Effect of positive expiratory pressure and type of tracheal cuff on
the incidence of aspiration in mechanically ventilated patients in
an intensive care unit*
Umberto Lucangelo, MD; Walter A. Zin, MD, PhD; Vittorio Antonaglia, MD; Lara Petrucci, MD;
Marino Viviani, MD; Giovanni Buscema, MD; Massimo Borelli; Giorgio Berlot, MD
I
t has been demonstrated thathigh-volume low-pressure (HVLP) tracheal tubes do not avoid the contamination of the lower airway by secretions leaking from the subglottis and above (1, 2). This finding owes to the longitudinal folds within the cuff wall that occur on its inflation within the tra-chea even if adequate pressure (25–30 cm H2O) is used (1, 3). Contamination of the lower airways may ensue and promote ventilator-associated pneumonia (4 – 6).
Ventilator-associated pneumonia is a very important cause of prolongation of stay and death in the intensive care unit (ICU).
Silent aspiration of upper airway se-cretions has been reported in patients undergoing general anesthesia (7, 8) and in the ICU (9 –11). In addition, interest-ing in vitromodels have been employed to demonstrate that even under favorable conditions, leakage can still be detected (12–14). However, leakage was
mini-mized with HVLP tubes, thus implying that the material used in the cuffs and their shape may be of paramount impor-tance in determining whether microinha-lation will take place. The use of positive expiratory pressure (PEP) may also con-tribute to seal the leakage (12).
Thus, the aims of this study were to evaluate the effects on microinhalation of the following: 1) the application of 5 cm H2O of PEP for five consecutive hours and subsequent PEP removal in the fol-lowing 7 hrs in mechanically ventilated patients; 2) the use of endotracheal tube cuffs made of polyvinyl chloride (PVC) or polyurethane inflated with 30 cm H2O; and 3) the identification of the minimum safe level of PEP to avoid leakage in a benchtop model.
MATERIALS AND METHODS
Patients and Design of the Study. Forty consecutive adult patients admitted to the ICU Objective:To test the effects of positive expiratory pressure on
the leakage of fluid around cuffs of different tracheal tubes, in mechanically ventilated patients and in a benchtop model.
Design: Randomized clinical trial and experimental in vitro study.
Setting:Intensive care unit of a university hospital.
Patients:Forty patients recovering in the intensive care unit were ventilated in volume-controlled mode. Twenty patients were randomly intubated with Hi-Lo tubes (HL group), whereas the remaining 20 subjects were intubated with SealGuard tubes (SG group).
Interventions: Immediately after intubation and cuff inflation with 30 cm H2O, Evans blue was applied onto the cephalic surface
of the tracheal tube cuff. A 5-cm H2O positive expiratory pressure
was used during the first 5 hrs of stay, and thereafter it was removed. Bronchoscopy verified whether the dye leaked around the cuff. The experiment lasted 12 hrs. Leakage was also tested in vitro with the same tracheal tubes with incremental level of positive expiratory pressure.
Measurements and Main Results:At 1 hr, 5 hrs, and thereafter hourly until 12 hrs, bronchoscopy was used to test the presence of dye on the trachea caudal to the cuff. At the fifth hour, two patients of the HL group failed the test. One hour after positive expiratory pressure removal, all subjects in group HL exhibited a dyed lower trachea. On the other hand, one patient in group SG presented a leak at the eighth hour, and at the 12th hour three of them were still sealed.In vitro, the same level of positive expi-ratory pressure delayed the passage of dye around the cuff; after 30 mins positive expiratory pressure was removed, and in 10 mins all dye leaked only in the Hi-Lo tube.
Conclusions: We found that 5 cm H2O positive expiratory
pressure was effective in delaying the passage of fluid around the cuffs of tracheal tubes bothin vivoandin vitro. The SealGuard tube proved to be more resistant to leakage than Hi-Lo. (Crit Care Med 2008; 36:●●●–●●●)
KEY WORDS: tracheal tube; microinhalation; ventilator-associ-ated pneumonia; positive expiratory pressure; volume-controlled mechanical ventilation; bronchoscopy; Evans blue
*See also p. 00.
