Abstract & Commentary

The Cuff-Leak Test for Predicting Post-Extubation Stridor: Are We Focusing on the Wrong Question?

By Richard H. Kallet, MS, RRT, FAARC, FCCM, Director of Quality Assurance, Respiratory Care Services, San Francisco General Hospital, is Associate Editor for Critical Care Alert.

Mr. Kallet reports no financial relationships relevant to this field of study.

Synopsis: This prospective, single-center, observational study investigated whether the presence of deterioration in the results of cuff-leak tests done immediately post-intubation as compared to the same test repeated just prior to extubation would improve sensitivity for detecting subsequent stridor. There was no difference between the testing strategies in sensitivity (86%), whereas the specificity of the pre-extubation test was much higher (76%) than the comparison method (48%).

Source: Gros A, et al. Intra-individual variation of the cuff-leak test as a predictor of post-extubation stridor. Respir Care 2012; May 4. [Epub ahead of print.]

The cuff-leak test (CLT) was performed prospectively in this study using the volume technique whereby the difference between inspired and expired tidal volume before and after cuff deflation is used to detect the presence of laryngeal edema and the likelihood of developing post-extubation stridor.1 The rationale for performing the CLT immediately post-intubation is to control for a false-positive test (that is, little or no leak and subsequent absence of stridor) that may occur when the endotracheal tube is too large relative to laryngeal diameter. The authors determined the change in CLT as the difference between pre-extubation and post-intubation results, so that a negative value (i.e., deterioration in leak volume) would signify a positive test for possible upper airway edema. Data were collected from 104 adult ICU patients requiring more than 48 hours of mechanical ventilation who were deemed ready for an extubation trial. Determinations of sensitivity, specificity, positive, and negative predictive values for stridor were made for each test strategy. The change in CLT using a pre-hoc cutoff of 0 mL was compared to the pre-extubation CLT in which the cutoff was determined post-hoc.

There were seven incidents of post-extubation stridor and respiratory distress (6.7%), all of which occurred within 12 hours, and reintubation was required in six. The CLT comparison method was not better than using a single pre-extubation CLT for detecting stridor. Post-hoc analysis demonstrated that using a cutoff value of 130 mL for the pre-extubation CLT produced the same sensitivity (86%) and a higher specificity (76% vs 48%). Likewise, the positive predictive value was higher for the single CLT vs the CLT comparison method (21% and 11% respectively), whereas the negative predictive values were essentially the same (99% and 98%, respectively).


Post-extubation stridor is a relatively rare but frequently life-threatening event with a reported incidence between 2-16%.2 There are several indirect risk factors (female gender, height, obesity, and neurologic injury) as well as direct risk factors (duration of endotracheal intubation, history of difficult or traumatic intubation, unplanned traumatic extubation, upper airway infection, trauma, or surgery). The ability to accurately predict post-extubation stridor is limited by crude measurement techniques and questionable inter-observer reliability. More importantly, the low prevalence of stridor limits the ability to establish a reliable positive predictive value for the CLT.

We addressed this issue at our institution following a sentinel event in which a patient with several unrecognized risk factors developed severe post-extubation stridor and subsequently suffered a cardiac arrest. Although we had been using various CLT techniques over the years, these were not standardized, nor were absolute cutoff criteria agreed upon. Also, patients were not routinely screened for risk factors. Since then, we have systematically screened all intubated patients for risk factors and stratified the number and type of risk factors that would automatically trigger a CLT.

Our institution has also incorporated the volume CLT introduced by Miller et al.1 The advantage of the volume CLT is that it uses standardized ventilator settings to minimize inter-observer variability. Pressure-based techniques typically rely on a manual resuscitator with an inline pressure manometer. The airway pressure at which a leak becomes "audible" (if at all) is used to assess whether upper airway edema is likely to be present. However, neither the specific amount nor the delivery technique of applied airway pressure is standardized. An acceptable cutoff value for airway pressure has never been established and therefore is arbitrary. In addition, the "quality" of the audible air leak has never been addressed. For example, should clinicians make the same decision regarding extubation at a leak pressure of 20 cm H2O when they hear a brief high-pitched squeak as in the case of a sustained sonorous sound?

More importantly, when we compared the volume CLT to our traditional pressure CLT, we discovered that approximately 100 patients had a negative-volume CLT despite having a positive-pressure CLT. Only four of these patients developed post-extubation stridor and only one required reintubation. Also, we found that the positive predictive value for our volume CLT is only about 50%, whereas our negative predictive value is 95%. Therefore, we have become more liberal with extubation trials in our unit despite a poor volume CLT, although this is done with an anesthesiologist at the bedside as a precaution.

The conditions under which the CLT is performed are important. Originally, the CLT was devised by anesthesiologists and done under general anesthesia when both the laryngeal and abdominal muscles are relaxed. This is important because laryngeal muscle adduction may produce a false-positive result, whereas forced expiration potentially may produce a false-negative result. When the CLT is done in sedated, relaxed patients test results are highly reproducible with little inter-observer variability.3 Clinicians are reluctant to sedate a patient just prior to extubation to perform a CLT. We have largely circumvented this by performing the CLT during the night prior to a planned extubation when extra sedation can be given. For patients who require 2 or more days of mechanical ventilation, it seems unlikely that upper airway edema would suddenly develop within a few hours prior to extubation.

The take-home message is that the volume CLT has a high negative predictive value and therefore is a useful screening tool for reliably predicting the absence of stridor. A positive test result is sufficiently sensitive to warrant more thorough airway evaluation prior to extubation. It behooves clinicians to extubate patients with a positive CLT under close scrutiny with reintubation equipment readily available at the bedside. Finally, from my experience I concur with the authors that because of the low incidence and known risk factors associated with post-extubation stridor, the CLT should be performed only in patients with significant or multiple risk factors.


  1. Miller RL, Cole RP. Association between reduced cuff leak volume and post-extubation stridor. Chest 1996;110:1035-1040.
  2. Jaber S, et al. Post-extubation stridor in intensive care unit patients. Risk factors, evaluation and importance of the cuff-leak test. Intensive Care Med 2003;29:69-74.
  3. Pettignano R, et al. Is the leak test reproducible? South Med J 2000;93:683-685.