Abstract
Background:
Cuff leak tests are done to evaluate the risk for post-extubation stridor. Miller and Cole reported that differences between inspired-expired tidal volume (VT) during cuff deflation while on volume control ventilation produced high positive predictive value (0.80) and specificity (0.98) using a cut-off of <110 mL in a medical cohort. 1 However, ventilators differ in reporting inspired and expired VT because they vary in compensating for the effects of heated humidification and circuit compression. 2 Therefore, we simplified Miller and Cole's method by only monitoring differences in expired VT between conditions of cuff inflation and deflation and using a similar cuff-value (100 mL). We speculated that this variation would still produce useful information in assessing stridor risk.
Methods:
In 2010 we initiated this procedure as part of systematic screening of all mechanically-ventilated patients at high risk for stridor. Test ventilator settings were standardized (Volume Control, VT: 8-10 mL/kg, constant flow: 50-60 L/min, inspiratory time: 1-1.5 s). Subjects had their upper airways suctioned prior to cuff deflation and were sedated for the procedure. Expired VT was recorded beginning 6 breaths after cuff deflation; the subsequent 3 breaths were averaged. Extubation trials occurred within 1-12 h following the test. Data was analyzed using Fisher Exact test with alpha set at 0.05.
Results:
Between May 2010 and February 2018, 756 cuff leak tests were performed; followed by a trial of extubation. The majority (73%) were either trauma/surgical or neurocritical care subjects. A cut-off for expired VT of < 100 mL had a high likelihood of association with subsequent stridor despite a moderate positive predictive value (Table). The incidence of post-extubation stridor was 57/756 (7.5%); similar to that reported by Miller and Cole.
Conclusions:
A simplified volume-based cuff leak test using only expired VT and a cut-off value of 100 mL was very reliable in predicting the absence of post-extubation stridor. Although leaks of < 100 mL were associated with a much higher likelihood for developing stridor, the condition was absent in 55% these subjects.
1. Miller RL, Cole RP. Association between reduced cuff leak volume and postextubation stridor. Chest 1996;110(4):1035-1040. 2. Phillips J, Pangilinan L, Mangalindan E, et al. Ventilator compression volume compensation does not maintain pre-set tidal volume. Respir Care 2016;61(10):OF9.
Key: CI = confidence interval, *P<0.0001 View all access options for this article.
Sensitivity [95% CI]
0.47 [0.34-0.61]
Specificity [95% CI]
0.95 [0.93-0.97]
Positive Predictive Value [95%CI]
0.45 [0.32-0.58]
Negative Predictive Value [95%CI]
0.96 [0.94-0.97]
Likelihood Ratio
10.0*
Get full access to this article