From the Department of Perioperative Medicine, Intensive Care and Emergency, Cattinara Hospital, Tri-este University School of Medicine, TriTri-este, Italy (UL, VA, LP, MV, GB, GB); Department of Mathematics and Computer Science, University of Trieste, Italy (MB); and Carlos Chagas Filho Institute of Biophysics, Federal Uni-versity of Rio de Janeiro, Rio de Janeiro, Brazil (WAZ).
Supported, in part, by the Department of Periopera-tive Medicine, Intensive Care and Emergency, Cattinara Hospital, Trieste University School of Medicine, Trieste, Italy.
None of the authors has any possible sources of financial support, corporate involvement, and patent holdings. The authors have not disclosed any potential conflicts of interest.
For information regarding this article, E-mail: u.lucangelo@fmc.units.it
Copyright © 2008 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins
of Cattinara Hospital from January to October 2006 who required immediate orotracheal in-tubation and mechanical ventilation because of deterioration of consciousness state (Glas-gow Coma Scale scoreⱕ8) were studied in the semirecumbent position (30 – 45°). The exclu-sion criteria were chronic obstructive pulmo-nary disease, active pneumonia, previous sur-gery of the upper airway or larynx, obesity, or other causes that increased abdominal pressure or jeopardized normal chest wall function.
The patients were randomly divided into two groups of 20 patients each at admission by a physician unaware of the study: those intu-bated with Hi-Lo tracheal tubes and those receiving SealGuard tubes (both from Mallinckrodt Medical, Cornamady, Athione County, Ireland), groups HL and SG, respec-tively. The geometric characteristics of the Hi-Lo cuffs (inflation pressure of 30 cm H2O) resting on the bench were length 3.7 cm, circumference 10 cm, volume 20 mL, diame-ter 3.2 cm, and spindle-like shape; the char-acteristics for SealGuard cuffs were length 4.5 cm, circumference 9.2 cm, volume 20 mL, diameter 2.9 cm, and cylindrical shape. Before intubation, propofol (1 mg/kg of body weight) and succinylcholine chloride (1 mg/kg of body weight) were injected intravenously if neces-sary. Appropriate sedoanalgesia (midazolam and remifentanil) and muscle paralysis (atra-curium), if required, were administered dur-ing the entire period of study. All patients were mechanically ventilated in volume-control mode with a 7200 Puritan-Bennett ventilator (Puritan-Bennett Corporation, Carlsbad, CA) with tidal volume of 8 –9 mL/kg and respira-tory rate to maintain normocapnia(12–14 breaths/min), and 5 cm H2O of PEP was im-mediately applied. Tracheal tube size was cho-sen following the Higenbottam-Payne equa-tion (15), and the cuff was inflated with air to a pressure of 30 cm H2O, as verified with an aneroid manometer (Mallinckrodt Medical). Careful aspiration of upper and lower airway secretions ensued. One milliliter of Evans blue diluted in 1 mL of saline solution (NaCl 0.9%) was then carefully placed on the top of the cuff. One hour and 5 hrs after intubation, fiber-optic bronchoscopy (Pentax FI-10P2, 3.4-mm distal tip diameter, PENTAX Imaging Com-pany, Golden, CO) was performed to detect the possible presence of blue dye in the trachea, according to the modified Evans blue dye test, as recently reevaluated (16). Thereafter, PEP was removed and bronchoscopy was per-formed hourly until 12 hrs after intubation. If on any occasion a blue spot was seen on the trachea caudal to the tube’s tip, leakage was confirmed and the experiment was finished. In all instances, the bronchoscope tip was never introduced below the tube’s distal end. Mouth and tracheal secretions were not aspirated during the whole experiment.
The investigation was approved by the local ethics committee, and informed consent was obtained from each patient’s next of kin.
