Author: Thomas Rosenthal
Because of its high specificity, a cuff leak test is better at ruling in than ruling out post-extubation airway obstruction in adults who are mechanically ventilated, a meta-analysis has concluded.
The analysis of 21 studies included 3,772 patients and was presented at the 2018 annual meeting of the Society of Critical Care Medicine (abstract 19). Although it has excellent specificity, the moderate sensitivity of the cuff leak test (CLT) may mean other options with higher sensitivity to predict post-extubation airway obstruction may be preferred.
“Our systematic review of 21 diagnostic accuracy studies found that the sensitivity and specificity of the cuff leak test to predict post-extubation laryngeal edema was 69% and 89%, respectively,” said Akira Kuriyama, MD, MPH, Emergency and Critical Care Center, Kurashiki Central Hospital, in Kurashiki, Japan, and the study’s lead author.
“Including other diagnostic accuracy parameters, the cuff leak test has an overall diagnostic accuracy,” Dr. Kuriyama said. “The cuff leak test can predict post-extubation laryngeal edema in the presence of cuff leak, given excellent sensitivity. However, the test may not be able to rule it out well in the absence of cuff leak, given the not-so-good sensitivity.”
Dr. Kuriyama and his co-researcher Jeffrey Jackson, MD, MPH, a professor of medicine at Medical College of Wisconsin in Milwaukee, said the 2017 American Thoracic Society (ATS) and American College of Chest Physicians (ACCP) clinical practice guideline recommends performing a CLT when liberating critically ill patients from mechanical ventilation to separate those at high risk for post-extubation airway events because prophylactic corticosteroids are indicated for these patients (Chest 2017;151:166-180).
“This statement was based on two systematic reviews on the diagnostic accuracy of the cuff leak test that were published over seven years ago,” Drs. Kuriyama and Jackson said. “To date, several new studies have been published.” They updated a meta-analysis of the diagnostic accuracy of the CLT to predict post-extubation airway obstruction in adults in their analysis of studies in PubMed, EMBASE and the Cochrane Central Register of Controlled Trials.
Best Use of the Test
The new meta-analysis found that the pooled sensitivity was 0.70 (95% CI, 0.52-0.83) and the specificity was 0.89 (95% CI, 0.85-0.92), with a positive likelihood ratio of 6.33 (95% CI, 4.45-9.03) and a negative likelihood ratio of 0.34 (95% CI, 0.21-0.56). The diagnostic odds ratio was 18.52 (95% CI, 9.06-37.86), according to the researchers.
“The authors’ analysis revealed that the CLT is a very good to excellent test, as reflected by the area under the summary receiver operating characteristic test of 0.9 and diagnostic odds ratio of 18,” said Curtis N. Sessler, MD, the Orhan Muren Professor of Medicine and the director of the Center for Adult Critical Care Virginia Commonwealth University Health System, in Richmond, who was not associated with the study and was asked to comment.
“However, they correctly point out that the test is no more than 70% sensitive and has a relativeweak negative likelihood ratio of about 0.4,” Dr. Sessler said. “Pooled results of prospective studies indicate that only about one in 14 unselected intubated ICU patients will develop post-extubation laryngeal edema and only one in 28 will require reintubation due to upper airway obstruction.
“Accordingly, applying a very good test to a group of patients who have a low likelihood of having the event … can result in some patients remaining intubated to receive corticosteroids who don’t really need them,” Dr. Sessler said. “Restricting the use of the CLT to patients at higher risk of post-extubation laryngeal edema should reduce the likelihood that patients with a falsely abnormal CLT (no leak) will be kept intubated unnecessarily.”
Dr. Sessler said the clinical risk factors that were identified in observational studies include female sex, duration of intubation of more than six days, larger-diameter endotracheal tube, traumatic intubation and previous unplanned extubation.
“Post-extubation laryngeal edema is what we would like to avoid and, given the moderate sensitivity of the cuff leak test, we may need to seek some other options to rule it out,” Dr. Kuriyama said. “However, the cuff leak test is still an important test for clinicians … involved in critical care.”
“Given the considerable burden associated with extubation failure, it is useful and reasonable for a cuff leak test to rule in high-risk patients to further prepare for a safe extubation,” the researchers concluded. “Our study thus supports the idea of performing a cuff leak test, as stated in the current guideline.”
Dr. Sessler said the CLT was an inexpensive, simple and important test to identify patients at high risk for developing post-extubation laryngeal edema. “Post-extubation laryngeal edema can result in rapid progression to acute respiratory failure due to upper airway obstruction—a potentially dangerous situation for which clinicians should be prepared with expertise and equipment to handle a difficult airway,” he said. “Additionally, pre-extubation treatment of high-risk patients with short-duration, moderate-dose corticosteroids significantly reduces the incidence of laryngeal edema.”
Dr. Sessler noted that the recent guidelines from CHEST and ATS indicate as a “weak recommendation” that a CLT be performed in intubated adults who meet extubation criteria and are at high risk for post-extubation stridor.
“What we need is a multicenter, randomized controlled trial that compares a strategy of CLT and pre-extubation corticosteroids with conventional practice and focuses on total duration of mechanical ventilation, as well as post-extubation laryngeal edema and safety,” Dr. Sessler said.