A critical function of the endotracheal tube cuff is to seal the airway, thus preventing aspiration of pharyngeal contents into the trachea and to ensure that there are no leaks past the cuff during positive pressure ventilation. However, complications have been associated with insufficient cuff inflation. Consequences of micro-aspiration of oropharyngeal secretions include nosocomial pulmonary infections 1. Conventional high-volume, low-pressure cuffs may not prevent micro-aspiration even at cuff pressures up to 60 cm H₂O 2, although some studies suggest that only 25 cm H₂O is sufficient 3. In contrast, newer ultra-thin cuff membranes made from polyurethane effectively prevent liquid flow around cuffs inflated only to 15 cm H₂O 2. In the absence of clear guidelines, many clinicians consider 20 cm H₂O a reasonable lower limit for cuff pressure in adults.

Catastrophic consequences of endotracheal tube cuff over-inflation such as rupture of the trachea 456, tracheo-carotid artery erosion 7, and tracheal innominate artery fistulas are rare now that low-pressure, high-volume cuffs are used routinely. However, post-intubation sore throat is a common side effect of general anesthetic and may partly result from ischemia of the oropharyngeal and tracheal mucosa 8910, and the most common etiology of non-malignant tracheoesophageal fistula remains cuff-related tracheal injury 1112. In addition, acquired laryngeal stenosis may be caused by mechanical abrasion or pressure necrosis of the laryngeal mucosa secondary to high cuff pressure 1314. Animal data indicate that a cuff pressure of only 20 cm H₂O may significantly reduce tracheal blood flow with normal blood pressure and critically reduces it during severe hypotension 15. Similarly, inflation of endotracheal tube cuffs to 20 cm H₂O for just four hours produces serious ciliary damage that persists for at least three days 16. Thus, appropriate inflation of endotracheal tube cuff is obviously important.

Lomholt et al. recommended selecting a cuff pressure of 25 cmH₂O as a safe minimum cuff pressure to prevent aspiration and leaks past the cuff 17; Bernhard et al. supported this recommendation 18. On the other hand, Nordin et al. studied the relationship between cuff pressure and capillary perfusion of the rabbit tracheal mucosa and recommended that cuff pressure be kept below 27 cm H₂O (20 mmHg) 19. Seegobin and Hasselt reached similar conclusions in an in vitro study and recommended cuff inflation pressure not exceed 30 cm H₂O 20. It is thus essential to maintain cuff pressures in the range of 20–30 cm of H₂O

Adequacy of cuff inflation is conventionally determined by palpation of the external balloon. Previous studies suggest that this approach is unreliable 2122. One study, for instance, found that cuff pressure exceeded 40 cm H₂O in 40-to-90% of tested patients 22. However, increased awareness of over-inflation risks may have improved recent clinical practice. Our first goal was thus to determine if cuff pressure was within the recommended range of 20–30 cmH₂O, when inflated using the palpation method. Because cuff inflation practices are likely to differ among clinical environments, we evaluated cuff pressure in three different practice settings: an academic university hospital and two private hospitals. It is also likely that cuff inflation practices differ among providers. We therefore also evaluated cuff pressure during anesthesia provided by certified registered nurse anesthetists (CRNAs), anesthesia residents, and anesthesia faculty.

Cuff pressure can be easily measured with a small aneroid manometer 23, but this device is not widely available in the United States. It would thus be helpful for clinicians to know how much air must be injected into the cuff to produce the minimum adequate pressure. We designed this study to observe the practices of anesthesia providers and then determine the volume of air required to optimize the cuff pressure to 20 cmH₂O for various sizes of endotracheal tubes. Taking another approach to the same question, we also determined compliance of the cuff-trachea system in vivo by plotting measured cuff pressure against cuff volume.

With approval of the University of Louisville Human Studies Committee and informed consent, we recruited 93 patients (42 men and 51 women) undergoing elective surgery with general endotracheal anesthesia from three hospitals in Louisville, Kentucky: 41 patients from University Hospital (an academic centre), 32 from Jewish Hospital (a private hospital), and 20 from Norton Hospital (also a private hospital).

Patients with emergency intubations, difficult intubations, or intubation performed by non-anesthesiology staff; pregnant women; patients with higher risk for aspiration (e.g., full stomach, history of reflux, etc.); and patients with known anatomical laryngeo-tracheal abnormalities were excluded from this study. The Human Studies Committee did not require consent from participating anesthesia providers. However, no data were recorded that would link the study results to specific providers.

Morphometric and demographic characteristics of the patients were similar at each participating hospital (Table 1). Measured cuff pressures averaged 35.3(21.6)cmH₂O; only 27% of the patients had measured pressures within the recommended range of 20–30 cmH₂O. Fifty percent of the values exceeded 30 cmH₂O, and 27% of the measured pressures exceeded 40 cmH₂O. Thus, 23% of the measured cuff pressures were less than 20 mmHg. Measured cuff volume averaged 4.4 ± 1.8 ml. Neither measured cuff pressure nor measured cuff volume differed among the hospitals (Table 2).