Statistical Analysis.Statistical analysis was performed by the open source statistical pack-age R (17). Likelihood ratio test was used to assess significance in Cox regression model analysis (18, 19), and a log-rank test was used to assess differences in survival analysis be-tween the two groups of tracheal tubes (20).
In VitroStudy. Two 100-mL Pyrex measur-ing cylinders (internal diameter 2.5 cm) served as a vertical trachea model. One was intubated with a size 7.5 (inner diameter) Hi-Lo tracheal tube and the other with a size 7.5 (inner diameter) SealGuard tube (both from Mallinckrodt Medical, Cornamady, Athione County, Ireland). The same pressure (30 cm H2O) was simultaneously applied to both cuffs via a Y tube connected to an aneroid manometer (Mallinckrodt Medical). Through the tubes, 5, 7.5, and 10 cm H2O of PEP was randomly applied to both measuring cylinders by means of a 3-L syringe. One milliliter of Evans blue diluted in 1 mL of saline solution (NaCl 0.9%) was added to the top of each cuff. After 30 mins, PEP was removed, and we looked for fluid leakage past tube cuffs. The study lasted 1 hr. The experiment was repeated in triplicate with new tubes of each kind.
RESULTS
Table 1 shows the anthropometric and functional characteristics of our patients at admission to the ICU. The two groups presented similar average values of height, weight, and age. In the HL group, nine patients received size 7.5 tubes and 11 patients were intubated with size 8 tubes, whereas in the SG group the re-spective data were 10 and 10. The pairing between the same tracheal tube size and the anteroposterior diameter of the glot-tis (15) was comparable in SG and HL group (Table 2).
Figure 1 shows that at the fifth hour after intubation, dye was detected for the first time in the lower trachea of two patients of the HL group. The broncho-scopic image of a representative patient presenting fluid leakage is shown in Fig-ure 2. At this time, PEP was removed. One hour afterward all the patients of the HL group presented fluid leakage around the tracheal tube cuff. At the eighth hour, the first patient of the SG group showed dye in the lower trachea. At the end of the experiment (12 hrs), three patients of this group still depicted sealed tracheal cuffs (Fig. 1). Evidence concerning the statisti-cally significant difference in global behav-ior between the two groups of patients was assessed by the log-rank test (z⫽ ⫺6.14, p⬍10⫺9).
Figure 3 depicts the two tubes (Hi-Lo on the left and SealGuard on the right)
with inflated cuffs (30 cm H2O) inside measuring cylinders (PEP below the cuffs 5 cm H2O) right after the application of Evans blue on top of the cuffs. While the SealGuard tube presented a clear cuff sealed against the tube wall, the Hi-Lo showed longitudinal and oblique folds within the cuff wall clearly dyed in blue. At 10 mins after the removal of PEP (40 mins after the application of Evans blue on the top of the cuffs), all liquid had passed around the cuff of the Hi-Lo tube and the folds were very well defined, whereas no fluid leaked around the Seal-Guard cuff (Fig. 4). The same results were found with 7.5 and 10 cm H2O of PEP (all liquid passed around the cuff of Hi-Lo tube after 11 and 10 mins, respectively).
DISCUSSION
The leakage of fluid past tracheal tube cuffs is a potential hazard for intubated
Table 1.Anthropometric data, clinical variables, and causes for admission to the intensive care unit
HL SG
No. of patients 20 20
Age, yrs 67.4⫾9.5 67.5⫾9.7 Height, cm 173.3⫾6.1 172.6⫾5.3 Weight, kg 75.1⫾10.7 73.2⫾12.8 Tube size
7.5 mm i.d.
9 10
Tube size 8 mm i.d.
11 10
VT, mL 615⫾53 605⫾69 RR, breaths/min 12.5⫾3.5 12.3⫾3.9 Subarachnoidal
hemorrhage
8 9
Cerebral neoplasia 12 11
HL, group of patients intubated with Hi-Lo tracheal tube; SG, group of patients intubated with SealGuard tracheal tube; i.d., inner diame-ter; VT, tidal volume; RR, respiratory rate.
