using this approach we found an excellent agreement between Ptr calculated with the method suggested by Guttmann and colleagues4 and the method suggested by Jarreau and colleagues2 (see Figs 13 and Appendix). Changing FIO2 and, hence, the density of the gas, affects resistance, which is particularly important in paediatric tubes,15 but for clinical purposes in adults the effect is negligible.4 While the gas density of room air is about 89% of the density of pure oxygen, density of helium is 12% of pure oxygen. Thus, the use of gases with major difference in density would have to be taken into account when calculating Ptr.
How large are potential errors when tracheal pressure is calculated?
All the methods for calculating Ptr depend on knowledge of the true actual ID of the tracheal tube. If secretions progressively narrow the tracheal tube while the algorithm continues to assume a native, unobstructed tracheal tube an error will result. The narrowing increases the flow-dependent resistive pressure decrease over the tracheal tube (PETT), and the calculated Ptr now overestimates the true Ptr. The potential error is estimated in Figure 5. The
PETT of the native tracheal tube with ID 8 is plotted for increasing flow rates. An assumed narrowing from ID 8 to 7.5 mm increases the tracheal tube-related resistance by about 25%, shifting the pressureflow curve upwards. The difference between the
PETT curves with 7.5 and 8.0 mm ID, respectively, yields the potential error (thick bottom line). With a constant inspiratory flow of 18 litre min1 (used in the present study) and a narrowing from 8 to 7.5 mm ID, the PRohrer overestimates the true Ptr by 0.11 cm H2O; with 48 litre min1 the overestimation is 1.06 cm H2O, and with 90 litre min1 the overestimation is 3.98 cm H2O. If this is added to the 0.61 cm H2O difference between PRohrer and measured Ptr it appears that up to an inspiratory flow rate of 48 litre min1 and with a minor narrowing (87.5 mm ID) the error is still acceptable for clinical purposes. With more complex flow patterns and/or more pronounced narrowing, estimation of the potential error is less straightforward, however. A comparison of the actual
PETT over time plot with the same plot of the initial native tracheal tube is of help. A comparison of the flow-dependent upward shift of the
PETT (see Fig. 5) with the native condition is more sensitive than monitoring tracheal tube obstruction with peak airway pressure (Ppeak). As discussed above, the effects of tracheal tube obstruction are flow-dependent. Not until the narrowing is severe enough to accelerate the flow rate up to a threshold above which flow changes from laminar to turbulent, will the increase in Ppeak become clinically obvious. We acknowledge, however, that progressive narrowing by secretions, and, hence, knowledge of the true actual ID, pose a problem for calculation of Ptr.
|
Breath-by-breath information on Ptrwhether measured or calculatedhelps in setting assisted modes of spontaneous breathing,2125 and other modes might also benefit from information on Ptr.
We conclude that calculating tracheal pressure is reliable and the results agree well with those determined by direct measurement. Calculation of Ptr uses pressureflow signals that are already available, is non-invasive, and can be done breath-by-breath. The resistive pressure decreases over the tracheal tube can be determined and the ventilator can be set with respect to tracheal pressure. If common clinical problems like progressive narrowing of the tracheal tube by secretions can be taken into account appropriately, this method can aid respiratory monitoring without the possible errors of direct tracheal pressure measurement.
![]() |
Acknowledgements |
---|
![]() |
Appendix |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
For the inspiratory and expiratory PETT it follows:
PETT-in=Paw-in Ptr-in(1)
PETT-ex= Paw-ex Ptr-ex(2)
with in and ex denoting inspiration and expiration, respectively. The inspiratory PETT is positive, the expiratory
PETT is negative.
From equations (1) and (2) the inspiratory and expiratory Ptr can be derived:
Ptr-in=Paw-in PETT-in(3)
Ptr-ex=Paw-ex + PETT-ex(4)
Two values are, hence, necessary for calculating Ptr, airway pressure (Pprox) and PETT. Calculation of
PETT requires a mathematical description of the non-linear pressureflow curve of the tracheal tube. This has been done either by using empirical equations such as the Rohrer equation14 or by using equations based on general physical laws such as the Ito and Blasius formula.2 2628 We investigated the approach using the Rohrer coefficients.4 For comparison, the approach based on the Blasius formula2 was also checked.