There was no correlation between the measured cuff pressure and the age, sex, height, or weight of the patients. Nor did measured cuff pressure differ as a function of endotracheal tube size. Measured cuff volumes were also similar with each tube size. Interestingly, the amount of air required to achieve a cuff pressure of 20 cmH₂O was similar with each tube size (Table 3). However, there was considerable variability in the amount of air required.

CRNAs (n = 72), anesthesia residents (n = 15), and anesthesia faculty (n = 6) performed the intubations. There were no statistically significant differences in measured cuff pressures among these three practitioner groups (P = 0.847).

The compliance of the tube was determined from the measured cuff pressure (cmH₂O) and the volume of air (ml) retrieved at complete deflation of the cuff; this showed a linear pressure-volume relationship: Pressure= 7.5. Volume+2.7, r² = 0.39 (Fig. 1).

Previous studies suggest that the cuff pressure is usually under-estimated by manual palpation. For example, Braz et al. 22 observed cuff pressure exceeding 40 cm H₂O in 91% of PACU patients after anesthesia with nitrous oxide, 55% of ICU patients, and 45% of PACU patients after anesthesia without nitrous oxide. In an experimental study, Fernandez et al. 21 observed that when the cuff was inflated randomly to 10, 20, or 30 cmH₂O, participating physicians and ICU nurses were able to identify the group in 69% of the high-pressure cases, 58% of the normal pressure cases, and 73% of the low pressure cases. Our results are consistent in that measured cuff pressure exceeded 30 cmH₂O in 50% of patients and were less than 20 cmH₂O in 23% of patients. Cuff pressures were thus less likely to be within the recommended range (20–30 cmH₂O) than outside the range. It was nonetheless encouraging that we observed relatively few extremely high values, at least many fewer than reported in previous studies 22. This result suggests that clinicians are now making reasonable efforts to avoid grossly excessive cuff inflation.

Measured cuff inflation pressures were virtually identical at the three study sites: one academic center and two private hospitals. These data suggest that management of cuff pressure was similar in these two disparate settings. Interestingly, there was also no significant or important difference as a function of provider – measured cuff pressures were virtually identical whether filled by CRNAs, residents, or attending anesthesiologists. Our results thus fail to support the theory that increased training improves cuff management.

We evaluated three different types of anesthesia provider in three different practice settings. Although we were unable to identify any statistically significant or clinically important differences among the sites or providers, our results apply only to the specific sites and providers we evaluated. While it is likely that these results are fairly representative, it is obvious that results would not be identical elsewhere because of regional practice differences.

Fernandez et al. 21 found that the volume of air required to inflate the endotracheal tube cuff varies as a function of tube size and type. But interestingly, the volume required to inflate the cuff to a particular pressure was much smaller when the cuff was inflated inside an artificial trachea; furthermore, the difference among tube sizes was minimal under those conditions. We similarly found that the volume of air required to inflate the cuffs to 20 cmH₂O did not differ significantly as a function of endotracheal tube size. These data suggest that tube size is not an important determinant of appropriate cuff inflation volume. A caveat, though, is that tube sizes were chosen by clinicians in our study and presumably matched patient size; results may well have differed if tube size had been randomly assigned. We intentionally avoided this approach since our purpose was to evaluate cuff pressures and associated volumes in three routine clinical settings.

A limitation of this study is that cuff pressure was evaluated just once 60 minutes after induction of anesthesia. Because nitrous oxide was not used, it is unlikely that the cuff pressures varied much during the first hour of the study cases. We recognize that people other than the anesthesia provider who actually conducted the case often inflated the cuffs. However, a full hour was plenty of time for the provider to have checked and adjusted cuff pressure to a suitable level.

We observed a linear relationship between the measured cuff pressure and the volume of air retrieved from the cuff. The regression equation indicated that injected volumes between 2 and 4 ml usually produce cuff pressures between 20 and 30 cmH₂O independent of tube size for the same type of tube. However, there was considerable patient-to-patient variability in the required air volume. Measuring actual cuff pressure thus appears preferable to injecting a given volume of air.

Cuff pressure should be measured with a manometer and, if necessary, corrected.

The author(s) declare that they have no competing interests.

SP oversaw day-to-day study mechanics, collected data on many of the patients, and wrote an initial draft of manuscript.

DIS contributed to study design, data analysis, and manuscript preparation.

PM, SW, and AV recruited patients and performed many of the measurements.

JD conceived of the study and participated in its design.

AW contributed to protocol development, patient recruitment, and manuscript preparation.

All authors read and approved the final manuscript.