Mean⫾SDvalues where applicable.
Table 2.Number of tracheal tube size used following the Higenbottam-Payne equation
Tube Size, i.d.
Tube Type,
No. of Patients AP, mm p
7.5 HL (9) 23.25⫾1.65 NS SG (10) 23.41⫾1.61
26.55⫾0.73
8 HL (11) 26.72⫾0.77 NS
SG (10)
i.d., internal diameter in mm; AP, anteropos-terior diameter of the glottis; HL, Hi-Lo; SG, SealGuard; NS, not statistically significant.
Data are expressed as mean andSD(unpaired
patients. Different cuff shapes and mate-rials have been continuously developed together with other measures, such as recumbent position of the thorax and continuous subglottic aspiration (4, 21). However, the solution still remains to be provided.
In benchtop models, the simulated use of positive pressure has been proven to minimize the microinhalation of liquid around the tube cuff (12, 22). Different cuff inflation pressures have also been tested, and low pressures, such as 10 cm H2O, resulted in a significant leakage (14). Thus, we aimed to test in intubated and mechanically ventilated patients us-ing HVLP tracheal tubes the possible
con-tribution of PEP to minimize the passage of fluid into the trachea. The same tubes were also testedin vitroto provide exper-imental support to thein vivoresults.
Two tubes were randomly tested in two similar groups of 20 patients each. Those belonging to the HL group were intubated with a Hi-Lo tracheal tube, whose cuff is made of PVC, whereas those in SG group received a SealGuard tube with a polyurethane cuff. The tube size for each patient was adequately calcu-lated according to his or her individual anthropometric data (15).
Appropriate cuff inflation pressure level is still debated (a range of 20 –30 cm H2O is recommended), and 25 cm H2O is
suggested as a minimal safe pressure value to prevent microinhalation (23). In our study, to highlight the sealing prop-erties due to different cuff materials (polyurethane vs. PVC), the cuff was in-flated to a pressure of 30 cm H2O as maximum value applicable to tracheal mucosa without ischemia (3). A limit of this study is that it did not compare re-sults (leakage and fold formation) at in-ferior cuff inflation pressures: however, PEP removal caused dye leakage in all Hi-Lo tubes, thus supporting the cen-tral role of PEP application when invag-ination of cuff wall is supposed to be minimal.
To our knowledge, no long-duration study has been conducted recently in pa-tients to address leakage of fluid past the tracheal tube with either PVC or polyure-thane cuffs.
During the first 5 hrs of the study, our patients were ventilated with a PEP of 5 cm H2O, and two patients of the HL group failed (Fig. 1). PEP was then re-moved and the remaining patients of this group presented a blue spot on the tra-chea 1 hr later. In the SG group, how-ever, the first patient failed on the eighth hour after intubation, and on the 12th hour three cases were still sealed. Lower tracheal aspiration did not collect blue secretions in these patients. The two groups were significantly different in global behavior (log-rank test,p⬍10⫺9). Although PEP was removed during the experiment, this factor can be considered as a time-dependent covariate in a Cox regression analysis framework. Studying the maximal model (i.e., time depen-dency and PEP included as a covariate) and simplifying from not significant in-teractions, a minimal adequate model can be fitted to the data, and a hypothesis test confirming the former significant difference between groups can be checked (likelihood ratio test statistics 86.2 on 1df, Weibull distributed,p⫽0). Only one study in the literature (8) compared the sealing properties of cuffs made of polyurethane and PVC. The lat-ter, in fact, was the Hi-Lo cuff used in the present investigation. The authors ap-plied methylene blue on top of the cuffs 1 hr before extubation, and immediately af-ter removal of the tube they performed a bronchoscopy to investigate the presence of blue dye on the lower trachea. They also found that the polyurethane cuff pro-vided a better sealing than the PVC cuff (0% vs. 31.2% seepage, respectively). This result is in line with ours: At 5 hrs under
Figure 1.Survival time plot of the percentage of leaking tubes as a function of time after intubation.