Rohrer equation
The Rohrer equation was first used to describe the airflow resistance of the lung14 and can describe the pressureflow curve of a single tube or a tubing system. It combines a linear with a non-linear pressureflow dependence. According to the Rohrer equation the inspiratory and expiratory pressure decrease across the tracheal tube PETT is described as:
PETT-in=k1 · |V·in| + k2 · V·in2(5)
PETT-ex=k1 · |V·ex| + k2 · V·ex2(6)
with k1 and k2 being the tube coefficients and V· being the flow rate. The tube coefficients are determined by fitting the Rohrer equation to the measured pressure and flow values. These have been measured in vitro taking into consideration the same gas mixture used by the patients, and for all sizes of tracheal tubes in clinical use the tube coefficients are available.4 29 To avoid the calculation of the pressure decrease to be affected by the direction of the flow its absolute value is used.
BlasiusIto approach (dimensional form)
To calculate the resistive pressure decrease across the tracheal tube we used the BlasiusIto approach in its simplified and dimensional form as proposed by Jarreau and colleagues.2 According to fluid dynamic theory the total inspiratory pressure decrease across the tracheal tube is divided into two components, the first component Pke being caused by the change in kinetic energy between two distinct cross sections (here tube connector and tracheal tube), and the second component
Pvd being the friction pressure decrease:
PETT-in=
Pke +
Pvd(7)
The first component Pke assuming a change in kinetic energy between a standard tracheal tube connector and a tracheal tube of ID
3 mm is given by:
Pke=0.98 · V·2 · (D4 5.72)(8)
where V· is the flow rate (litre s1), and D is the ID of the tracheal tube (cm).
The second component Pvd is calculated using the Blasius formula in its simplified and dimensional form:
Pvd=0.0104 · L · D4.75 · V·1.75(9)
where L is the length of the tracheal tube (cm), and V· is the flow rate (litre s1).
This gives the following simplified formula:
Pke=0.98 · V·2 · [DETT4 DCONN4] when D
3 mm
Pke=0.98 · V·2 · [1.5 · DETT4 DCONN4] when D=2.5 mm
where DETT is the diameter of the tracheal tube and DCONN is the diameter of the tracheal tube connector (usually 1.15 cm).
All Ptr calculations were performed with a tracheal tube of 32 cm length and of 7 mm ID.
|
|
|
|
|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2 Jarreau PH, Louis B, Dassieu G, et al. Estimation of inspiratory pressure drop in neonatal and pediatric endotracheal tubes. J Appl Physiol 1999; 87: 3646
3 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 30710[ISI][Medline]
4 Guttmann J, Eberhard L, Fabry B, Bertschmann W, Wolff G. Continuous calculation of intratracheal pressure in tracheally intubated patients. Anesthesiology 1993; 79: 50313[ISI][Medline]
5 Czosnyka M, Harland S, Piechnik S, Kirkpatrick P, Czosnyka Z, Menon D. Systematic overestimation of intracranial pressure measured using a Camino pressure monitor. J Neurol Neurosurg Psychiatry 1996; 61: 4278[ISI][Medline]
6 Czosnyka M, Czosnyka Z, Pickard JD. Laboratory testing of three intracranial pressure microtransducers: technical report. Neurosurgery 1996; 38: 21924[ISI][Medline]
7 Sondergaard S, Karason S, Hanson A, et al. Direct measurement of intratracheal pressure in pediatric respiratory monitoring. Pediatr Res 2002; 51: 33945
8 Chang HK, Mortola JP. Fluid dynamic factors in tracheal pressure measurement. J Appl Physiol 1981; 51: 21825
9 Loring SH, Elliott EA, Drazen JM. Kinetic energy loss and convective acceleration in respiratory resistance measurements. Lung 1979; 156: 3342[ISI][Medline]
