Pre-extubation ultrasonographic measurement of intracricoid peritubal free space: A pilot study to predict post-extubation airway obstruction in children

Objective: Post-extubation airway obstruction (PEAO) is common and difficult to predict in children. We hypothesized that Intracricoid Peritubal Free Space (IPFS) obtained by deducting the outer diameter of the endotracheal tube in situ (ODTT - provided by the manufacturer) from the ultrasonographically measured internal transverse cricoid diameter (ICDt) is likely to be inversely proportional to the risk of developing PEAO. This prospective observational study was planned to evaluate this hypothesis. Methods: This study was conducted in a Pediatric Intensive Care Unit of a tertiary care teaching hospital in a low-middle income economy. Laryngotracheal ultrasound was performed just prior to the first elective extubation in 93 patients (3mo-12yrs) intubated for >= 48 h, to calculate the IPFS. Patients with pre-existent upper airway conditions, chronic respiratory diseases and poor airway reflexes were excluded. Patients with Westley's Croup Score (WCS) >= 4 were classified as PEAO, and those with WCS >= 7, as extubation failure (EF). Results: Thirty-two (34%) patients developed PEAO, while seventeen (18%) developed EF. Baseline clinical characteristics were similar in patients with and without PEAO. IPFS was lesser in patients who developed PEAO (4.16 +/- 1.18 mm vs. 5.28 +/- 1.51 mm, p < 0.001) and EF (4.13 +/- 1.44 mm vs. 5.07 +/- 1.46 mm, p = 0.019) compared to those who did not. IPFS < 5.16 mm predicted PEAO [sensitivity, 84%; positive predictive value (PPV), 87%; AUC, 0.714), while IPFS < 3.77 mm predicted EF (specificity, 80%; PPV, 88%; AUC, 0.679). Combining clinical risk factors (presence of clinical edema, prolonged ventilation and younger age) and lesser IPFS helped develop a clinico-sonographic prediction model with improved predictability for PEAO and EF (AUC, 0.820 for both). Conclusions: Lesser IPFS is reasonably sensitive and specific to predict PEAO and EF respectively with high PPV. Combining clinical risk factors and IPFS improved the PPV further. Further studies with larger samples stratified for different age groups in different clinical settings are required to confirm these observations.