Dashed line, patients intubated with Hi-Lo (HL) tracheal tubes;solid line, patients intubated with SealGuard (SG) tubes. The first leakage was detected in two patients of the HL group at the fifth hour. Positive expiratory pressure (5 cm H2O) was then removed. One hour afterward, all the HL patients
presented fluid leakage past the cuffs. In the SG group, the first microinhalation was detected at the eighth hour, and at the end of the experiment (12 hrs) three patients were still sealed. There was a statistically significant difference in global behavior between the two groups of patients (p⬍10⫺9).
PEP, 10% of the patients in the HL group had positive findings in the modified Evans blue dye test against 0% in the other group. Petring et al. (8) did not use PEP in their study. Another work dem-onstrated in humans that HVLP cuffs tended to be superior regarding cuff seal in comparison with other designs (24), although controversial results can also be found in the literature (25).
In attempt to prevent leakage, a pro-totype of an HVLP cuff with a smooth tapering shape of compliant latex was proposed (26). Cuff lubrication with wa-ter-soluble gel was also proposedin vitro
and in mechanically ventilated patients. The authors of that study found a differ-ent incidence of leakage in bench model (0% vs. 100% in lubricated and nonlubri-cated tubes, respectively) and in anesthe-tized patients (11% vs. 83%, respec-tively), suggesting a plugging effect of gel in the cuff wall (27). Recently Young et al. (28) describedin vitroand in patients the clinical advantages of a volume low-pressure silicone cuff compared with an HVLP cuff. The HVLP cuff does not per-mit an accurate tracheal wall pressure control and is associated with tracheal injury. On the contrary, only a fixed
pro-portion of the intracuff pressure is trans-mitted to tracheal mucosa using a low-volume low-pressure cuff; however, that cuff required a constant inflation pres-sure device to enpres-sure continuous cuff in-flation and airway protection (28).
Many studies have addressed the seal-ing properties of the tracheal tube cuff. Some authors, using pig or model tra-cheas, compared Hi-Lo tubes with other tubes and concluded that the former avoided seepage around the cuffs to a larger extent than the others (13, 14, 29). These studies, however, did not take into consideration the pressure beyond the cuffs, although they verified the effects of different pressures inside the cuffs and the time required for the leakage to oc-cur. No polyurethane cuff has been tested in vitro.
Considering the data gathered from our patients, we decided to analyze a benchtop model in which a 5-cm H2O PEP and PVC and polyurethane cuffs could be tested under a cuff inflation pressure equal to 30 cm H2O. The same types of tubes used in our clinical study were used, but only tube size 7.5 was employed. In the present case, SealGuard cuffs were longer and narrower than the Hi-Lo cuff; both were wider than the tra-chea. Additionally, the former is made of polyurethane while the latter uses PVC. The former is more pliable and more rub-berlike than the latter, which may im-prove its ability to avoid seepage.
The shape of the tubes and cuff phys-ical characteristics, such as diameter, thickness (or thinness), and compliance, have been reported to play a role in the tube’s sealing properties (30). In humans (without PEP), the polyurethane tube proved to better seal the trachea in a time span of 3 hrs (8). To our knowledge, ours is the firstin vitro study comparing the sealing properties of these two tubes.
Regarding the use of PEP, it has been reported that 5 cm H2O PEP delays the leakage in model trachea (22) and sili-cone tube (12) preparations. We used measuring cylinders (inner diameter 2.5 cm) to simulate the trachea and only a small (⬵7 mm H2O) head of pressure above the cuff (Fig. 3) to reproduce the same experimental condition of in vivo study and to minimize gravitational forces. Qualitatively our results resemble those of Janson and Poulton (22), who reported a smaller leakage and a longer time to seepage when PEP was applied to their model (size 7.5 HVLP Mallinckrodt tube). In our case, with PEP no leakage
Figure 3.Two tracheal tubes (Hi-Lo on the left and SealGuard on the right) with inflated cuffs (30 cm H2O) inside measuring cylinders (positive expiratory pressure of 5 cm H2O) right after the application
of Evans blue on the top of the cuffs. One can easily note dyed longitudinal folds within the cuff wall on the left tube.