10 Sullivan M, Paliotta J, Saklad M. Endotracheal tube as a factor in measurement of respiratory mechanics. J Appl Physiol 1976; 41: 5902
11 Wright PE, Marini JJ, Bernard GR. In vitro versus in vivo comparison of endotracheal tube airflow resistance. Am Rev Respir Dis 1989; 140: 1016[ISI][Medline]
12 Navalesi P, Hernandez P, Laporta D, et al. Influence of site of tracheal pressure measurement on in situ estimation of endotracheal tube resistance. J Appl Physiol 1994; 77: 2899906
13 Conti G, De Blasi RA, Lappa A, et al. Evaluation of respiratory system resistance in mechanically ventilated patients: the role of the endotracheal tube. Intensive Care Med 1994; 20: 4214[ISI][Medline]
14 Rohrer F. Der Strömungswiderstand in den menschlichen Atemwegen und der Einfluß der unregelmäßigen Verzweigungen des Bronchialsystems auf den Atmungsverlauf in verschiedenen Lungenbezirken. Pfluegers Arch 1915; 162: 22599
15 Wall MA. Infant endotracheal tube resistance: effects of changing length, diameter, and gas density. Crit Care Med 1980; 8: 3840[ISI][Medline]
16 VanSurell C, Louis B, Lofaso F, et al. Acoustic method to estimate the longitudinal area profile of endotracheal tubes. Am J Respir Crit Care Med 1994; 149: 2833[Abstract]
17 Louis B, Glass GM, Fredberg JJ. Pulmonary airway area by the two-microphone acoustic reflection method. J Appl Physiol 1994; 76: 223440
18 Jarreau PH, Louis B, Desfrere L, et al. Detection of positional airway obstruction in neonates by acoustic reflection. Am J Respir Crit Care Med 2000; 161: 17546
19 Guttmann J, Eberhard L, Haberthur C, et al. Detection of endotracheal tube obstruction by analysis of the expiratory flow signal. Intensive Care Med 1998; 24: 116372[CrossRef][ISI][Medline]
20 Karason S, Sondergaard S, Lundin S, Wiklund J, Stenqvist O. Direct tracheal airway pressure measurements are essential for safe and accurate dynamic monitoring of respiratory mechanics. A laboratory study. Acta Anaesthesiol Scand 2001; 45: 1739[CrossRef][ISI][Medline]
21 Bolder PM, Healy TE, Bolder AR, Beatty PC, Kay B. The extra work of breathing through adult endotracheal tubes. Anesth Analg 1986; 65: 8539[Abstract]
22 Habib MP. Physiologic implications of artificial airways. Chest 1989; 96: 1804[ISI][Medline]
23 Haberthur C, Fabry B, Stocker R, Ritz R, Guttmann J. Additional inspiratory work of breathing imposed by tracheostomy tubes and non-ideal ventilator properties in critically ill patients. Intensive Care Med 1999; 25: 5149[CrossRef][ISI][Medline]
24 Messinger G, Banner MJ, Blanch PB, Layon AJ. Using tracheal pressure to trigger the ventilator and control airway pressure during continuous positive airway pressure decreases work of breathing. Chest 1995; 108: 50914
25 Messinger G, Banner MJ. Tracheal pressure triggering a demand-flow continuous positive airway pressure system decreases patient work of breathing. Crit Care Med 1996; 24: 182934[CrossRef][ISI][Medline]
26 Ito H. Friction factor for turbulent flow in curved pipes. J Basic Eng 1959; 81: 12334
27 Blasius W. A new model for demonstrating the course of blood pressure and blood flow intensity of arterial circulation in various physiological and pathological conditions. Z Kreislaufforsch 1968; 57: 792800[ISI][Medline]
28 Lofaso F, Isabey D, Lorino H, Harf A, Scheid P. Respiratory response to positive and negative inspiratory pressure in humans. Respir Physiol 1992; 89: 7588[CrossRef][ISI][Medline]
29 Guttmann J, Kessler V, Mols G, Hentschel R, Haberthur C, Geiger K. Continuous calculation of intratracheal pressure in the presence of pediatric endotracheal tubes. Crit Care Med 2000; 28: 101826[ISI][Medline]