Figure 4.Two tracheal tubes (Hi-Lo on the left and SealGuard on the right) with inflated cuffs (30 cm H2O) inside measuring cylinders (positive expiratory pressure of 5 cm H2O) 10 mins after the removal
was detected in both tubes until 30 mins had elapsed, although the longitudinal and oblique folds found in the Hi-Lo tube displayed colored liquid (Fig. 3). Addi-tionally, we noted tiny bubbles moving proximally through the folds in the Hi-Lo tubes, which possibly reflect the pneu-matic effect generated by the higher pres-sure distal to the cuff. In the study of Janson and Poulton, when PEP was nil, the investigators found a shorter average time (7 mins) before leakage than in the presence of PEP (21 mins) (22). We found that 10 –11 mins after PEP removal, all liquid (2 mL) had passed around the cuff of the Hi-Lo tube, whereas the SealGuard retained the dye above the cuff (Fig. 4). These data show a protective effect for any PEP applied, and the loss of pneu-matic effect produced immediate leakage. The application of 7.5 and 10 cm H2O PEP did not improve sealing during the observational period but only increased the bubble formation above the cuff at the highest level of PEP applied. We ob-served in our study that the invaginations of the cuff wall did not disappear during PEP trial, suggesting that PEP acts as a protective counterbalance pressure with-out modifying cuff sealing property. Probably the best protective PEP effect is coupled with the maximum safe cuff pressure applicable. These findings are in agreement with those of Young et al. (28), demonstrating that at different PEP levels, the rate of flow secretion passing through the cuff are inversely related to the cuff pressure (12).
Thus, the presentin vitroobservations support our clinical findings. Theoreti-cally, spontaneous breathing and tracheal aspiration may lower the pressure in the cuff, thus increasing leakage around it. To avoid this confounding factors, all the patients were mechanically ventilated in volume control mode and were unable to breathe spontaneously; moreover, mouth and tracheal secretions were not aspi-rated during the whole experiment.
CONCLUSIONS
We found that 5 cm H2O PEP was the minimum safe level of pressure that im-proved the sealing around HVLP cuffs, especially in the polyurethane cuff. Con-sidering that 5 cm H2O PEP is a relatively low pressure level, its daily application should be recommended, if not
contrain-dicated, to prevent contamination of lower airways in mechanically ventilated ICU patients.
ACKNOWLEDGMENT
Dr. W.A. Zin is a researcher of the Brazilian National Council for Scientific and Technological Development (CNPq/ MCT), Brazil, and a visiting professor at the Trieste University School of Medicine, Italy.
REFERENCES
1. Pavlin EG, Van Nimwegard D, Horbein TF: Failure of a high-compliance low-pressure cuff to prevent aspiration. Anesthesiology
1975; 42:216 –219
2. MacRae W, Wallace P: Aspiration around high-volume low-pressure endotracheal cuff.
BMJ1981; 283:1220
3. Seegobin RD, Van Hasselt GL: Endotracheal cuff pressure and tracheal mucosal blood flow: Endoscopic study of effects of four large volume cuffs.BMJ1984; 288:965–968 4. Cook D, De Jonghe B, Brochard L, et al:
Influence of airway management of ventila-tor-associated pneumonia.JAMA 1998; 279: 781–787
5. Cook D, Walter SD, Cook RJ, et al: Incidence of and risk factors for ventilator-associated pneumonia in critically ill patients.Ann In-tern Med1998; 129:433– 440
6. Heyland DK, Cook DJ, Griffith L, et al: The attributable morbidity and mortality of ven-tilator-associated pneumonia in the critically ill patient.Am J Respir Crit Care Med1999; 159:1249 –1256
7. Bernard WN, Cottrell JE, Sivakumaran C, et al: Adjustment of intracuff pressure to pre-vent aspiration. Anesthesiology 1979; 50: 363–366
8. Petring OU, Adelhøj B, Jensen BN; Preven-tion of silent aspiraPreven-tion due to leaks around cuffs of endotracheal tubes.Anesthesia Analg
1986; 65:777–780
9. Browning DH, Graves SA: Incidence of aspi-ration with endotracheal tubes in children.
J Pediatr1983; 102:582–584
10. McCleave DJ, Fisher M: Efficacy of high vol-ume low pressure cuffs in preventing aspira-tion. Anaesth Intensive Care 1977; 5:167–168
11. Spray SB, Zuidema GD, Cameron JL: Aspira-tion pneumonia. Am J Surg 1976; 131: 701–703
12. Young PJ, Rollinson M, Downward G, et al: Leakage of fluid past the tracheal tube cuff in a benchtop model. Br J Anaesth1997; 78: 557–562
13. Young PJ, Blunt MC: Compliance character-istics of the Portex Soft Seal Cuff improves
seal against leakage of fluid in a pig trachea model.Crit Care1999; 3:127–130
14. Dullenkopf A, Gerber A, Weiss M: Fluid leak-age past tracheal tube cuffs: Evaluation of the new Microcuff endotracheal tube.J In-tensive Care Med2003; 29:1849 –1853 15. Higenbottam T, Payne J: Glottis narrowing
in lung disease. Am Rev Respir Dis 1982; 125:746 –750
16. Belafsky PC, Blumenfeld L, LePage A, et al: The accuracy of the modified Evans blue dye test in predicting aspiration. Laryngoscope
2003; 113:1969 –1972
17. R Development Core Team: R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vi-enna, Austria. http://www.R-project.org. Ac-cessed 2006
18. Vittinghoff E, Glidden DV, Shibolski SC, et al: Regression Methods in Biostatistics: Lin-ear, Logistic, Survival, and Repeated Mea-sures Models. New York, Springer, 2005 19. Crawley MJ: Statistics: An Introduction
Us-ing R. Chichester, UK, Wiley, 2005 20. Glantz SA: Primer of Biostatistics. New York,
McGraw-Hill, 2003
21. Torres A, Serra-Batlles J, Ros E, et al: Pul-monary aspiration of gastric content in pa-tients receiving mechanical ventilation: The effect of body position.Ann Intern Med1992; 116:540 –543
22. Janson BA, Poulton TJ: Does PEP reduce the incidence of aspiration around endotracheal tubes?Can Anesth Soc J1986; 33:157–161 23. Lomholt N: A device for measuring the
lat-eral wall cuff pressure of endotracheal tubes.
Acta Anaesthesiol Scand1992; 36:775–778 24. Hahnel J, Treiber H, Konrad F, et al: A
com-parison of different endotracheal tubes: Tra-cheal cuff seal, peak centering and the inci-dence of postoperative sore throat.
Anaesthesist1993; 42:232–237
25. Seegobin RD, Van Hasselt GL: Aspiration be-yond endotracheal cuffs.Can Anaesth Soc J
1986; 33:273–279
26. Young PJ, Ridley SA, Downward G: Evalua-tion of a new design of tracheal tube cuff to prevent leakage of fluid to the lungs.Br J Anaesth1998; 80:796 –799
27. Blunt MC, Young PJ, Patil A, et al: Gel lubri-cation of the tracheal tube cuff reduces pul-monary aspiration.Anesthesiology2001; 95: 377–381
28. Young PJ, Pakeerathan S, Blunt MC, et al: A low-volume, low-pressure tracheal tube cuff reduces pulmonary aspiration.Crit Care Med
2006; 34:632– 639
29. Asai T, Shingu K: Leakage of fluid around high-volume, low-pressure cuffs apparatus: A comparison of four tracheal tubes. Anaesthe-sia2001; 56:38 – 42
